Cross-Platform Release Notes for Cisco IOS Release 12.2SRE

---

Resolved Caveats—Cisco IOS Release 12.2(33)SRE15a

Table 1 Resolved Caveats for Cisco IOS Release 12.2(33)SRE 15a

 

Resolved Caveats—Cisco IOS Release 12.2(33)SRE15

Table 2 Resolved Caveats for Cisco IOS Release 12.2(33)SRE 15

 

Resolved Caveats—Cisco IOS Release 12.2(33)SRE14

Table 3 Resolved Caveats for Cisco IOS Release 12.2(33)SRE 14

 

Resolved Caveats—Cisco IOS Release 12.2(33)SRE13

Table 4 Resolved Caveats for Cisco IOS Release 12.2(33)SRE 13

 

Resolved Caveats—Cisco IOS Release 12.2(33)SRE12

Table 5 Resolved Caveats for Cisco IOS Release 12.2(33)SRE12

Resolved Caveats—Cisco IOS Release 12.2(33)SRE11

Table 6 Resolved Caveats for Cisco IOS Release 12.2(33)SRE11

Resolved Caveats—Cisco IOS Release 12.2(33)SRE10

• CSCtk61069

Symptoms: The Cisco IOS router crashes.

Conditions: This symptom occurs while performing “write memory” or “show running configuration” on the router after configuring “privilege exec level 15 show adjacency”.

Workaround: Do not set the privilege exec level for any form of the show adjacency command.

• CSCts39290

Symptom: Under runs counter increments report for native Gigabit interfaces of NPE-G1.

Conditions: This symptom is not observed any specific conditions.

Workaround: This issue can be resolved by clearing the interface.

• CSCue00996

Symptom: The Cisco IOS Software implementation of the Network Address Translation (NAT) feature contains two vulnerabilities when translating IP packets that could allow an unauthenticated, remote attacker to cause a denial of service (DoS) condition.

Cisco has released free software updates that address these vulnerabilities.

There are no workarounds to mitigate these vulnerabilities.

This advisory is available at the following link:

http://tools.cisco.com/security/center/content/CiscoSecurityAdvisory/cisco-sa-20140326-nat

---

Note The March 26, 2014, Cisco IOS Software Security Advisory bundled publication includes six Cisco Security Advisories. All advisories address vulnerabilities in Cisco IOS Software. Each Cisco IOS Software Security Advisory lists the Cisco IOS Software releases that correct the vulnerability or vulnerabilities detailed in the advisory as well as the Cisco IOS Software releases that correct all Cisco IOS Software vulnerabilities in the March 2014 bundled publication.

Individual publication links are in Cisco Event Response: Semiannual Cisco IOS Software Security Advisory Bundled Publication at the following link:

http://www.cisco.com/web/about/security/intelligence/Cisco_ERP_mar14.html

Conditions: See published Cisco Security Advisory

Workaround: See published Cisco Security Advisory

PSIRT Evaluation: The Cisco PSIRT has assigned this bug the following CVSS version 2 score. The Base and Temporal CVSS scores as of the time of evaluation are 7.1/5.9:

https://intellishield.cisco.com/security/alertmanager/cvssCalculator.do?dispatch=1&version=2&vector=AV:N/AC:M/Au:N/C:N/I:N/A:C/E:F/RL:OF/RC:C

CVE ID CVE-2014-2111 has been assigned to document this issue.

Additional information on Cisco’s security vulnerability policy can be found at the following URL:

http://www.cisco.com/en/US/products/products_security_vulnerability_policy.html

• CSCug84789

Symptom: A vulnerability in the Cisco 7600 Series Route Switch Processor 720 with 10 Gigabit Ethernet Uplinks models RSP720-3C-10GE and RSP720-3CXL-10GE could allow an unauthenticated, remote attacker to cause the route processor to reboot or stop forwarding traffic. The vulnerability is due to an issue in the Kailash field-programmable gate array (FPGA) versions prior to 2.6.

Cisco has released free software updates that address this vulnerability.

Workarounds that mitigate this vulnerability are not available.

This advisory is available at the following link:

http://tools.cisco.com/security/center/content/CiscoSecurityAdvisory/cisco-sa-20140326-RSP72010GE

---

Note The March 26, 2014, Cisco IOS Software Security Advisory bundled publication includes six Cisco Security Advisories. All advisories address vulnerabilities in Cisco IOS Software. Each Cisco IOS Software Security Advisory lists the Cisco IOS Software releases that correct the vulnerability or vulnerabilities detailed in the advisory as well as the Cisco IOS Software releases that correct all Cisco IOS Software vulnerabilities in the March 2014 bundled publication.

Individual publication links are in Cisco Event Response: Semiannual Cisco IOS Software Security Advisory Bundled Publication at the following link:

http://www.cisco.com/web/about/security/intelligence/Cisco_ERP_mar14.html

Conditions: See published Cisco Security Advisory

Workaround: See published Cisco Security Advisory

PSIRT Evaluation: The Cisco PSIRT has assigned this bug the following CVSS version 2 score. The Base and Temporal CVSS scores as of the time of evaluation are 7.1/5.9:

https://intellishield.cisco.com/security/alertmanager/cvssCalculator.do?dispatch=1&version=2&vector=AV:N/AC:M/Au:N/C:N/I:N/A:C/E:F/RL:OF/RC:C

CVE ID CVE-2014-2107 has been assigned to document this issue.

Additional information on Cisco’s security vulnerability policy can be found at the following URL:

http://www.cisco.com/en/US/products/products_security_vulnerability_policy.html

• CSCuh45042

Symptom: Traffic on some GIG subinterfaces are seen to be dropped at the SPA. The SPA TCAM is seen to have two entries sharing the same logical address as a result of which one entry is seen to overwrite the other.

Conditions: This symptom was observed after a router/LC/SPA reload. The exact condition that triggers this symptom is not known.

Workaround: There is no workaround.

• CSCuh91645

Symptom: WS-SUP720-3B crashes while receiving DHCP packets.

Conditions: This symptom occurs with the ip dhcp relay information policy-action encapsulate command.

Workaround 1. Use the ip dhcp relay information policy-action replace command.

Workaround 2. Use the no ip dhcp relay information trusted command.

• CSCui29745

Symptom: Member links under MLPPP go down as the LCP negotiation of those PPP links fails.

Conditions: This symptom occurs after the router reloads and the traffic is flowing through the multilink.

Workaround: Reload SPA/LC on the other end of the link.

• CSCui65914

Symptom: Minor or major temperature alarms are reported in the syslog along with the following DATACORRUPTION logs:

Conditions: The symptom is observed on ES+ series linecards of Cisco 7600 Series Routers.

Workaround: There is no workaround.

• CSCuj30702

Symptom: This bug is specific to port channel sub interface configuration in ES+ card. This bug is not relevant to any other port channel configuration in ES+, that is, EVC/Bridge-Domain over PoCH sub-int etc, and other card types, such as ES20/ LAN cards are free from this bug. Any type of IP communication on port channel sub interfaces in ES+ cards fail. Such an issue is seen only with port channel sub interfaces on ES+ and not seen with port channel main interfaces.

Conditions: This symptom will only be seen with images where the fix of CSCuh40617 is integrated.

Workaround: The connections will work fine if it is moved to the main interface or by using EVC BD configurations.

• CSCuj82897

Symptom: The “control-word” length is not set properly for small HDLC packets running over HDLC AToM VC with SIP-200. For example: SPA-8XCHT1/E1.

Conditions: This symptom occurs when HDLC AToM VC with SIP-200 is deployed, for example, SPA-8XCHT1/E1, will result in a packet length mismatch issue or dropping by the remote PE router when HDLCoverMPLS runs over the Ethernet link adding an additional padding which cannot be classified at all.

Workaround: Use SIP-400.

• CSCul27327

Symptom: On the Cisco c7600 router, if PIM is configured on the port-channel and on the port members, any failure on one of the port members will disable the FE CAM.

Conditions: This symptom occurs when PIM is configured on the port members.

Workaround: Perform the following workaround:

1. Do not configure PIM sparse-mode on the port members even though the CLI is accepted.

2. In case the PIM sparse-mode is configured on the port members, remove it from the port members and the port-channel and then reapply the PIM configuration on the port-channel only.

Further Problem Description: A similar issue (CSCtf75608) is seen on the Cisco Catalyst 6500 Series Switches, but the workaround is to configure PIM on the port-channel and the port members to avert the FE CAM to be disabled in the event of one of the port members failing.

• CSCul52239

Symptom: Multicast traffic might get affected after an interface delete and reconfiguration. This is more likely to happen in dot1q sub-interfaces in ES+ and specifically only if the delete and reconfiguration of the interface is done within 30 seconds.

Conditions: This symptom occurs in Cisco IOS Release 12.2SREx and Cisco IOS 15S based releases.

Workaround: Perform interface delete and reconfiguration with a time gap of one minute.

Further Problem Description: To check whether the issue is hit:

Note down the interface’s internal vlan: PE2#sh vlan int usage | i GigabitEthernet2/24.904 2000 GigabitEthernet2/24.904

Get to SP console and do “sh fid start <internal vlan> end <internal vlan>”

PE2-sp#sh fid start 2000 end 2000 FID Id Protocol Bkt Enabled FE CAM Enabled Vlan Don’t Learn Age group ------ -------------------- -------------- ---- ----------- --------- 2000 no no 2000 yes 0x00

The issue is hit if “FE CAM Enabled” bit is set to “no”.

• CSCul65614

Symptom: The FAN-MOD-6SHS module consumes more power than expected (should be around 180W).

Conditions: This symptom occurs when the ES+ Combo card is placed in slot-1 of 7600 chassis.

Workaround: Place ES+ Combo cards in any other slot other than slot-1 of 7600 chassis.

• CSCum16315

Symptom: Upon reload of a Cisco 7600 router configured with a CoPP policy containing IPv6 ACLs and DSCP matching, the CoPP is only applied to the active RSP as shown below.

After reload:

CoPP is applied to only the active RSP/SUP after reload:

When this issue is triggered, the following error will be seen in the logs:

This issue potentially exposes the device to a DoS vulnerability.

Conditions: This symptom occurs under the following conditions:

1. 7600 HA Environment.

2. CoPP IPV6 ACL with DSCP match.

3. Reload or Switchover.

Workaround: There are two workarounds for this issue.

1. Modify the CoPP Policy to remove IPV6 ACL/DSCP matching.

2. Remove and reapply the CoPP configuration as shown below:

CoPP is applied to all modulues as required:

PSIRT Evaluation: The Cisco PSIRT has evaluated this issue and does not meet the criteria for PSIRT ownership or involvement. This issue will be addressed via normal resolution channels.

If you believe that there is new information that would cause a change in the severity of this issue, please contact psirt@cisco.com for another evaluation.

Additional information on Cisco’s security vulnerability policy can be found at the following URL:

http://www.cisco.com/en/US/products/products_security_vulnerability_policy.html

Resolved Caveats—Cisco IOS Release 12.2(33)SRE9a

Cisco IOS Release 12.2(33)SRE9a is a rebuild release that addresses a critical issue for Cisco IOS Release 12.2(33)SRE.

• CSCuj30702

Symptom: This bug is specific to port channel sub interface configuration in ES+ card. This bug is not relevant to any other port channel configuration in ES+, that is, EVC/Bridge-Domain over PoCH sub-int and other card types, such as ES20/ LAN cards, are free from this bug. Any type of IP communication on port channel sub interfaces in ES+ cards fails. Such an issue is seen only with port channel sub interfaces on ES+ and not seen with port channel main interfaces.

Conditions: This symptom will only be seen with images where the fix of CSCuh40617 is integrated.

Workaround: The connections will work fine if it is moved to the main interface or by using EVC BD configurations.

Open Caveats—Cisco IOS Release 12.2(33)SRE9

Cisco IOS Release 12.2(33)SRE9 is a rebuild release for Cisco IOS Release 12.2(33)SRE.

• CSCtn00145

Symptom: Standby sup reloads on issuing the no ip route command.

Conditions: This symptom is observed when static route statements present on the active and is missing on the stand-by. Perform these steps to recreate the issue:

1. Configure a static route with an interface dependency (eg: ip static route 1.1.1.0 255.255.255.0 eth 0/0).

2. Shutdown the Hardware Module/SPA corresponding to that interface. When the hardware module/SPA corresponding to that interface is shutdown, the static route entry is hidden from the configuration.

3. Reload the stand-by. When stand-by comes up, it performs a bulk sync to the running configuration of master. However the static routes which are HIDDEN, won't be present in the running configuration of master and hence will not be created in stand-by.

4. Bring up the hardware module/SPA. As this is still treated as OIR, the HIDDEN flag will be removed from the static routes and other related configurations will now be part of running configurations in the master. However these static routes are not present in stand-by as these information is lost due to reload and there will not be a disclosure of routes in standby and standby will be missing these static routes.

5. User executes no ip route. The route will be removed from master but as standby doesn't have these routes, it will result in a PRC failure leading to a standby reload.

6. Stand-by comes up after the reload. Now, it will have the entire configuration along with static routes in sync as the unhidden static routes are now part of the running configurations that are synced to standby

Workaround: Reloading the standby (most likely in a maintenance window) is the only way to sync missing ip route statements.

Resolved Caveats—Cisco IOS Release 12.2(33)SRE9

Cisco IOS Release 12.2(33)SRE9 is a rebuild release for Cisco IOS Release 12.2(33)SRE. The caveats in this section are resolved in Cisco IOS Release 12.2(33)SRE9 but may be open in previous Cisco IOS releases.

• CSCsl38246

Symptom: When console logging is turned on, a flood of the messages shown below can potentially lead to watchdog invocation and a subsequent crash:

Conditions: A single-bit correctable error is detected on a CPU read from DRAM. As long as the errors remain correctable, and the performance of the processor does not deteriorate, the module is usable.

Workaround: Since this is a parity error you can prevent the issue from happening in the future by reseating the module. If the issue still persists after reseating the module then we may be facing a hardware issue.

• CSCsv74508

Symptom: If a linecard is reset (either due to an error or a command such as hw-module slot reload) at the precise time an SNMP query is trying to communicate with that linecard, the RP could reset due to a CPU vector 400 error.

Conditions: This symptom occurs when the linecard is reset(either due to error or a command such as hw-module slot reload) at the precise time an SNMP query is received.

Workaround: There is no workaround.

• CSCtb34814

Symptom: The following error message is reported just before a crash:

There may not be any tracebacks given for the crash.

Conditions: This symptom is observed under normal conditions.

Workaround: There is no workaround.

• CSCtd72271

Symptom: On an L3 port-channel with member interfaces of a 7600-ES+4TG3CXL module, after a member interface has recovered from a failure or when a shut/no shut is performed, this member interface starts flapping and the output traffic is dropped.

Conditions: This symptom occurs when a service policy is attached or VRF is configured on a port-channel interface and there are two members on it.

Workaround: Remove the service policy from the port-channel interface

• CSCtf31377

Symptom: IOS crashes due to processor pool memory corruption.

Conditions: This symptom occurs due to processor pool memory corruption. IOS generates one or more CLUE memory error messages similar to the following messages:

This issue could also be seen on LAN cards of a Cisco 7600 router.

Workaround: There is no workaround.

• CSCtj29261

Symptom: Customer received the following error messages on a Cisco ASR router:

Conditions: This issue can occur when the Cisco ASR 1000 router receives STP packet because it is not yet supported on Cisco ASR 1000.

Workaround: Stop sending STP packet.

• CSCtj61284

Symptom: NAT overload does not work for non-directly connected destinations in MPLS-VPN configurations.

Conditions: The symptom is observed with NAT overload configured to NAT traffic coming over an MPLS VPN to internet (via a VRF-enabled interface).

Workaround: There is no workaround.

• CSCtk32299

Symptom: SVI state remains down.

Conditions: This symptom occurs when scaled MTP and SVI configurations are copied to the router.

Workaround: Unconfigure and configure MTP and SVI configurations.

• CSCty12641

Symptom: CFM ethernet ping fails with 7600 as the remote MEP end.

Conditions: This symptom is observed after remote CFM over Xconnect MEPs with MEPs terminating on 7600 and having ESM20 LC.

Workaround: There is no workaround. Consider using ES+LC.

• CSCue01146

Symptom: SNMP GET fails for VPDN-related MIB.

Conditions: This symptom occurs while receiving an SNMP GET for the MIB before all VPDN configurations are applied.

Workaround: Reload the router.

• CSCue05492

Symptom: The DHCP snooping client ignores the IPC flow control events from CF.

Conditions: This symptom is observed when CF gives flow control off event and the DHCP snooping client does not handle it.

Workaround: There is no workaround.

• CSCue44554

Symptom: Traffic stops forwarding over port-channels configured with FAST LACP after an RP switch over.

Conditions: This symptom occurs after an RP fail over.

Workaround: A shut/no shut interface will help recover.

• CSCue59592

Symptom: Multiple crashes observed with the following tracebacks after upgrading the Cisco IOS Release from 12.2(33)SRC1 to 12.2(33)SRE6:

Conditions: This symptom is observed with a combination of BGP VPNv4 prefixes + PBR enabled on the interface for the VRF and during upgrade of image or reload of the device. If “mls mpls recirc agg” is enabled in global mode, then this crash will not be observed.

Workaround: Enable “mls mpls recirc agg” in global mode.

• CSCue86845

Symptom: An unexpected behavior caused with Ingress QoS, caused by commit CSCuc01040.

Conditions: This symptom is observed with Ingress QoS, caused by commit CSCuc01040.

Workaround: There is no workaround.

• CSCue94653

Symptom: When the port-security configured interface goes to blocking state (MST), the VLANs configured on the port go to not-forwarding state temporarily. The secure mac-addresses are not added back resulting in loss of traffic.

Conditions: The symptom is observed when the port-security configured interface goes to blocking state.

Workaround: Shut and no shut the port-security interface to re-add the mac-addresses.

• CSCuf17009

Symptom: With PIM enabled on a P2P GRE tunnel or IPSec tunnel, the SP of the Cisco 7600 series router might crash.

Conditions: This symptom occurs when there are more number of tunnels going via the same physical interface. This issue is seen in Cisco IOS SREx and Cisco IOS 15.S based releases only.

Workaround: There is no workaround.

• CSCuf64313

Symptom: Linecard crash is seen with machine-check exception.

Conditions: There is no trigger. The crash is random.

Workaround: There is no workaround.

• CSCuf73798

Symptom: Tracebacks are seen when Lc/Sp is comes up.

Conditions: This symptom occurs in Cisco IOS Release 15.3(3)S.

Workaround: There is no workaround.

More Info: This issue was not reproducible. So a code walk-through and some safe checks and code strengthening was done to avoid such an error.

• CSCuf81275

Symptom: Some ISG sessions do not pass traffic.

Conditions: This symptom is observed when you have more than one Line Card for the ISG sessions.

Workaround: There is no workaround.

• CSCug23348

Symptom: The “mod” value in the SSRAM may be inconsistent to the number of ECMP paths.

Conditions: This occurs with ECMP TE tunnels with tunnel mpls traffic-eng load-share value commands configured.

Workaround: Remove the tunnel mpls traffic-eng load-share value commands from the TE tunnels.

• CSCug34485

Symptom: Multiple Cisco products are affected by a vulnerability involving the Open Shortest Path First (OSPF) Routing Protocol Link State Advertisement (LSA) database. This vulnerability could allow an unauthenticated attacker to take full control of the OSPF Autonomous System (AS) domain routing table, blackhole traffic, and intercept traffic.

Conditions: The attacker could trigger this vulnerability by injecting crafted OSPF packets. Successful exploitation could cause flushing of the routing table on a targeted router, as well as propagation of the crafted OSPF LSA type 1 update throughout the OSPF AS domain.

To exploit this vulnerability, an attacker must accurately determine certain parameters within the LSA database on the target router. This vulnerability can only be triggered by sending crafted unicast or multicast LSA type 1 packets. No other LSA type packets can trigger this vulnerability.

OSPFv3 is not affected by this vulnerability. Fabric Shortest Path First (FSPF) protocol is not affected by this vulnerability.

Workaround: Cisco has released free software updates that address this vulnerability. Workarounds that mitigate this vulnerability are available. This advisory is available at the following link: http://tools.cisco.com/security/center/content/CiscoSecurityAdvisory/cisco-sa-20130801-lsaospf.

PSIRT Evaluation: The Cisco PSIRT has assigned this bug the following CVSS version 2 score. The Base and Temporal CVSS scores as of the time of evaluation are 5.8/5.8: https://intellishield.cisco.com/security/alertmanager/cvssCalculator.do?dispatch=1&version=2&vector=AV:N/AC:M/Au:N/C:N/I:P/A:P/E:H/RL:U/RC:C

CVE ID CVE-2013-0149 has been assigned to document this issue. Additional information on Cisco’s security vulnerability policy can be found at the following URL: http://www.cisco.com/en/US/products/products_security_vulnerability_policy.html

• CSCug50208

Symptom: A crash is seen due to double free of memory.

Conditions: The symptom is seen when the accept interface VLAN goes down.

Workaround: There is no workaround.

• CSCug68193

Symptom: Multicast traffic across ES+ cards stop flowing across subinterfaces.

Conditions: The symptom is observed after a linecard OIR. After the linecard comes up, multicast traffic stops flowing across subinterfaces.

Workaround: Shut/no shut the subinterface.

• CSCug94275

Symptom: ES+ card crashes with an unexpected exception to CPU: vector 200, PC = 0x0.

Conditions: The symptom is observed on the ES+ series linecards on a Cisco 7600 series router. Symptom is reported on the ES+ console and in the crashinfo file on the ES+ flash disk. It is not reported in the syslog.

Workaround: There is no workaround.

• CSCuh07349

Symptom: A Cisco 7600 Sup may crash due to SP memory corruption.

Conditions: This issue is observed on an REP enabled router, which is part of an REP segment. The exact trigger for this issue is not clear.

Workaround: There is no workaround.

• CSCuh16927

Symptom: Mac entries learned on a trunk link are flushed after removing VLANs.

Conditions: The symptom is observed when some allowed VLANs are removed on a trunk link, all mac address entries learned on this link are flushed. This is issue is specific to extended VLAN IDs.

Workaround: Executing ping to destination IP after removing VLANs will recover this condition.

• CSCuh40617

Symptom: Ping fails when “encap dot1q” is configured on an FE SPA inserted in bay 1 of flexwan.

Conditions: This symptom is observed when FE SPA is inserted in bay 1 of flexwan.

Workaround: Move the SPA to bay 0 of flexwan.

• CSCuh48840

Symptom: A Cisco Router crashes.

Conditions: This symptom is observed under the following conditions:

a. sup-bootdisk formatted and copied with big size file, like copy 7600 image file around 180M size

b. reload box, and during bootup try to write file to sup-bootdisk (SEA write sea_log.dat 32M bytes)

c. then the issue appear

– When the issue seen, check the sea_log.dat always with 0 byte

– No matter where (disk0 or bootdisk) to load image.

– No matter sea log disk to sup-bootdisk or disk0:. I reproduced the issue with “logg sys disk disk0:” config.

SEA is calling IFS API to create sea_log.dat, looks like IFS creating file hungs SP.

Workaround: There is no workaround.

Resolved Caveats—Cisco IOS Release 12.2(33)SRE8

Cisco IOS Release 12.2(33)SRE8 is a rebuild release for Cisco IOS Release 12.2(33)SRE. The caveats in this section are resolved in Cisco IOS Release 12.2(33)SRE8 but may be open in previous Cisco IOS releases.

• CSCsx46323

Symptoms: When a span monitor source or destination is a port-channel that is automatically created by a service module, the monitor configuration will be discarded on reloads. Also, for a redundant system, restarting the standby will cause the standby to continually reset, until the monitor session source/destination using the PO is removed.

Conditions: This symptom occurs when the configuration is “monitor session x source interface port-channel yyy”, where yyy is the port-channel that is automatically created by the service module.

Workaround: Remove the monitor session created on internal port-channels of service modules before any redundant SUP reset or reload. A full system reload will also cause the monitor session to be discarded and will therefore prevent a continual reload cycle of the standby.

• CSCtd87110

Symptoms: IP session fails to sync as DHCP static bindings do not exist on Standby router.

Conditions: This symptom is seen in HA setup.

Workaround: There is no workaround.

• CSCtg39957

Symptoms: Spurious memory access with MPLS-TE is enabled.

Conditions: This symptom is seen when PATH message without any Session Attribute object is being received from TE head end. Note that Cisco IOS and XR routers always send the Session Attribute object, so this is only a potential issue with other vendors which do not do this (most do).

Workaround: There is no workaround.

• CSCtg47129

The Cisco IOS Software implementation of the virtual routing and forwarding (VRF) aware network address translation (NAT) feature contains a vulnerability when translating IP packets that could allow an unauthenticated, remote attacker to cause a denial of service (DoS) condition.

Cisco has released free software updates that address this vulnerability. Workarounds that mitigate this vulnerability are not available.

This advisory is available at the following link:

http://tools.cisco.com/security/center/content/CiscoSecurityAdvisory/cisco-sa-20130327-nat

Note: The March 27, 2013, Cisco IOS Software Security Advisory bundled publication includes seven Cisco Security Advisories. All advisories address vulnerabilities in Cisco IOS Software. Each Cisco IOS Software Security Advisory lists the Cisco IOS Software releases that correct the vulnerability or vulnerabilities detailed in the advisory as well as the Cisco IOS Software releases that correct all Cisco IOS Software vulnerabilities in the March 2013 bundled publication.

Individual publication links are in “Cisco Event Response: Semiannual Cisco IOS Software Security Advisory Bundled Publication” at the following link:

http://www.cisco.com/web/about/security/intelligence/Cisco_ERP_mar13.html

• CSCtg39957

The Resource Reservation Protocol (RSVP) feature in Cisco IOS Software and Cisco IOS XE Software contains a DoS vulnerability.

Cisco has released free software updates that address this vulnerability. There are no workarounds available to mitigate this vulnerability.

This advisory is available at the following link:

http://tools.cisco.com/security/center/content/CiscoSecurityAdvisory/cisco-sa-20130327-rsvp

Note: The March 27, 2013, Cisco IOS Software Security Advisory bundled publication includes seven Cisco Security Advisories. All advisories address vulnerabilities in Cisco IOS Software. Each Cisco IOS Software Security Advisory lists the Cisco IOS Software releases that correct the vulnerability or vulnerabilities detailed in the advisory as well as the Cisco IOS Software releases that correct all Cisco IOS Software vulnerabilities in the March 2013 bundled publication.

Individual publication links are in “Cisco Event Response: Semiannual Cisco IOS Software Security Advisory Bundled Publication” at the following link:

http://www.cisco.com/web/about/security/intelligence/Cisco_ERP_mar13.html

• CSCtk15775

Symptoms: One or more prefixes are missing on line cards, switch processors and standby supervisor, compared to the active supervisor.

Conditions: This symptom occurs after one of the following occurs:

– switchover

– standby supervisor reload

– multiple line cards reloading

Then a race may be hit such that one or more prefixes are not distributed to line cards, switch processors and the standby supervisor causing traffic to be misforwarded or not at all. The issue only affects prefixes downloaded at first priority, as listed in the “show cef table download priority” output.

Workaround: The default behavior is for routes with recursive dependents and the default route are downloaded as first priority, so configuring “cef table download recursive-dependents priority 2” and “cef table download default- route priority 2” will cause such prefixes to be downloaded as second priority and the defect will not be hit. Configuring these commands may have a slight adverse effect on convergence performance.

Alternatively, the CEF consistency checker can be used to detect prefixes missing on the line cards and switch processors and automatically repair the inconsistencies. This can be configured by the following:

cef table consistency-check IPv4 type scan-lc-rp
cef table consistency-check IPv4 type scan-rp-lc
cef table consistency-check IPv4 data-checking
cef table consistency-check IPv4 error-message

If the issue is hit, then the system state can be fixed by doing: clear ip route *

• CSCtk31340

Symptoms: Cisco route processor (RP) crashes when a port-channel is removed and the member link is defaulted.

Conditions: When a port-channel is removed (no int port-channel 200) and the member link is defaulted, the port-channel does not automatically remove the configurations on the member link. This crashes the route processor.

Workaround: There is no workaround.

• CSCtk67455

Symptoms: The fragmented traffic is dropped when the LOG option is set for IPv6 ACLs on 3CXL PFC-based supervisors.

Conditions: This symptom is observed when the LOG keyword is specified for IPv6 ACLs on 3CXL PFC mode.

Workaround: There is no workaround.

• CSCtl70143

Symptoms: LAC does not forward a PPP CHAP-SUCCESS message from LNS to client sometimes.

Condition: This symptom is seen when T1/PRI is used between the client and LAC.

Workaround: There is no workaround.

• CSCtn41225

Symptoms: The IPC port disappears and error messages are displayed on the Cisco CMTS. On the Cisco 7600 platform, it causes a crash.

Conditions: This symptom is observed while doing a quit operation using the show ipc util command.

Workaround: Do not use the show ipc util command.

• CSCtq09206

Symptoms: Traffic flowing via MPLS TE tunnels gets blackholed after FRR-protected primary link flaps initiate an FRR cutover. CEF Backwalk failure messages may be observed on the SP/DFC console.

Conditions: This symptom is observed with TE/FRR configuration with node protection.

Workaround: There is no workaround.

• CSCtu42387

Symptoms: Gigabit and 10 Gigabit Fiber link reporting threshold violation alarm in Cisco ME3600, Cisco ME3800, and Cisco 7600 devices. The “%SFF8472-3-THRESHOLD_VIOLATION: Gi0/11: Rx power high alarm” error message is seen on ports.

Conditions: This symptom is observed on Cisco ME3600, Cisco ME3800, and Cisco 7600 devices. The messages are seen with the interface shut or no shut.

SFF8472-3-THRESHOLD_VIOLATION Gi5/1: Rx power low alarm; Operating value:
-28.5 dBm, Threshold value: -24.0 dBm

Workaround: Fixing the fiber signal issue or disconnecting the fiber from the transceiver has been known to stop the messages.

• CSCtw70298

Symptoms: A router crashes after the router boots up with scaled configuration and L2/L3 Distributed EtherChannel (DEC) or port-channel.

Conditions: This symptom occurs only if there are more than two or more DFCs (ES+ and other equivalents) and with L2/L3 DEC configured such that one of the DFCs takes much longer to come up than the other.

Workaround: There is no workaround.

• CSCty07558

Symptoms: DHCPv6 packets are dropped on a Cisco 7600 switch. For example, they are not flooded.

Conditions: This symptom is observed when there is no IPv6 address on an SVI or if l2 VLAN has SVI in shut state (default existence after a new ACL feature).

Workaround: Two possible workarounds which essentially serve as the fix due to the limitations they impose:

1. When working with a pure L2 VLAN, remove ttl rate limiter (selected as default rate limiter on Cisco7600, but not on other boxes) using “no mls rate-limit all ttl-failure”.

2. If the design permits and TTL rate limiter is necessary, put a dummy IPv6 address on the SVI or simply configure IPv6 enable on the SVI.

• CSCtz33778

Symptoms: MDT remains in deleted state after several successive mdt removes/adds under the VRF.

Conditions: The issue is seen when remove and add mdt is done quickly for multiple VRFs through a script.

Workaround: Remove and re-add the VRF.

• CSCtz34869

Symptoms: Aps-channel stops working.

Conditions: This symptom occurs with an open ring and is seen in the following scenario:

A1(po2)(RPL)<======>(po2)A3 (gig3/2)<=======>(gig3/3)A4

Shut down gig3/2 on A3. Does not make A1 into protection.
=> Debugs show no SF packets are being transmitted to A1 which is connected to A3 via “Port-channel”
=> A1 (po2) is RPL of the ring. It is not going to unblocked even after A3-A4 link goes down.

Workaround: Reload the line card.

• CSCtz43626

Symptoms: Minor or major temperature alarms reported in the syslog:

%C7600_ENV-SP-4-MINORTEMPALARM: module 2 aux-1 temperature crossed threshold #1(=60C). It has exceeded normal operating temperature range.

%C7600_ENV-SP-4-MINORTEMPALARM: EARL 2/0 outlet temperature crossed threshold #1(=60C). It has exceeded normal operating temperature range.

Conditions: The symptom is observed on ES+ series line cards of Cisco 7600 series routers. Specifically, the reported temperature will be far off from reading of other sensors on the line card.

Workaround: There is no workaround.

• CSCtz68466

Symptoms: Router crashes when the QoS policies are applied through radius server under multiple virtual-access interfaces, which are cloned from virtual-template configurations.

Conditions: This symptom occurs when the router is using the following configurations:

– Virtual-template settings

– L2TP sessions

– Radius server for the QoS policies

Workaround: There is no workaround.

• CSCua25671

Symptoms: After adding the source interface in RSPAN, there is huge flooding to all trunks allowing RSPAN VLAN starts, even if there is no traffic on the RSPAN source interface.

Conditions: This symptom is observed under the following conditions:

1. The router has a RSPAN source session.

2. The source interface being added to the RSPAN source session is on ES+.

3. Any of the ES+ modules in the system has an interface on the RSPAN VLAN (that is, at least one of the interfaces on an ES+ module carries RSPAN replicated traffic).

4. The online diagnostic tests, TestFabricCh0Health and TestFabricCh1Health, are enabled on the ES+ module which has 2 and 3 mentioned above.

Workaround 1: Disable the online diagnostic tests, TestFabricCh0Health and TestFabricCh1Health, on the ES+ module which has the RSPAN source.

Workaround 2: If you have to use an interface on the ES+ module as a SPAN source, make sure that no other interface on any of the ES+ modules in the system is in the RSPAN VLAN. If you have to use an interface on the ES+ module to carry RSPAN replicated traffic, make sure that no other interface on any of the ES+ modules in the system is being monitored as an RSPAN source.

• CSCua25748

Symptoms: The PW receive counter does not work.

Conditions: This symptom is observed only with the ES+ card. This issue is not seen always due to timing events.

Workaround: Flap VC again, and check if the counter works. If it does not work, reconfigure the VC.

• CSCua68243

Symptoms: IGMP and PIM control packets are not reaching RP. As a result, the mac-address table for IGMP snooping entries is not populated.

Conditions: This can be seen on a Cisco 7600 series router that is running IOS where IGMP and PIM control packets come in on an SVI only after the condition where the SVI link state goes down and comes up again. This does not affect routed ports.

Workaround: In the SVI configuration mode:

1. Unconfigure PIM by using no ip pim.

2. Unconfigure IGMP snooping by using no ip igmp snooping.

3. Re-enable both PIM and IGMP snooping.

• CSCua85239

Symptoms: Flapping BGP sessions are seen if large BGP update messages are sent out and BGP packets are fragmented because midpoint routers have the smaller “mtu” or “ip mtu” configured.

%TCP-6-BADAUTH: Invalid MD5 digest from 6.6.6.5(179)

to 2.2.2.5(17744) tableid - 0

%TCP-6-BADAUTH: Invalid MD5 digest from 6.6.6.5(179)

to 2.2.2.5(17744) tableid - 0

%BGP-5-ADJCHANGE: neighbor 6.6.6.5 Down BGP

Notification sent

%BGP-3-NOTIFICATION: sent to neighbor 6.6.6.5 4/0

(hold time expired) 0 bytes

%BGP_SESSION-5-ADJCHANGE: neighbor 6.6.6.5 VPNv4

Unicast topology base removed from session BGP Notification sent

%TCP-6-BADAUTH: Invalid MD5 digest from 6.6.6.5(179)

to 2.2.2.5(17744) tableid - 0

%BGP-5-ADJCHANGE: neighbor 6.6.6.5 Up

%TCP-6-BADAUTH: Invalid MD5 digest from 6.6.6.5(179)

to 2.2.2.5(17744) tableid - 0

%TCP-6-BADAUTH: Invalid MD5 digest from 6.6.6.5(179)

to 2.2.2.5(17744) tableid - 0

Conditions: This symptom is observed between two BGP peers with matching MD5 passwords configured and can be triggered by the following conditions:

– If the midpoint path has the “mtu” or “ip mtu” setting that is smaller than the outgoing interface on BGP routers, it will be force the BGP router to fragment the BGP packet while sending packets through the outgoing interface.

– Peering down and the MD5 error do not always occur. They occur only once or twice within 10 tests.

Workaround: There is no workaround.

• CSCub23231

Symptoms: Very specific events/packet types cause the ES20 LC to stop passing traffic. Information on these events and packets that lead to the issue are not known currently.

Conditions: This symptom occurs when the ES20 interface has an EVC or MPLS configuration.

Workaround: Reload LC.

• CSCub36356

Symptoms: Scaling up routes result in huge memory allocations, eventually depleting the SP memory, leading to MALLOC FAIL and subsequent system crash.

Conditions: This symptom occurs in normal conditions.

Workaround: There is no workaround.

• CSCub39296

Symptoms: Unexpected exception to CPU: vector 200, PC = 0x0. Traceback decode is irrelevant.

Conditions: The symptom is observed on the ES+ series line cards on a Cisco 7600 series router. Symptom is reported on the ES+ console and in the crashinfo file on the ES+ flash disk. It is not reported in the syslog.

Workaround: There is no workaround.

• CSCub47520

Symptoms: “Match dscp default” matches router initiated ARP packets.

Conditions: This issue seen on Cisco 7600 ES+ line cards.

Workaround: Classify router generated packets using source mac address using a MAC ACL.

• CSCub58312

Symptoms: When IGMP query with source IP address 0.0.0.0 is received on an interface, it is marked as mrouter port for that VLAN.

Router#show ip igmp int vlan 1

Vlan1 is up, line protocol is up

IGMP querying router is 0.0.0.0 <----

Router#sh ip igmp snooping mrouter

vlan ports

-----+----------------------------------------

1 Po1,Po8,Router<-----

Conditions: This symptom is seen when IGMP query with source IP address 0.0.0.0 is received.

Workaround: There is no workaround.

• CSCub60678

Symptoms: Standby RSP is periodically reset after memory exhaustion. This can be checked by checking free memory on standby SP by the show memory statistic command.

Conditions: This symptom is triggered by standby RSP restart or router reload.

Workaround: There is no workaround.

• CSCub87579

Symptoms: Multicast traffic gets forwarded to the wrong tunnel protected with IPsec.

Conditions: This symptom is observed when Multicast (PIM) is enabled on the GRE tunnel protected with IPsec on the Cisco 7600.

Workaround: Shut/no shut on the tunnel protected with IPsec resolves the issue.

• CSCub90038

Symptoms: A 7600 series router may log IPC and ICC error messages.

Conditions: The symptom is observed when multipoint GRE tunnels are configured in combination with a large number of VLANs. The continual buildup of IPC/ICC messages on the supervisor engine may cause TestMacNotification diagnostic failures for modules in the router.

Workaround: Remove the GRE tunnel configuration or increase the GRE keepalive timers using the command keepalive period retry on the tunnel interface.

• CSCub91428

Symptoms: Internal VLAN is not deleted even after waiting for 20 minutes, and the VLAN cannot be reused.

Conditions: This symptom is seen with any internal VLAN allocated dynamically that is not freed up after 20 minutes in the pending queue.

Workaround: There is no workaround.

• CSCub91546

Symptoms: Traffic is dropped silently on the VLAN.

Conditions: This symptom is observed when all the VLANs in the router are used (0 free VLAN). Any new internal VLAN creation will fail, and an appropriate error message is not shown.

Workaround: There is no workaround.

• CSCuc28757

Symptoms: IPv6 HbH Traffic traversing across BD SVIs will not be rate-limited by HbH rate-limiter that is configured.

Conditions: This symptom is seen when enabling HbH rate-limiter on an NP of ES+ and IPv6 HbH traffic traversing across SVIs part of EVC BD of ES+ interface.

Workaround: There is no workaround.

• CSCuc41369

Symptoms: Complete traffic loss occurs for V6 mroutes.

Conditions: This symptom occurs during deletion and addition of VRFs for the MVPNv6 inband signaling configuration.

Workaround: There is no workaround.

• CSCuc44306

Symptoms: The IPv6 HbH packets get punted to RP as a result of HbH rate-limiter not working.

Conditions: This symptom is observed when IPv6 HbH packets hit the bridged interface on SIP400/SIP200 with IPv6 HbH rate-limiter configured.

Workaround: There is no workaround.

• CSCuc46356

Symptoms: Router hangs and crashes by WDOG.

Conditions: This symptom occurs when IPv6 ACL is applied to a port-ch sub-if. The sub-if is deleted followed by deletion of the ACL.

Workaround: Delete the ACL before deleting the port-ch sub-if.

• CSCuc48162

Symptoms: EVC Xconnect UP MEP is sending CCMs when the remote EFP is shut.

Conditions: This symptom occurs when EFP is admin down.

Workaround: There is no workaround.

• CSCuc50482

Symptoms: With P2P GRE PIM enabled tunnels in Cisco 7600, high CPU might occur in standby SP. It could happen mostly when the tunnel is configured under a MVPN.

Conditions: This symptom may happen when running Cisco IOS Release 12.2(33) SRE7 and later releases.

Workaround: Boot up the standby with tunnel shut and then enable it once the standby is up.

• CSCuc60245

Symptoms: Pseudowires stop passing traffic until the LSP is reoptimized.

Conditions: This symptom is observed when pseudowires stop passing traffic until the LSP is reoptimized.

Workaround: The common fix is reoptimizing the LSP onto a new path in one or both directions.

• CSCuc72244

Symptoms: On the Cisco 7600, both sides running Cisco IOS Release SRE4, Ethernet SPA configured with “negotiation Auto” and changed to “no negotiation auto”. The interface is operating in half-duplex instead of full-duplex mode.

Conditions: This is a timing issue seen when configuring/un-configuring auto-negotiation or when doing continuous router reload.

Recovery action: Configuring “shut” and “no shut” on the interface changes the duplex state to full-duplex.

Workaround: There is no workaround.

• CSCuc85297

Symptoms: If a Cisco 7600 router’s IOS image is upgraded using the ISSU procedure, the P2P tunnel interface multicast traffic might be affected.

Conditions: This occurs with a SRE7-based release.

Workaround: Perform a shut/no shut of the P2P tunnel interface.

• CSCuc96345

Symptoms: ARP exchange between the Cisco 7600 and the client device fails. The Cisco 7600 has an incomplete ARP entry in its ARP table for the client. This issue is likely to be seen between the Cisco 7600 and other Cisco platforms with MAC address 6073.5Cxx.xxxx. The incoming ARP reply is parsed by the platform CEF as an IP packet and dropped.

The following OUIs (as of October 30, 2012) are affected: (first 3 bytes from MAC address/MAC starts with)

14-73-73

20-73-55

4C-73-67

4C-73-A5

54-73-98

60-73-5C (One of Cisco's OUI ranges)

64-73-E2

70-73-CB

8C-73-6E

98-73-C4

A0-73-32

C4-73-1E

D0-73-8E

F0-73-AE

F4-73-CA

Conditions: This symptom is observed with the EVC pseudowire and 802.1q subinterface on the same physical interface, and connectivity via the subinterface is affected.

Sample configuration:

interface TenGigabitEthernet3/1

service instance 2013 ethernet

encapsulation dot1q 411 second-dot1q 200

rewrite ingress tag pop 2 symmetric

xconnect 10.254.10.10 3350075 encapsulation mpls

interface TenGigabitEthernet3/1.906

encapsulation dot1Q 906

ip address 10.10.10.1 255.255.255.0

Workaround:

– There should be a static ARP entry on the Cisco 7600 for the client’s MAC and IP.

– Change the MAC address of client to a nonaffected OUI.

Note: This ddts is caused/exposed due to fix of CSCtc22745.

• CSCuc97711

Symptoms: After SSO, traffic on the P2P-GRE tunnel within an MVPN may be affected.

Conditions: This symptom is observed with Cisco IOS Release SREx- and RLSx-based releases.

Workaround: Shut/no shut the P2P tunnel interface.

• CSCud13862

Symptoms: The Cisco WS-SUP720 running Cisco IOS Release 12.2(33)SRE3 crashes.

Conditions: This symptom occurs during a CPU process history update.

Workaround: The issue can be avoided by removing the configuration statement for “CPU Utilization Statistics”.

conf t no process cpu statistics limit

• CSCud19230

Symptoms: ES+ line card reload occurs with the following error messages:

%PM_SCP-SP-1-LCP_FW_ERR: System resetting module 2 to recover from error:

x40g_iofpga_interrupt_handler: LINKFPGA IOFPGA IO Bus Err val: 4214784 Bus

Error Add:332 Bus Err

data: 0

%OIR-SP-3-PWRCYCLE: Card in module 2, is being power-cycled Off (Module Reset

due to exception or

user request)

%C7600_PWR-SP-4-DISABLED: power to module in slot 2 set Off (Module Reset due

to exception or user

request)

Conditions: This symptom is observed with the ES+ line card.

Workaround: There is no workaround.

• CSCud27379

Symptoms: WS-SUP720-3B running Cisco IOS Release 12.2(33)SRE4 crashes at get_alt_mod after issuing “sh run int g4/13” with several trailing white spaces until the cursor stops moving.

Conditions: This symptom occurs when you issue the show run interface command with trailing spaces until the cursor stops moving.

Workaround: Do not specify trailing spaces at the end of the show run interface command.

• CSCud28759

Symptoms: SPA crash is seen when invoking spa_choc_dsx_cleanup_atlas_ci_config with no data packed.

Conditions: This symptom is observed when the packed data size should be 1 and the status should be success.

Workaround: There is no workaround.

• CSCud71211

Symptoms: The mpls traffic-eng reoptimize timers delay cleanup command does not take effect in the path protection. When path protection kicks in and “mpls traffic-eng reoptimize timers delay installation” expires, the new best LSP is installed, but the protection path is torn down at the same time. This can cause a few seconds of packet drops, which are being carried over the protection LSP.

Conditions: This symptom occurs when the path protection switchover is triggered on the protected tunnel.

Workaround: There is no workaround.

• CSCud90950

Symptoms: Multicast traffic might not flow through when the P2P tunnel is the incoming interface in the Cisco 7600 router.

Conditions: This symptom occurs in the Cisco IOS Release 12.2SREx and Cisco IOS Release 15.0x.

Workaround: Shut and no shut of the P2P tunnel interface.

• CSCue01579

Symptoms: Receivers on slot10 - 13 of the Cisco 7613 chassis cannot receive multicast traffic when the egress replication mode is used.

Conditions: This symptom occurs on RSP720-10G + CISCO7613 chassis and when using the egress replication mode.

Workaround: Change the replication mode to ingress by using the below given CLI:

mls ip multicast replication-mode ingress

• CSCue03598

Symptoms: Carrier-delay does not work on an ES+ card under the following specific condition:

Carrier-delay configured on gig 4/13 does not work on an ES+ card when we sh down gig0/1 on peer Cisco 3560 in the below given situation:

gig4/3[no sh] gig0/1[no sh]

7600 =================== 3560

gig4/13[sh] gig0/2[no sh]

1. do [no sh] on gig 4/13

2. do [sh] on gig0/1 right after 1

gig4/1 will go up as soon as gig 4/3 gets down instead of waiting till the configured carrier-delay timer expires.

Conditions: This symptom occurs when we enter show on the peer device.

Workaround: There is no workaround.

• CSCue32450

Symptoms: Filtering based on L4 ports does not happen for redirection to CE.

Conditions: This symptom occurs when the WCCP service uses a redirect-list and this ACL has its first entry as a “deny”.

Workaround: Make the first entry in the redirect-list ACL as a “permit”.

• CSCue61286

Symptoms: RSA keys fail to replicate.

Conditions: This issue is specific to adventerprisek9 image and is seen in the following condition:

1. In HA setup, if router has RSA configuration and Active supervisor goes down, then RSA keys are not getting replicated on Standby Supervisor.

2. During ISSU upgrade, upgrade will fail because of failure of RSA key replication.

Workaround: There is no workaround.

Open Caveats—Cisco IOS Release 12.2(33)SRE7a

Cisco IOS Release 12.2(33)SRE7a is a rebuild release for Cisco IOS Release 12.2(33)SRE. The caveats in this section are open in Cisco IOS Release 12.2(33)SRE7a. This section describes only select open caveats.

• CSCuc85297

Symptoms: If a Cisco 7600 router’s IOS image is upgraded using the ISSU procedure, the P2P tunnel interface multicast traffic might be affected.

Conditions: This occurs with a SRE7-based release.

Workaround: Perform a shut/no shut of the P2P tunnel interface.

Resolved Caveats—Cisco IOS Release 12.2(33)SRE7a

Cisco IOS Release 12.2(33)SRE7a is a rebuild release for Cisco IOS Release 12.2(33)SRE. The caveats in this section are resolved in Cisco IOS Release 12.2(33)SRE7a but may be open in previous Cisco IOS releases.

• CSCtz43626

Symptoms: Minor or major temperature alarms reported in the syslog:

Conditions: The symptom is observed on ES+ series linecards of Cisco 7600 series routers. Specifically, the reported temperature will be far off from reading of other sensors on the linecard.

Workaround: There is no workaround.

• CSCua68243

Symptoms: IGMP and PIM control packets are not reaching RP. As a result, the mac-address table for IGMP snooping entries is not populated.

Conditions: This can be seen on a Cisco 7600 series router that is running IOS where IGMP and PIM control packets come in on an SVI only after the condition where the SVI link state goes down and comes up again. This does not affect routed ports.

Workaround: In the SVI configuration mode:

1. Unconfigure PIM by using no ip pim.

2. Unconfigure IGMP snooping by using no ip igmp snooping.

3. Re-enable both PIM and IGMP snooping.

• CSCub39296

Symptoms: Unexpected exception to CPU: vector 200, PC = 0x0. Traceback decode is irrelevant.

Conditions: The symptom is observed on the ES+ series linecards on a Cisco 7600 series router. Symptom is reported on the ES+ console and in the crashinfo file on the ES+ flash disk. It is not reported in the syslog.

Workaround: There is no workaround.

• CSCub74451

Symptoms: EARL inlet/outlet displays incorrect temperature values. If the temperature crosses the minor/major threshold false alarms will be generated. In case of a major alarm the linecard will shut down as a preventive measure.

Conditions: There is no trigger for the issue.

Workaround: Reload the linecard.

• CSCub90038

Symptoms: A 7600 series router may log IPC and ICC error messages.

Conditions: The symptom is observed when multipoint GRE tunnels are configured in combination with a large number of VLANs. The continual buildup of IPC/ICC messages on the supervisor engine may cause TestMacNotification diagnostic failures for modules in the router.

Workaround: Remove the GRE tunnel configuration or increase the GRE keepalive timers using the command keepalive period retry on the tunnel interface.

Resolved Caveats—Cisco IOS Release 12.2(33)SRE7

Cisco IOS Release 12.2(33)SRE7 is a rebuild release for Cisco IOS Release 12.2(33)SRE. The caveats in this section are resolved in Cisco IOS Release 12.2(33)SRE7 but may be open in previous Cisco IOS releases.

• CSCtb30811

Symptoms: The aggregate route is not removed in certain scenarios.

Conditions: This symptom is observed when all aggregated child routes are marked for deletion and the periodic function (which processes these routes to be deleted) gets to remove the route from routing table before the aggregate processing function gets a chance to process the aggregate route to which they belong. That is, the aggregate routes remain in the routing table although the child routes get deleted.

Workaround: Configuring “bgp aggregate-timer” to 0, which is the lowest value, would reduce the chances of hitting this problem considerably.

In case this issue occurs, then in order to delete the stale aggregate route, you need to configure a temporary local BGP route (say, redistribute a static route or network a loopback) with the address being a subnet of the stale aggregate address, and then remove the aggregate address and the added route. This should delete the route from the table and also send withdrawals to other routers.

Further Problem Description: The periodic function is by default called at a 60-second interval. The aggregate processing is normally done based on the CPU load. If there is no CPU load, then the aggregate processing function would be triggered within 1 second from when the tie routes are marked for deletion, and hence the chances of hitting this problem is very minimal. As the CPU load increases, this function call will be triggered at higher intervals, and if the CPU load is very high, it could go as high as the maximum aggregate timer value configured via the command. By default, this maximum value is 30 seconds and is configurable with a range of 6-60 seconds. So, if default values are configured, then as the CPU load increases, the chances of hitting this defect is higher.

• CSCtc73759

Summary: The H.323 implementation in the Cisco IOS software contains two vulnerabilities that may be exploited remotely to cause a denial of service (DoS) condition on a device that is running a vulnerable version of the Cisco IOS software.

Cisco has released free software updates that address these vulnerabilities. There are no workarounds to mitigate these vulnerabilities other than disabling H.323 on the vulnerable device.

This advisory is posted at http://www.cisco.com/warp/public/707/cisco-sa-20100922-h323.shtml.

• CSCtf71636

Symptoms: The router crashes while configuring/unconfiguring random detect.

Conditions: This symptom is observed with Cisco IOS interim Release 12.2 (31.17.01)SB. The policy given below has to be applied on FR DLCI interface for this issue to occur:

Next, the following command sequence causes the router to crash:

Workaround: There is no workaround.

• CSCth35873

Symptoms: The Subject Alternative Name (SAN) is incorrect in persistent self-signed certificates.

Conditions: This symptom occurs in persistent self-signed certificates. The SAN is displayed as DNS:secp6-11.cisco.com, but it is supposed to be DNS:zzztmp.cisco.com

Workaround: There is no workaround.

• CSCth96200

Symptoms: A continuous traceback is seen:

Conditions: This symptom occurs when the FRR feature is configured and possibly under scale conditions.

Workaround: There is no workaround.

• CSCtl01184

Symptoms: Sometimes an EVC that is configured on ES+ sends frames out with CFI bit set in the VLAN tag.

Conditions: This symptom is observed on EVCs that are configured on ES+.

Workaround: There is no workaround.

• CSCtn93891

Symptoms: Multicast traffic is getting blocked.

Conditions: This symptom occurs after SSO with mLDP and P2MP-TE configurations.

Workaround: There is no workaround.

• CSCto04463

Symptoms: A device crashes upon rebooting the routers with BGP PIC configurations for IPv6 networks. The crash is seen in both SRE code and 15.1 code.

Conditions: This symptom is observed mainly with the code-affected handles. This issue is seen with the following conditions:

1. Some error conditions by adding defensive code.

2. Proper extraction of adjacency subblock in BGP PIC cases.

Workaround: There is no workaround.

• CSCtq24526

Symptoms: A memory corruption crash is seen on a device running crypto.

Conditions: This symptom occurs on a device running crypto.

Workaround: There is no workaround.

• CSCtq77973

Symptoms: An ATM interface is created on a CEoP SPA, followed by range-pvc configurations on it. Later, when trying to delete and recreate the interface, the configuration fails with the below message.

Conditions: This symptom is not seen easily and seems to be related to the timing of events.

Workaround: Reload the SPA.

• CSCtr20762

Symptoms: L3VPN tunnel is not coming up after the router is reloaded.

Conditions: This symptom is observed with “aaa system accounting” configured and when the TACACS server is not reachable.

Workaround 1: Disable “aaa system accounting”.

Workaround 2: Ensure that the TACACS server is reachable.

• CSCtr27674

Symptoms: A SIP-200 linecard crashes.

Conditions: This symptom is observed with a POS SPA on SIP-200 line card after performing an ISSU upgrade on a Cisco 7600 series router.

Workaround: There is no workaround.

• CSCtr44384

Symptoms: Spurious memory access will be seen while deleting a subinterface with two VLAN tags.

Conditions: This symptom is observed only when a subinterface has a policy with two VLAN tags.

Workaround: There is no workaround.

• CSCts12499

Symptoms: SPA firmware crash at one bay leads to SPA crash in another bay.

Conditions: This symptom is observed when “test crash cema” is executed from the SPA console. leading to the SPA in the other bay to reload. Also, the crashinfo is not present in the RP disk.

Workaround: There is no workaround.

• CSCts20857

Symptoms: This issue is actually a fix for CSCtj96916, which is the original issue

Conditions: This symptom occurs when changing the card type from T3 to E3.

Workaround: There is no workaround.

• CSCts69115

Symptoms: Remote router (PE2) crash is seen after 500 shut/no shut of UUT primary paths.

Conditions: This symptom occurs after shut/no shut (500 times) of primary paths on UUT (PE1). The router is in hung state, and after rebooting, the crash file is not generated. From the logs, it is seen that the number of siblings in a chunk has gone above the threshold.

Workaround: Reload the router.

• CSCts81427

Symptoms: With a scaled dLFIoATM configuration on FlexWAN, after issuing SSO, some of the interfaces stop pinging.

Conditions: This symptom is observed after doing SSO.

Workaround: Shut/no shut of the ATM interface helps to resolve the problem.

• CSCts84700

Symptoms: The incoming VRF stream over the MDT data tunnel is vanishing.

Conditions: This symptom occurs when deleting and recreating the VRF.

Workaround: Introduce a delay of 6 minutes between the deletion and the recreation of the VRF.

• CSCtt99627

Symptoms: The lacp rate and lacp port priority commands may disappear following a switchover from active to standby RP.

Conditions: This symptom affects the Cisco 7600 platform.

Before performing a switchover, one may check the configuration on the standby RP to see if the commands are present or not. If the commands are not present on the standby RP then they will disappear if a switchover occurs.

Workaround: Prior to switchover, if the commands do not show up on the standby RP, as described above, then unconfiguring and reconfiguring the command on the active RP will fix the issue.

Otherwise, if the commands disappear after a switchover, then the commands must be reconfigured on the newly active RP.

• CSCtu23195

Symptoms: SNMP ifIndex for serial interfaces (PA-4T/8T) becomes inactive after PA OIR.

Conditions: This symptom is observed with a PA OIR.

Workaround: Unconfigure and reconfigure the channel-groups of the controller and reload the router.

• CSCtu90140

Symptoms: A chunk memory leak is observed.

Conditions: A chunk memory leak is seen after configuring the IP source guard.

Workaround: There is no workaround.

• CSCtw46229

Symptoms: Small buffer leak. The PPP LCP configuration requests are not freed.

Conditions: This symptom is observed with PPP negotiations and the session involving PPPoA.

Workaround: Ensure that all your PPP connections stay stable.

• CSCtw53121

Symptoms: ES+ goes into major state occasionally on reload or SSO.

Conditions: This symptom occurs in the Cisco 7600 router with a 40 gig ES+ line card that is running Cisco IOS Release 15.2(2)S.

Workaround: There is no workaround.

• CSCtw55401

Symptoms: The SPA-1XCHSTM1/OC3 card goes to out of service after SSO followed by OIR.

Conditions: This symptom is observed with the SPA-1XCHSTM1/OC3 card with Cisco 7600-SIP-200 combination.

Workaround: There is no workaround.

• CSCtw80678

Symptoms: Multilink PPP ping fails when the serial interfaces experience QMOVESTUCK error.

Conditions: This symptom may be observed if multilink PPP member links and serial interfaces on which the QMOVESTUCK error is reported are on the same SPA.

Workaround: “no shut” the interface in the QMOVESTUCK error message, remove QoS policies on the interface and subinterfaces, and remove the interface from T1/T3 controller. Then, rebuild the configuration.

• CSCtw88599

Symptoms: If “port acl” is configured, diagnostics for the port fail during bootup. If the port ACL is on a supervisor port then the router goes for a reload.

Conditions: This symptom is observed when you configure “port acl” on a switch port and reload the router.

Workaround: Disable diagnostics for the module.

This issue will effect only if there is a switchport configured on the router. The issue will not affect the traffic or the filtering based on the ACL, even if the testAclDeny fails and the card is on MajFail (due to this test only).

As a workaround, we can remove the switchport configurations for the ports (if they exist), then reload and apply the configurations after the router has come up. Alternatively, do a “no diagn crash” and try to bring up the router.

In case the router reloads, the ports will not go into shutdown state. Hence, it is a cosmetic issue. It can be ignored. If reloaded in presence of the switchport configurations, it should come up after two reloads into a minor error state.

• CSCtx23014

Symptoms: HSRP hellos cannot be sourced from certain IPs.

Conditions: This symptom is observed when HSRP hellos cannot be sourced for an IP address with a standby IP address in the same subnet and both are configured in the global VRF. For example:

Workaround: Use an IP from the subnet for the SVI interface in the same VRF.

• CSCtx23593

Symptoms: Some virtual circuit (VC) information is missing in the Simple Network Management Protocol (SNMP) MIB object cAal5VccEntry from the output of the snmpwal command, but not in the router configuration command. For example, The ATM VCs 4/0.120 exist on the router but are missing in the MIB.

Conditions: This symptom is observed on a Cisco 7204VXR (NPE-G2) router that is running 12.2(33)SRE5 (c7200p-advipservicesk9-mz.122-33.SRE5.bin) image in the customer network. This issue may also occur in other releases.

This issue typically occurs over a period of time due to create/delete of subinterfaces. It also occurs if the customer uses the snmp ifmib ifIndex Persist command, which retains ifIndicies assigned to the @~@subinterfaces across router reload.

Workaround: There can be two workarounds where there is no fix present in the Cisco IOS code for this bug.

Workaround 1:

– Enter the show atm vc privileged EXEC command on the same device to obtain a complete list of all the VCs. Or,

– Do the SNMPWALK suffixing the ifIndex of the interface to get the value.

Doing the SNMPWALK suffixing the ifindex and getting the value can be one workaround.

Workaround 2:

1. Under configuration mode: no snmp ifmib ifIndex Persist.

2. On all the ATM main interfaces: no snmp ifindex persist.

3. Save the configuration: copy running start.

4. Reload the box: reload. Reapply the persist configurations.

5. Configure in configuration mode: snmp ifmib ifIndex Persist.

6. Under the ATM main interface: snmp ifindex persist.

After this workaround, the problem may reappear over a period of time, but chances are very less.

The workaround/fix which needs to be enabled where the code fix is present in the Cisco IOS code for this bug.

Since this will go over all the possible ifIndicies, it will take more CPU cycles, causing some delay. The below global CLI can be used to enable/disable the fix based on the need.

CLI: snmp-server enable traps atm snmp-walk-serial

• CSCtx36095

Symptoms: A traceback is seen after applying DMLP configurations while doing a line card reload.

Conditions: This symptom occurs during a line card reload.

Workaround: There is no workaround.

• CSCtx48010

Symptoms: PIM neighbors on MDT tunnel continuously flap on a Cisco 7600 series router. Decapsulation path shows wrong rewrite index for flapping peers, instead of expected 7FFA recirculation index.

Conditions: This symptom is observed with Cisco IOS Release 15.1(3)S1, with ES20 card as core-facing.

Workaround: Identifying the adjacency of the flapping peer and changing the rewrite index to 7FFA manually stops the flap:

• CSCtx48473

Symptoms: A router crashes when the following command is executed:

No crashinfo is generated on the RP and SP. Please see the attached console before the crash.

Conditions: The above command must be executed.

Workaround: There is no workaround.

• CSCtx57154

Symptoms: RP crashes and brings down the router.

Conditions: This symptom is observed upon shut/no shut of an ES+ access interface configured with 5K EVC.

Workaround: There is no workaround.

• CSCtx62138

Symptoms: Standby resets continuously due to Notification timer that Expired for RF Client: Cat6k QoS Manager.

Conditions: This symptom is observed on a Cisco 7600 HA loaded with scale QoS and GRE + IPsec configurations.

Workaround: There is no workaround.

• CSCtx77501

Symptoms: Traffic is dropped at the decap side of a PE box.

Conditions: This symptom occurs with SSO at the decap side of an MVPN setup, with DFC core-facing and 6748 access facing.

Workaround: Do a switchover.

• CSCtx79462

Symptoms: OSPF neighborship does not get established.

Conditions: This symptom is observed when Enabling PFC on a multilink bundle in SIP-400. The OSPF neighborship does not get established.

Workaround: There is no workaround. Disable PFC to bring up the OSPF neighborship.

Further Problem Description: The OSPF hello packets get dropped by the peer end because the IP header is corrupted.

• CSCtx85247

Symptoms: An ES20 line card is reset on doing redundancy switchover of RSPs.

Conditions: This symptom is seen with redundancy switchover of RSPs.

Workaround: There is no workaround.

• CSCty12312

Symptoms: Multilink member links move to an up/down state and remain in this condition.

Conditions: This symptom occurs after multilink traffic stops flowing.

Workaround: Remove and restore the multilink configuration.

• CSCty13647

Symptoms: Symptoms vary from one image to another. The following symptoms have been mostly observed:

1. Spurious memory access tracebacks from SPAN code even when SPAN is not configured.

2. RP crash when unconfiguring a SPAN session with a particular session number.

Conditions: This symptom is always seen on a particular SPAN session number.

Workaround: Use a different a SPAN session number for SPAN configurations to avoid the router crash. Shutdown the SPAN session if not in use. There is no workaround to avoid spurious memory access messages.

• CSCty29230

Symptoms: CMFIB entries are not being programmed on the SP and DFCs. Mroute shows both Accept and OILS, and ip mfib output also shows Accept interface and Forwarding interface, but CMFIB entries are not programmed.

Conditions: This symptom is observed with a Cisco 7600 running a Cisco IOS Release 15.1(3)S throttle.

Workaround: There is no workaround.

• CSCty32851

Symptoms: A Cisco router may unexpectedly reload due to software forced crash exception when changing the encapsulation on a serial interface to “multilink ppp”.

Conditions: This symptom is observed when the interface is configured with a VRF.

Workaround: Shut down the interface before making the encap configuration change.

• CSCty34020

Symptoms: A Cisco 7201 router’s GigabitEthernet0/3 port may randomly stop forwarding traffic.

Conditions: This only occurs on Gig0/3 and possibly Fa0/0 as they both are based on different hardware separate from the first three built-in gig ports.

Workaround: Use ports Gig0/0-Gig 0/2.

• CSCty37396

Symptoms: The remote PE sees label 442.

However, this label is punted and passed on to the rate limiter as it is with destination index 0x7FFF, and packet drops are seen.

Conditions: This symptom occurs in a 6PE setup and when the IPV6 address is aggregated.

Workaround: Do not use the aggregate address.

• CSCty51172

Symptoms: The MAC address learned on L2 DEC on 7600-ES+40G3CXL is not installed as the primary entry on all the member interfaces, if the ingress traffic is on the nonhashed interface for that EFP.

Conditions: This symptom occurs when Layer 2 distributed Etherchannel traffic is learned on a hashed interface first and then moved to a nonhashed interface.

Workaround: Do not use Layer 2 distributed Etherchannel.

• CSCty94289

Symptoms: The drop rate is nearly 1 Mbps with priority configuration.

Conditions: This symptom is observed when traffic received in the MSFC router class-default is the same as on the other end of the MSFC2 router.

Workaround: Unconfigure the priority and configure the bandwidth, and then check for the offered rate in both the routers. This issue is only seen with the Cisco 7600 series routers (since the issue is with the FlexWAN line cards). The issue is seen with a priority configuration and does not show up when the priority is unconfigured, so there is no workaround as such for this issue otherwise.

• CSCty99331

Symptoms: CPU hog messages are seen on the console.

Conditions: This symptom is seen when applying huge rmap with more than 6k sequences on an interface.

Workaround: There is no workaround.

• CSCty99711

Symptoms: SIP-400 crash may be observed due to illegal memory access.

Conditions: This symptom is observed with Cisco IOS Release 12.2(33)SRE4 when SIP-400 has PPPoE session scale.

Workaround: There is no workaround.

• CSCtz13465

Symptoms: High CPU is seen on Enhanced FlexWAN module due to interrupts with traffic.

Conditions: This symptom is observed with an interface with a policy installed.

Workaround: There is no workaround.

• CSCtz26188

Symptoms: Packet loss is observed on platforms in certain deployments having a large number of prefixes routing traffic onto a TE tunnel.

Conditions: This symptom occurs if the configured value of the cleanup timer is 60 seconds, then packets might be lost on the platforms where the forwarding updates take longer.

Workaround: Configure the value of the cleanup timer to 300 seconds.

• CSCtz28024

Symptoms: On a TE tunnel midpoint, the incoming label shown as allocated for the LSP by the TE does not match the label allocated for the LSP in the forwarding. Packets going over the TE tunnel are dropped.

Conditions: This symptom can occur when the midpoint router receives multiple RSVP RESV messages for the same LSP and requests TE to admit the reservation. This could be due to receiving duplicate RSVP RESV messages within a very short duration or if the downstream neighbor should change certain session attributes for the RSVP RESV message. Changes to the RSVP RESV message are not a normal occurrence and this issue is not easily reproducible.

In the case observed, the downstream router was modifying the Record Route Object (RRO) unexpectedly. However, changes in the RESV outgoing label or next-hop would also trigger the problem. In the case where the issue was observed, the downstream next-hop router was a CRS running Cisco IOS XR. The conditions in which this problem was observed were during a reload of the tailend router and multiple manually triggered TE reoptimizations throughout the network. It is not known what caused the CRS to send RSVP RESV change to RRO.

Should the midpoint router process a change to an existing RSVP RESV message, TE will not handle reprogramming the forwarding properly and arrive at the problem. This problem exists in Cisco IOS Release 12.2SRE and later releases.

Workaround: Shut and no shut the tunnel.

• CSCtz30983

Symptoms: Crash on ES+ line card upon issuing the show hw-module slot X tech-support or show platform hardware version command. This is similar to CSCti78408 but not to CSCti78408.

Conditions: This symptom occurs on an ES+ line card.

Workaround: Do not issue the show hw-module slot X tech-support or show platform hardware version command on an ES line card unless explicitly mentioned by Cisco.

• CSCtz31888

Symptoms: After state change of one of the L3 uplink interfaces, STP cost of BPDU PW increases from 200 to 2M, which can lead to blocking state in STP for this PW.

Conditions: This symptom occurs with state change of one of the uplink L3 interfaces.

Workaround: Increase the cost of access ring to more then 2M to avoid blocking of the BPDU PW.

• CSCtz35040

Symptoms: The router crashes when “clear up route *” is issued.

Conditions: This symptom is observed when a router is operating few thousand BGP/IGP routes and “clear ip route *” is issued.

Workaround: There is no workaround.

• CSCtz46399

Symptoms: VLAN RAM is not programmed with the correct VPN number for a few VRFs.

Conditions: This symptom is observed with the OIR of the ES-20 line card with scaled MVPNv6 configuration and 50 VRFs configured.

Workaround: There is no workaround.

• CSCtz62680

Symptoms: “DFC WAN Line Card Software Element Not Found - QOS: TCAM Class ID” errors appear, along with BADCHUNKFREEMAGIC errors, leading to an ES20 crash.

Conditions: When service policies less than 128 kb are added or removed.

Workaround: There is no workaround.

• CSCtz66770

Symptoms: When under ATM PVC (SPA-4XOC3-ATM or v2) with MUX encapsulation and OAM enabled, L3 policy-map is applied and PVC goes down.

Conditions: This symptom occurs when policy-map sets DSCP (to 7) for default-class, and it affects OAM communication.

Workaround: Use aal5snap encapsulation.

• CSCtz86024

Symptoms: There is a long delay in joining mcast stream with Cisco IOS 15S releases that are running on RP.

Conditions: This symptom is seen when there is no (*,G) on the box, and the first packet for the stream creates this entry.

Workaround: With static joins we can make sure that entry is present in mroute table.

• CSCua09073

Symptoms: If 6708 generates txCRC errors, CSCtq73026 accounts for these errors in TestErrorMonitor diagostic test and takes the necessary recovery action. But, for CSCtq73026 to be invoked, the TestErrorMonitor test should be included in the test suite for 6708. This test is missing and hence, the fix in CSCtq73026 will also not be invoked.

Conditions: See the description of CSCtq73026. For this fix to be take effect, TestErrorMonitor should be added in the test suite. In this DDTS, this test has been added so that in case of an error, as mentioned in CSCtq73026, recovery action will be triggered.

Workaround: There is no workaround.

• CSCua10377

Symptoms: A Cisco router with Circuit Emulation SPA may suffer an SPA crash.

Conditions: This symptom occurs when the CE T1 circuit is configured by the end user for AT&T FDL, and the end user transmits FDL messages requesting 4-hour or 24-hour performance statistics.

Workaround: There is no workaround. If possible, contact the end user and have them reconfigure their device for ANSI FDL.

• CSCua16786

Symptoms: When a Cisco 7600 router is placed as a mid-hop-router between the first hop router (FHR) and rendezvous point (RP), with P2P GRE tunnel interface as the FHR facing interface, then PIM-registration might not get completed. The unicast PIM-registration packet might get dropped at the Cisco 7600 router.

Conditions: This symptom is seen in Cisco IOS Release 12.2(33)SRE6 and RLSx releases.

Workaround: Delete and create the FHR facing p2p tunnel interface at Cisco 7600 router, which is acting as mid-hop-rtr.

• CSCua18382

Symptoms: There are locally originated packets larger than 1476 bytes and are going through the IPsec tunnel with GRE.

Conditions: This symptom occurs when the Cisco 7600 router has encryption processing on the c7600-spa-ipsec-2g and is configured with the crypto engine gre vpnblade command. However, only locally originated traffic to CPE is prone to not break through traffic via the tunnel.

Workaround: Configure “crypto engine gre supervisor” instead of the VPN blade under the tunnel interface. However, this can be useful only if there is one tunnel. If there is more than one tunnel, the crypto would be offloaded to the VPN blade irrespective of the configuration.

• CSCua25943

Symptoms: CPU Hog is observed on the LC when the number of IPv6 prefixes pumped in is more than 10,000.

Conditions: This symptom is observed when more than 10,000 IPv6 prefixes are pumped into the router.

Workaround: There is no workaround.

• CSCua33287

Symptoms: ES+: L2TPv3 entries are programmed incorrectly after restart.

Conditions: This symptom is observed when some L2TPv3 sessions are established on ES+ module. After the restart of ES+, some np entries may not program correctly. As the result, ES+ will stop to transmit packets.

This condition will recover after executing shut/no shut on physical interfaces.

Workaround: There is no workaround.

• CSCua41398

Symptoms: The Cisco SUP720 crashes.

Conditions: This symptom occurs when you issue the sh clns interface | i ^[A-Z] | Number of active command multiple times via script with the following error and decodes:

Workaround: There is no workaround.

• CSCua42089

Symptoms: Configuring Ingress redirection for service group 61 (Mask) and applying an extended ACL in the outbound direction on the same interface causes software switching even when there are no punt entries in the TCAM.

Conditions: This symptom is observed when WCCP service 61 with Mask assignment in the Ingress indirection, along with an outbound ACL, is configured on the same interface.

Workaround: Do not configure the outbound ACL along with a WCCP service.

• CSCua57728

Symptoms: Traffic loss of ~25s is seen upon doing TE FRR Cutover with IPv6 prefixes.

Conditions: This symptom is observed with four core-facing tunnels, and 100,000 IPv6 prefixes. Shut the primary interface and check for the traffic loss.

Workaround: There is no workaround.

• CSCua67532

Symptoms: IPsec sessions fail to come up.

Conditions: This symptom occurs when Site-Site crypto configuration using crypto map is applied on SVI, and when no ISAKMP profile is configured under that crypto map.

Workaround: There is no workaround.

• CSCua85837

Symptoms: Depending on the aces in the redirect-list ACL used, only a subnet of the traffic gets redirected based on ports.

Conditions: This symptom is observed when the WCCP service has a redirect-list ACL applied which in turn has port-specific aces in them.

Workaround: Remove the L4 operators from the redirect-list ACL.

• CSCua98690

Symptoms: The ES+/ES20/SIP-400 (any card which supports EVC) card may crash due to memory corruption.

Conditions: This symptom is observed when the MAC ACL is configured on EFP.

Workaround: There is no workaround.

• CSCub21468

Symptoms: UDP header is corrupted randomly.

Conditions: This symptom is observed with the Cisco 7609-S (RSP720-3C-GE) running Cisco IOS Release 12.2(33)SRE5, with the VRF Aware LI feature.

Workaround: There is no workaround.

• CSCub31902

Symptoms: Alignment correction tracebacks are seen from within the diag_dump_lc_l2_table() cosmetic issue, which create temporary memory inconsistencies in the function.

Conditions: This symptom occurs in normal conditions, during bootup time, provided testMacNotification fails.

Workaround: Disable bootup diagnostics or disable the testMacNotification health monitoring test.

• CSCts70790

Symptoms: A Cisco 7600 router ceases to advertise a default route configured via “neighbor default-originate” to a VRF neighbor when the eBGP link between a Cisco 7600 router and its VRF eBGP peer flaps.

Conditions: This symptom is observed when another VPNv4 peer (PE router) is advertising a default route to the Cisco 7600 router with the same RD but a different RT as the VRF in question. When the VRF eBGP connection flaps, the VRF default is no longer advertised.

Workaround: Remove and readd the neighbor default-originate command on the Cisco 7600 router and do a soft clear for the VRF neighbor.

• CSCtw76855

Symptoms: IPC times out, and hence disconnects on line cards, resulting in MFIB table resync.

Conditions: This symptom is observed when IPC times out.

Workaround: This fix is an enhancement to avoid such unnecessary table disconnects due to IPC timeouts.

• CSCub67101

Symptoms: The POS interface line protocol is down with encapsulation PPP in an MPLS setup.

Conditions: This symptom occurs when configuring encapsulation PPP on both ends of PE1 and CE1, and then configuring xconnect in the customer-facing interface of PE1.

Workaround: Reconfigure the xconnect settings. Then, the interface will come up in the proper state.

Resolved Caveats—Cisco IOS Release 12.2(33)SRE6

Cisco IOS Release 12.2(33)SRE6 is a rebuild release for Cisco IOS Release 12.2(33)SRE. The caveats in this section are resolved in Cisco IOS Release 12.2(33)SRE6 but may be open in previous Cisco IOS releases.

• CSCee38838

Symptoms: A crashdump may occur during a two-call-per-second load test on a gateway, and the gateway may reload.

Conditions: This symptom is observed on a Cisco 3745 that runs Cisco IOS Release 12.3(7)T and that functions as a gateway when you run a two-call-per-second load test that uses H.323, VXML, and HTTP. The crash occurs after approximately 200,000 calls.

Workaround: There is no workaround.

• CSCsb53810

Symptoms: A Cisco Catalyst 6500 series switch may not block traffic, which is supposed to be denied by an outbound ACL on a VLAN interface.

Conditions: This issue is under investigation.

Workaround: Reload the switch.

• CSCsd98525

Symptoms: An SSH version 2 (SSHv2) session is terminated prematurely.

Conditions: This symptom is observed when large chunks of data are transferred in the SSHv2 session, for example, when the show tech command is entered and the command output is transferred in the SSHv2 session.

Workaround: Use SSH version 1.

• CSCsf17406

Symptoms: A memory leak is seen.

Conditions: The symptom is observed when a Cisco router performing crypto certificate revocation checks downloads a Certificate Revocation List (CRL). It is seen in the “Crypto PKI-CRL” process when the CRL is larger than roughly 70 kbs in size.

Workaround: Reduce the CRL size or disable the CRL revocation check.

Further Problem Description: Cisco IOS Release 12.3(11)T and earlier releases (including Cisco IOS Release 12.3/12.2, etc.) are not affected by this issue.

• CSCta27728

Symptoms: A Cisco router may crash.

Conditions: This symptom is observed on a Cisco ASR1002 router running Cisco IOS Release 15.1(2)S1 with RSVP for MPLS TE tunnel signaling.

Workaround: There is no workaround.

• CSCte53162

Symptoms: In radius messaging, nas-port-id is not prepended to “acct-session- id” when the nas-port format e encoding string command is configured.

Conditions: This symptom is observed when the nas-port format e encoding string command is configured.

Workaround: Use the nas-port format d encoding bits command.

• CSCtg47129

The Cisco IOS Software implementation of the virtual routing and forwarding (VRF) aware network address translation (NAT) feature contains a vulnerability when translating IP packets that could allow an unauthenticated, remote attacker to cause a denial of service (DoS) condition.

Cisco has released free software updates that address this vulnerability. Workarounds that mitigate this vulnerability are not available.

This advisory is available at the following link:

http://tools.cisco.com/security/center/content/CiscoSecurityAdvisory/cisco-sa-20130327-nat

Note: The March 27, 2013, Cisco IOS Software Security Advisory bundled publication includes seven Cisco Security Advisories. All advisories address vulnerabilities in Cisco IOS Software. Each Cisco IOS Software Security Advisory lists the Cisco IOS Software releases that correct the vulnerability or vulnerabilities detailed in the advisory as well as the Cisco IOS Software releases that correct all Cisco IOS Software vulnerabilities in the March 2013 bundled publication.

Individual publication links are in “Cisco Event Response: Semiannual Cisco IOS Software Security Advisory Bundled Publication” at the following link:

http://www.cisco.com/web/about/security/intelligence/Cisco_ERP_mar13.html

• CSCtg57657

Symptoms: A router is crashing at dhcp function.

Conditions: This issue has been seen on a Cisco 7206VXR router that is running Cisco IOS Release 12.4(22)T3.

Workaround: There is no workaround.

• CSCth10764

Symptoms: PPP Negotiation not working correctly between Cisco GSR XR and Cisco 7200.

Conditions: Max-header size different on both ends, PPP not negotiating lower size.

Workaround: There is no workaround.

• CSCth84995

Symptoms: The router may reload when performing an ISSU upgrade or downgrade.

Conditions: This symptom occurs when performing an ISSU upgrade or downgrade.

Workaround 1: Reload the router.

Workaround 2: Upgrade from Cisco IOS Release 12.2(33)SRD4 to 12.2(33)SRD5 first and then from 12.2(33)SRD5 to 12.2(33)SRE6.

• CSCti81539

Symptoms: Some of the ACLs related to TCP cannot be removed from a router.

Conditions: This symptom is observed while unconfiguring ACLs.

Workaround: Remove the entire ACL, and recreate it again.

• CSCtj84234

Symptoms: With multiple next-hops configured in the set ip next-hop clause of route-map, when the attached interface of the first next-hop is down, packets are not switched by PBR using the second next-hop.

Conditions: This symptom is seen only for packets switched in software and not in platforms where packets are PBR’d in hardware. This symptom is observed with route-map configuration, as given below:

Workaround: There is no workaround.

• CSCtk03371

Symptoms: SVI-based EoMPLS/VPLS VC fails to forward traffic even when VC is up.

Conditions: This happens when the ip cef accounting non-recursive command is configured on the router. This command is documented as an unsupported command on the Cisco 7600 platform, but it should also generate an error message when configured on the Cisco 7600. Preferably it should not take any action, for example, it should not affect any other working features.

Workaround: Unconfigure the command by typing “no ip cef accounting non- recursive”.

• CSCtk46363

Symptom: A device running Cisco IOS and acting as a DHCP server crashes.

Conditions: This symptom is observed when a client requests a specific IP address.

Workaround: Disable duplicate address detection check using the ip dhcp ping packet 0 command.

• CSCtk62763

Symptoms: A Cisco 7600 router equipped with multiple DFC line cards may experience an unexpected reload because of increased IGMP activity.

Conditions: This symptom is observed when IGMP joins and leaves (OIF churn) at approximately 160pps or more on DFCs with around 600 mroutes that have SVIs as OIFs.

Workaround: There is no workaround.

• CSCtl09030

Symptoms: The Cisco ASR1k configured to function as ISG and DHCP relay/server crashes in the ARP input process or IP inband session initiator process in dhcpd_find_binding function.

Conditions: This symptom is observed when the Cisco ASR1k is configured with DHCP relay or server and DHCP initiated IP sessions are configured. This issue is seen when the ISG inband IP session initiator is configured and an ARP request is received from a client whose DHCP IP session has timed out or cleared.

Workaround: Disable ISG DHCP session initiator.

• CSCtl86141

Symptoms: Data traffic is switched over BPDU PW. This issue can result in MAC-FLAPs over the multiple MST rings configured to be part of MST 0 instances (configured along with RL2GP) as BPDU PW acts as a loop in the network connectivity for those MST rings.

Conditions: This symptom is observed when data traffic is received on the access VLAN on which BPDU PW is configured.

Workaround: There is no workaround. This issue is not expected to be seen on the customer VLAN as data traffic is not expected over the native VLAN.

• CSCtn07696

Symptoms: The Cisco 6506-E/SUP720 may crash while redirecting the show tech-support command output using the ftp command due to TCP-2-INVALIDTCB.

Conditions: This symptom is observed with the following CLI:

During the FTP operation, if the interface fails or shuts down, it could trigger this crash.

Workaround: This is an FTP-specific issue. Redirect the output by TFTP or other protocols.

• CSCtn22523

Symptoms: IPSLA udp-jitter probes may crash at saaAddSeqnoDupQ in Cisco IOS Release 12.4T/15.0M. There is no impact to other releases.

Conditions: This symptom is observed when the network experiences delay, and reordered and duplicate packets can trigger this problem when IPSLA udp-jitter is scheduled.

Workaround: Disable udp-jitter probes.

• CSCtn65116

Symptoms: Some VPNv4 prefixes may fail to be imported into another VRF instance after a router reload or during normal operation.

Conditions: The symptom is observed with a router that is running BGP and Cisco IOS Release 12.2(33)SB or Release 12.2(33)SRB or later. Earlier versions are not affected. This occurs with the same prefixes with different mask lengths, e.g.: 10.0.0.0/24 and 10.0.0.0/26 (but not for 10.0.0.0/24 and 10.0.0.1/32, because 10.0.0.0 is not the same prefix as 10.0.0.1). It is seen with the following process:

1. Assume the prefix, 10.0.0.0/24, is imported from VPNv4 to VRF. It has been allocated a label of 16.

2. The allocated label changes from 16 to 17, e.g.: due to interface flapping or BGP attribute change.

3. However, before the BGP import happens, a more specific prefix (e.g.: 10.0.0.0/26) is added to the BGP radix tree, but it is denied for importing due to, say, RT policy.

Workaround: Remove RT or import map and add it back. Note, however, that if the above conditions occur again, the issue could reappear.

• CSCto13462

Symptoms: A Cisco router may reload when it is acting as a DHCP server.

Conditions: The symptom is observed when there are two DHCP clients using the same client ID. The reload is triggered when the second client releases its DHCP address.

Workaround: There is no workaround.

• CSCto71004

Symptoms: Router crashes with high scale and a lot of BGP routes and scaled MPLS L3 VPNs enabled. This crash is seen in the router when core links flap.

Conditions: The symptom is seen when a scaled router with a lot of BGP routes crashes when some of the core links flap. Setup has scaled MPLS L3 VPNs enabled.

The following messages are seen:

Workaround: There is no workaround.

• CSCto72927

Symptoms: Configuring an event manager policy may cause a Cisco router to stop responding.

Conditions: This issue is seen when a TCL policy is configured and copied to the device.

Workaround: There is no workaround.

• CSCtq49325

Symptoms: A router reloads when a graceful shutdown is done on EIGRP.

Conditions: The router reload occurs only when multiple EIGRP processes redistributing each other run on two redundant LANs and a graceful shutdown is done on both EIGRP processes simultaneously.

Workaround: Redundant LANs may not be necessary in first place. If it is required, if mutual redistribution is done, then while doing graceful shutdown, sufficient time should be given for one process to be shutdown completely before executing the second shutdown command. This should resolve the problem.

Further Problem Description: In a normal scenario, a zombie DRDB or path entry (a temporary DRDB entry which is deleted as soon as processing of the packet is done) would be created only for reply message. But here, due to the redundancy in LAN and EIGRP processes in this scenario, a query sent on one interface comes back on the other which causes this zombie entry creation for the query also. In the query function flow it is expected that this zombie entry will not be deleted immediately, rather it is to be deleted only after a reply for the query is sent successfully. At this point, (i.e.: before a reply is sent) if a shutdown is executed on the EIGRP process, then all the paths and prefixes will be deleted. If a particular path is threaded to be sent - in this case it is scheduled for a reply message - the path is not deleted and an error message is printed. However the flow continues and the prefix itself is deleted. This results in a dangling path without the existence of any prefix entry. Now when the neighbors are deleted, the flushing of the packets to be sent will lead to crash since it does not find the prefix corresponding to the path. The solution is to unthread from the paths from sending before deletion. A similar condition will occur if the packtization timer expiry is not kicked in immediately to send the DRDBs threaded to be sent and a topology shutdown flow comes to execute first.

• CSCtr04829

Symptoms: A device configured with “ip helper-address” drops packets because of a zero hardware address check.

Conditions: This symptom occurs when the hardware address is zero.

Workaround: There is no workaround.

• CSCtr34960

Symptoms: A router that is running Cisco IOS may run out of IO memory.

The show buffers command shows that the count reaches 0 in free list.

Conditions: This issue is seen post bootup. The Cisco 7600 in HA is required to hit the issue. The show buffers old command shows some buffers hanging on EOBC buffers list for a long time, weeks or more. The issue is a corner case, and buffer leak rate is slow.

This DDTS fixes leaks for the mls cef maximum-routes and mls cef adjacency-mcast commands.

See the output from the show buffers old pack:

The show buffers old packet command output will be either 000603E7 OR 00480002.

Workaround: Reload the supervisor to clear the leaked buffers.

• CSCtr47317

Symptoms: After a switchover, a Cisco Catalyst 6500 series switch may be replicating some spanned traffic indefinitely and flooding the network with the span copies.

Conditions: The issue is seen after the following sequence:

– An internal service module session for a FWSM or other service modules exists:

– If you attempt to configure a span session with the session number already in use:

– The command seems to be rejected, but it is synchronized to the standby supervisor.

– A switchover happens.

Workaround: There is no workaround.

• CSCtr51926

Symptoms: IPv6 packets are not classified properly in a subinterface when a service-policy is applied on the main interface.

Conditions: The symptom is observed when a service-policy is applied on the main interface.

Workaround 1: Enable IPv6 explicitly on the main interface:

Workaround 2: Reconfigure the IPv6 address on the subinterface:

• CSCtr58140

Symptoms: PFR-controlled EIGRP route goes into Stuck-In-Active state and resets the neighbor.

Conditions: This symptom is observed when the PFR inject route in an EIGRP topology table after the policy decision. The issue was first seen on an MC/BR router running PFR EIGRP route control and with EIGRP neighbors over GRE tunnels.

Workaround: There is no workaround.

• CSCtr79347

Symptoms: crashes at BGP Task without a BGP configuration change or BGP neighbor up/down event.

Conditions: No conditions but this is a rarely observed issue.

Workaround: There is no workaround.

• CSCtr88739

Symptom 1: Routes may not get imported from the VPNv4 table to the VRF. Label mismatch may also be seen.

Symptom 2: The routes in BGP may not get installed to RIB.

Conditions: The symptoms are only observed with routes with the same prefix, but a different mask length. For example, X.X.X.X/32, X.X.X.X/31, X.X.X.X/30..... X.X.X.X/24, etc. These issues are not easily seen and are found through code walkthrough.

For symptom 1, each update group is allocated an advertised-bit that is stored at BGP net. This issue is seen when the number of update groups increases and if BGP needs to reallocate advertised-bits. Also, this symptom is observed only with a corner case/timing issue.

For symptom 2, if among the same routes with a different prefix length, if more specific routes (15.0.0.0/32) do not have any bestpath (for example, due to NH not being reachable or inbound policy denying the path, but path exists due to soft-reconfiguration), then even if a less specific route (15.0.0.0/24) has a valid bestpath, it may not get installed.

Workaround for symptom 1: Remove “import-route target” and reconfigure route-target.

Workaround for symptom 2: Clear ip route x.x.x.x to resolve the issue.

• CSCtr91106

A vulnerability exists in the Cisco IOS Software that may allow a remote application or device to exceed its authorization level when authentication, authorization, and accounting (AAA) authorization is used. This vulnerability requires that the HTTP or HTTPS server is enabled on the Cisco IOS device.

Products that are not running Cisco IOS Software are not vulnerable.

Cisco has released free software updates that address these vulnerabilities.

The HTTP server may be disabled as a workaround for the vulnerability described in this advisory.

This advisory is available at the following link:

http://tools.cisco.com/security/center/content/CiscoSecurityAdvisory/cisco-sa-20120328-pai

• CSCtr92285

Symptoms: The following log is seen, and VCs cannot be configured.

Conditions: This symptom is observed with the access circuit interface shut and core flaps occurring, along with pseudowire redundancy. Also, leaks occur per flap.

Workaround: There is no workaround. If VCs can be removed, do so to release some IDs. Otherwise, try a redundancy switchover.

• CSCts06929

Symptoms: Disposition traffic gets dropped after SSO as the new local labels allocated by AToM do not get programmed on the line cards.

Conditions: This symptom occurs when pseudowires are configured on the setup without graceful restart configured. Then, SSO is performed and two local labels have the same disposition information. This really manifests as a traffic drop issue when the scale is high.

Workaround: Configuring graceful restart resolves this issue.

• CSCts13255

Symptoms: Standby SUP720 crash is observed on the Cisco 7600 router in c7600s72033-advipservicesk9-mz.150-1.S3a.bin. This issue is random and recurring. Tracebacks are generated with the following error message:

Conditions: This symptom is observed on the Cisco 7600 router with mistral based supervisors like SUP720. This issue is fairly uncommon, but affects all the versions after Cisco IOS Release 12.2(33)SRE, including Cisco IOS Releases 15.0S, 15.1S and 15.2S. This does not affect RSP 720.

Workaround: There is no workaround.

• CSCts32920

Symptoms: Traffic gets punted to the RP.

Conditions: This symptom occurs when there are multiple P2P-GRE tunnels in a particular VRF. Remove one particular P2P-GRE tunnel from that VRF.

Workaround: Shut/no shut P2P-GRE tunnels in that particular VRF, for which traffic is getting punted to the RP.

• CSCts38429

The Cisco IOS Software Internet Key Exchange (IKE) feature contains a denial of service (DoS) vulnerability.

Cisco has released free software updates that address this vulnerability. This advisory is available at the following link:

http://tools.cisco.com/security/center/content/CiscoSecurityAdvisory/cisco-sa-20120328-ike

• CSCts41217

Symptoms: Memory leak occurs in const_mfib_lc_* processes.

Conditions: This symptom is observed during deletion and addition of VRFs. Multicast needs to be enabled in the VRF and the P2P tunnels should be present in the VRF. VRF should be removed before deleting the P2P tunnels for this to happen.

Workaround: Delete all the P2P tunnels before removing the VRF

• CSCts47566

Symptoms: During standby RP bootup, an incorrect ICMP punt entry for MPLS interfaces gets programmed. This entry appears even if the interface has no ACL applied in the input/output direction.

Conditions: This symptom is observed with the ICMP punt entry, which is required to allow trace route across MPLS cloud. This is a workaround for a hardware problem with Tycho. The feature is called “FM_FEATURE_MPLS_ICMP_BRIDGE”. This workaround is required only if there are aggregate labels programmed in the superman VPN cam, but can get set incorrectly even when there is no VRF configuration on the box.

Workaround: To clear the entry set unnecessarily, disable/enable MPLS on the interface for which it appears.

• CSCts58394

Symptoms: The SNMP graph traffic rate (collected from the port-channel subinterface) does not match the 5-minute offered rate from “show policy-map inter port-channel x.x”.

Conditions: This symptom occurs on the Cisco 7600-S running Cisco IOS Release 15.0(1)S4 with the port-channel subinterface on 76-ES+XC-40G3CXL. This issue is seen only when there is EARL recirculation of packets and affects only the ingress traffic rate.

Workaround: There is no workaround.

• CSCts59014

Symptoms: Only one ATM VC shaper token is updated per cycle in a high-scale scenario.

Conditions: This symptom is observed with HQOS on ATM VC with many ATM VCs per interface.

Workaround: There is no workaround.

• CSCts88467

Symptoms: Drops happen earlier than expected.

Conditions: This symptom occurs if the queue-limit is incorrectly calculated.

Workaround: Configure a queue-limit explicitly to fix this issue, then remove and reapply the policy. Configuring queue-limit in parent policy automatically triggers calculation based on the parent queue-limit value on the child queue-limits based on bandwidth allocated to various classes.

• CSCtt02313

Symptoms: When a border router (BR) having a parent route in EIGRP is selected, “Exit Mismatch” is seen. After the RIB-MISMATCH code was integrated, RIB-MISMATCH should be seen, and the TC should be controlled by RIB-PBR, but they are not.

Conditions: This symptom is observed when two BRs have a parent route in BGP and one BR has a parent route in EIGRP. The preferable BR is the BR which has a parent route in EIGRP. The BRs having BGP have no EIGRP configured.

Workaround: There is no workaround.

• CSCtt03485

Symptoms: ES40: IDBMAN crash is seen with “no ip flow-export destination <> vrf <>”.

Conditions: This symptom occurs when “ip flow-export destination 10.21.1.1 3000 vrf vrf_1120” is removed.

mapping 1013 is cleared, but 1015 is not cleared from idbman mapping

mapping 1015 is still present in IDBMAN, eventhough 1015 is a free VLAN, so, it can be allocated to any new interface

Now, 1015 can be allocated for any other new interface, as it is cleared from “internal VLAN table”, whereas it is not cleared from IDBMAN mapping. Thus, you can reproduce the IDBMAN inconsistency with NDE interfaces.

When a new interface comes UP, the IDBMAN set will fail, as there is already an old mapping existing (NDE_1015). When you try to delete this new interface, it will try to clear the mapping in IDBMAN. But, it finds the old mapping (NDE_1015); hence, you must perform forced crash in idbman_if_clear_vlan_id and configure “ip flow-export destination 10.21.1.1 3000 vrf vrf_1120”.

Workaround: Reload.

• CSCtt06191

Symptoms: You cannot FTP copy file from the router to the FTP server.

Conditions: The symptom is observed when using Cisco IOS Release 12.2(33)SRE3 and an FTP server configuration with overwrite off.

Workaround: Change the FTP server configuration to allow overwrite.

• CSCtt19442

Symptoms: Cisco 7600 subinterface that is configured for bridging after router reload sends traffic even when being shutdown. This traffic is sent from physical interface to which subinterface correspond and further received on the other side of the link.

Conditions: This symptom is seen when bridging is configured on subinterface.

Workaround:

– Doing a no shutdown, then shutdown on the subinterface clears the issue.

– Remove bridging configuration from subinterface.

Deleting subinterface, and then recreating it does not fix the issue.

• CSCtt23367

Symptoms: The status on active PoA is A/U. The status on standby PoA is S/A.

Conditions: This symptom is seen after HA switchover. When configuring a new mLACP port-channel on new ACTIVE RP, it may get stuck in A/U state.

Workaround: Remove the port-channel and RG configuration and add back again.

• CSCtt25612

Symptoms: The router crashes with traceback error messages and the standby takes over. After this, the router is stable.

Conditions: There is no known trigger or changes that were made as per the user update.

Workaround: There is no workaround.

• CSCtt36757

Symptoms: The following error message is noticed when configuring QoS on the interface of an ES+ card:

Conditions: The symptom is observed after a misconfiguration in the interface. The interface was misconfigured as switchport which removed the QoS configuration from the interface configuration but not from the linecard. After the interface was configured back to an L3 port, the issue started occurring when the same policy was reapplied.

Workaround: A new policy can be applied but the required policy cannot be applied again.

• CSCtt43834

Symptoms: Netflow counter gets incremented when sending SSM group range as v2.

Conditions: The symptom is observed when doing an SSO.

Workaround: There is no workaround.

• CSCtt46638

Symptoms: A Cisco 7604 running Cisco IOS Release 12.2(33)SRE4/SRE5 crashes when changing the tunnel source and destination of an IPsec sVTI.

Conditions: The symptom is observed once the IPsec session is up and traffic is flowing through. If the tunnel source or destination is changed, the router crashes. This does not occur with a plain GRE tunnel.

Workaround: There is no workaround.

• CSCtt46873

Symptoms: In an MVPN setup, when the mdt default command is removed from under the VRF, unicast packets coming from the core, such as LDP and BGP, get dropped, leading to router isolation.

Conditions: This issue is primarily seen when mls mpls tunnel-recir is not configured on the box (or does not get enabled due to the absence of a sip10g device). In such a case, MDT tunnel VLAN gets allocated, but is never released, until the mdt default command is removed. Since the decap adjacency handling the unicast packets is a GRE decap, with an MDT tunnel VLAN allocated, removal/re-add of mdt default command will program the adjacency with the MDT tunnel VLAN. Another removal along with a race condition might leave the adjacency with the tunnel VLAN (now deallocated), thereby causing the unicast packets to be dropped.

Workaround: Configure mls mpls tunnel-recir on the box and remove/re-add the mdt default command or reload with mls mpls tunnel-recir configured to be safe.

• CSCtt90672

Symptoms: CFM MEP enters the INACTIVE state on deleting the subinterface.

Conditions: This symptom is observed under the following conditions:

1. Create a subinterface (vlan 104) for EOAM communication. Check “CC-Status” = Enabled.

2. Create a QinQ subinterface (vlan tags: 104 128) for subscriber on the same physical interface. Check “CC-Status” = Enabled.

3. Later, delete the QinQ subinterface from the step 2 above (DT’s provisioning system does it, for example, for a new policy change). The “CC-Status” goes to inactive.

Workaround: Unconfigure and reconfigure the continuity check command under the corresponding Ethernet CFM domain/service global configuration for this CFM MEP.

• CSCtu12574

Symptoms: The show buffers command output displays:

1. Increased missed counters on EOBC buffers.

2. Medium buffer leak.

The leak is small. It is a leak of 64 bytes per buffer that is leaked, and the leak appears to be very slow.

Conditions: The show buffers old command output displays some buffers hanging on the EOBC buffers list for a really long time, such as weeks or even more. This issue is a corner case and the buffer leak rate is slow.

This DDTS tracks the leak specific to IPC application l3-mgr.

From the show buffers old pack output:

And, if we look at the ICC header at the underscored items 00520002:

Workaround: Reload the system.

• CSCtu30649

Symptoms: Standby is reset.

Conditions: This issue is seen when the ISSU standby is reset because of MCL failure.

Workaround: There is no workaround.

• CSCtu35116

Symptoms: VPDN session keeps on trying to come up with MPLS MTU higher than 1500.

Conditions: The symptom is observed when you upgrade a Cisco 7200VXR from the c7200-a3jk91s-mz.122-31.SB18 to the c7200-adventerprisek9-mz.122-33.SRE4 image.

Workaround: There is no workaround.

• CSCtu36674

Symptoms: Packets stop being transmitted in the output direction on L2transport local connect PVC on the ATM interface.

Conditions: This symptom is observed when local connect is configured and a new ATM subinterface is configured on the same ATM main interface as the one with local connect PVC.

Workaround 1: Perform shut/no shut on local connect.

Workaround 2: Unconfigure/reconfigure local connect.

• CSCtu42550

Symptoms: ARP failure.

Conditions: Occurs when IPv6 address or IPv6 multicast is configured.

Workaround: There is no workaround.

• CSCtu60863

Symptoms: IGMP reports do not get installed in the IGMP group list.

Conditions: The symptom is observed when the port-security feature is enabled on the switchport which is part of the VLAN on which the IGMP reports are received.

Workaround: Remove “switchport port-security” from ports associated with the VLAN on which the IGMP reports are received.

• CSCtv19529

Symptoms: Router crashes on unconfiguring the last available DHCP pool. Crash will also be seen on running the no service dhcp.

Conditions: This crash can happen only if “DHCP Client” process is running on the router along with the DHCP relay processes (DHCPD Receive, DHCPD Timer, DHCPD Database).

The client process can be started:

1. from an DHCP autoinstall attempt during router startup (with no nvram config).

2. if the ip address dhcp is run on one of the interfaces.

3. if the router was used for DHCP proxy client operations.

The relay processes are started when a DHCP pool is created by the ip dhcp pool pool command.

Workaround: Have a dummy DHCP pool created using the ip dhcp pool dummy_pool command, and never delete this pool. Other pools can be created and removed at will, the dummy_pool should not be removed. In addition, do not execute the no service dhcp command.

• CSCtw64040

Symptoms: Crash due to MPLS, which appears to be associated with load- balancing.

Conditions: This symptom occurs when MPLS is configured.

Workaround: There is no workaround.

• CSCtw99290

Symptoms: The source or destination group-address gets replaced by another valid group-address.

Conditions: The symptom is observed during the NVGEN process if it suspends (for example: when having a huge configuration generating the running-config for local viewing or during the saving of the configuration or during the bulk sync with the standby and the NVGEN process suspends). The global shared buffer having the address gets overwritten by another process before the NVGEN completes.

Workaround: There is no workaround.

• CSCtx08147

Symptoms: Mismatch in next-hop interface is observed in the outputs of the CEF table in show ipv6 cef as compared to show mls cef ipv6. For example:

Conditions: The issue is triggered with link flaps and the traffic shifting completely to use an alternate path. Issue is observed with recursive routes enabled over BGP and ISIS. It is associated with ECMP paths. It occurs for IPv6 prefixes for which the CEF OCE chain (output chain: field in the output given by show ipv6 cef prefix/mask internal) starts with a load balance object.

Workaround: You can recover by these methods:

1. Clear IPv6 routes.

2. Add/remove static IPv6 routes for the bad prefixes.

3. Shut/no shut the interfaces.

• CSCtx39936

Symptoms: A Cisco 7600 router configured for MPLS TE with tunnel load-sharing may punt traffic to MSFC when multiple TE paths to a given destination exist.

Conditions: The symptom is observed with a Cisco 7600 router that is running Cisco IOS Release 12.2(33)SRE4, configured with multiple MPLS TE tunnels with load-sharing.

Workaround 1: In some cases, clearing the router may trigger proper reprogramming of the prefix in the hardware.

Workaround 2: Remove load-sharing from the TE tunnels.

• CSCty06990

Symptoms: Intercepted packets are not forwarded to MD.

Conditions: The symptom occurs randomly after applying an LI tap on a Cisco 7600 with a SIP400 as the dedicated LI service card.

Workaround: Remove and reapply TAP.

Resolved Caveats—Cisco IOS Release 12.2(33)SRE5

Cisco IOS Release 12.2(33)SRE5 is a rebuild release for Cisco IOS Release 12.2(33)SRE. The caveats in this section are resolved in Cisco IOS Release 12.2(33)SRE5 but may be open in previous Cisco IOS releases.

• CSCso88042

Symptoms: The VLANs allowed on the trunk to WiSM are lost on every reload.

Conditions: This symptom occurs when the number of entries in the allowed-VLAN statement exceeds five.

Workaround: Limit the number of entries to five or less, using ranges instead of single VLANs.

Further Problem Description: When the entries in the VLAN-allowed statement are more than five, two WiSM module allowed-VLAN statements are seen, even though one line was allowed during configuration. When reloaded, only one WiSM module allowed-statement is taken, and the first statement is lost.

• CSCsq45560

Symptoms: The port-channel member link stays as a standalone port with LACP.

Conditions: This symptom is observed only with the “vlan dot1q tag native” feature enabled.

Workaround: There is no workaround.

• CSCsy82121

Symptoms: Flooding of multicast traffic can be seen on source-only multicast networks.

Conditions: The symptom is observed on a Cisco Catalyst 6500 series switch that is running a SUP720 and with Cisco IOS SXF and SXH Releases.

When the issue occurs, the show mac-address multicast command does not show the group to be installed. It is expected that the mrouter ports installed would be seen so that traffic is directed upstream. Once an IGMP join is received from a receiver, the group will not be source-only and will be installed in the MAC table.

Workaround: There is no workaround.

• CSCtb24959

Symptoms: The router may crash while clearing a large number of RP mappings.

Conditions: This symptom occurs when you configure the router as an RP agent and candidate RP for a large number of RPs. This issue is seen when you run the clear ip pim rp-map command several times.

Workaround: Do not run the clear ip pim rp-map command several times in succession.

• CSCtd15853

Symptoms: When removing the VRF configuration on the remote PE, the local PE receives a withdraw message from the remote PE to purge its MDT entry. However, the local PE does not delete the MDT entry.

Conditions: This symptom occurs under the following conditions:

– mVPN is configured on the PE router.

– Both Pre-MDT SAFI and MDT-SAFI IOS are running in a multicast domain.

For more information about MDT SAFI, see the following document: http://www.cisco.com/en/US/prod/collateral/iosswrel/ps6537/ps6557/ps6604/ps6651/prod_white_paper0900aecd80581f3d.html

Workaround: There is no workaround.

• CSCtg57599

Symptoms: Lots of SNMP CPUHOG messages are seen and there is a crash due to a watchdog timeout:

Conditions: The symptom is observed when polling Dot3Stats.

Workaround 1: Use the no snmp-server sparse-tables command.

Workaround 2: Block the objects in dot3 mib that contains this table from being polled:

Then, to apply the view, use:

and then put it back so it looks like:

• CSCtg59328

Symptoms: When IPCP renegotiates for an existing PPPoE session, the new IPv4 address does not get synced up with the standby.

Conditions: This symptom is observed when the following tasks are completed:

– Bring up a PPPoE session and ensure that it is synced to standby.

– From the PPPoE client, run the no ip address command followed by the ip address negotiated command under the Virtual-template interface.

– As part of the no ip address command, the session would first go down on both active and standby. The ip address negotiated command would then trigger IPCP renegotiation and the session would come up on active. On standby, the session remains down and the new IP address is not synced.

Workaround: There is no workaround.

• CSCth08505

Symptoms: PPPoE sessions may not sync to the standby RP.

Conditions: This symptom is observed after the first attempt at establishing a PPPoE session fails.

Workaround: Reloading the standby RP may resolve this issue.

• CSCth33804

Symptoms: Traffic is dropped at CPP with the error message “noipv4route” after RP switchover, and traffic on few sessions is dropped.

Conditions: This symptom occurs when the VRF is configured for PPPoE sessions and RP switchover is done with traffic flowing.

Workaround: Do not configure the VRF.

• CSCth74953

Symptoms: The SPI value is shown as 0x0. Hence, the ipsec sa validation is failing.

Conditions: This symptom is observed when the crypto profiles are being applied. The symptom is not observed with simple crypto maps.

Workaround: There is no workaround.

• CSCth87458

Symptoms: Memory leak is detected in SSH process during internal testing. Authentication is required in order for a user to cause the memory leak.

Conditions: This symptom is observed during internal protocol robustness testing.

Workaround: Allow SSH connections only from trusted hosts.

PSIRT Evaluation: The Cisco PSIRT has assigned this bug the following CVSS version 2 score. The Base and Temporal CVSS scores as of the time of evaluation are 6.8/5.6: https://intellishield.cisco.com/security/alertmanager/cvssCalculator.do?dispatch=1&version=2&vector=AV:N/AC:L/Au:S/C:N/I:N/A:C/E:F/RL:OF/RC:C

CVE ID CVE-2011-2568 has been assigned to document this issue.

Additional information on Cisco’s security vulnerability policy can be found at the following URL: http://www.cisco.com/en/US/products/products_security_vulnerability_policy.html

• CSCti04919

Symptoms: While unconfiguring and reconfiguring the VRF, PIM neighborship goes down in a specific scenario.

Conditions: This symptom occurs if the PIM MDT GRE tunnel takes more time to come up compared to other interfaces in the VRF.

Workaround: Toggle the default MDT.

• CSCti23324

Symptoms: With some L2 DEC configurations, recirculation may be added during packet forwarding.

Conditions: This symptom is seen with L2 DEC and PFC3B configurations.

Workaround: This is not a forwarding issue. Remove L2 DEC or use PFC3C in the L2 DEC.

• CSCti42671

Symptoms: The state of MLP bundles on the active RP and the standby RP is not in sync. Some of the bundles that are active on the active RP show up as inactive on the standby RP. This may result in the bundles going to down state after switchover.

Conditions: This symptom occurs under the following conditions:

1. Configure scaled number of MLP bundles on 1xCHOC12 SPA.

2. Reload the SPA.

Workaround: Reload the standby RP.

• CSCti82670

Symptoms: An RSP will crash when the CFM automated test script (consisting of 53 tests) is run twice in succession. With Cisco SUP720, the crash is seen with a single run.

Conditions: The automated test script must be run on three connected routers.

Workaround: Adding a no shut on the UUT interface with UP-MEPS before doing the LeakConfig seems to prevent the crash and provide a clean run.

Further Problem Description: Other problems observed are:

– The CFM MIB will return infinite results for getmany.

– A show command will crash the router.

– Stale Earl Adjacency entries remain while adding or removing EVCs. The workaround is to reload the LC.

• CSCti87194

Symptoms: The last fragment causes a crash because of an invalid zone value.

Conditions: This symptom occurs when a Big IPC message is fragmented. Then, the last fragment causes the crash because of an invalid zone value.

Workaround: There is no workaround.

• CSCtj56551

Symptoms: The Cisco 7600 crashes in a very rare case.

Conditions: This symptom is observed very rarely when route-churn/sessions come up.

Workaround: There is no workaround.

• CSCtj94490

Symptoms: The Route Processor (RP) reloads after 30 RP switchovers.

Conditions: This symptom occurs after 30 RP switchovers during 28000 PPPoEoA sessions while traffic is flowing.

Workaround: There is no workaround.

• CSCtk18404

Symptoms: Per-user route is not installed after IPCP renegotiation.

Conditions: The symptom is observed with the following conditions:

1. A PPP session comes up, and NAS installs static routes that are sent as an attribute from the RADIUS server.

2. After a while, if CPE asks for IPCP renegotiation, IPCP is renegotiated, but the static routes are lost.

Workaround: There is no workaround.

• CSCtl67150

Symptoms: PPP multilink interfaces fail to come up on the serial interface in Cisco ASR1K series routers.

Conditions: This symptom occurs under the following conditions:

1) When you create one or more than one t1 channel groups in a CT3 interface.

2) When you create a multilink interface.

3) When you create one link per channel group.

4) When the encapsulation for every link is PPP, authentication CHAP.

5) When the multilink interfaces fail to come up.

Workaround: There is no workaround.

• CSCtl82681

Symptoms: IPv6 configuration is blocked in the subinterface when xconnect is configured in the main interface under a service instance.

Conditions: This symptom is seen if the xconnect is configured under a service instance that has its encapsulation configured as untagged, default or dot1q vlan range.

Workaround: Change the encapsulation of the service instance that has the xconnect.

• CSCtl90292

Symptoms: The following error messages are displayed:

Conditions: This symptom is observed when several hits and failures are seen for medium buffers. All are linktype IPC.

For example:

Workaround: There is no workaround.

• CSCtl97230

Symptoms: When configuring 42 MLPP bundles, each bundle has two members. In 10 members, interleave is enabled and the policy is attached. After SSO is done under traffic, the LC crashes continuously in new active.

Conditions: This symptom is observed when LC crash occurs with the Cisco 7600 software data plane environment. This issue is seen with SSO with the following conditions:

1. PPP multilink interleave.

2. Policy with priority feature.

Workaround: There is no workaround.

• CSCtn16855

Symptoms: The Cisco 7200 PA-A3 cannot ping across ATM PVC.

Conditions: This symptom occurs due to a high traffic rate, and the output policy applied under PVC.

Workaround: There is no workaround. Removing the policy will resolve this issue, but the QoS functionality will not be present in this case.

• CSCtn19444

Symptoms: mLACP memberlinks may be bundled on an isolated PoA with a core failure, resulting in both PoAs becoming active.

Conditions: This symptom occurs when running mLACP. The ICRM connection between the PoAs is lost. The PoAs are in a split brain situation and both PoAs attempt to become active. If the interface configured as “backbone interface” goes down on one of the PoAs, that PoA may keep the port-channel memberlinks bundled. The end result is that both PoAs are in mLACP active state, and both have their port-channel memberlinks bundled. After the fix, the PoA with the backbone interface failure will unbundle its port-channel memberlinks, leaving only one PoA as active.

Workaround: Configure shared control by configuring “lacp max-bundle” on the Dual Homed Device (DHD) if the device supports it. This would prevent the DHD from bundling the memberlinks to both PoAs at the same time.

• CSCtn31333

Symptoms: CPU utilization is high due to the process Net Background.

Conditions: This symptom is observed on a router used for LNS with an L2TP application after upgrading to Cisco IOS Release 12.4(24)T.

Workaround: There is no workaround.

• CSCtn58128

Symptoms: The BGP process in a Cisco ASR 1000 router that is being used as a route reflector may restart with a watchdog timeout message.

Conditions: This symptom may be triggered by route-flaps in a scaled scenario, where the route reflector may have 4000 route reflector clients and may be processing one million+ routes.

Workaround: Ensure that “no logging console” is configured.

• CSCtn62287

Symptoms: The standby router may crash while flapping the interface or while doing soft OIR of the SPA.

Conditions: This symptom is observed when interfaces are bundled as a multlink and traffic flows across the multilink.

Workaround: There is no workaround.

• CSCtn95122

Symptoms: The ECC double-bit error is seen and the line card crashes. The following error message is reported by the LC:

Conditions: This symptom occurs when several sessions are deleted at the same time. In addition, there are other unknown race conditions that can cause this.

Workaround: There is no workaround.

• CSCtn97451

Symptoms: The bgp peer router crashes after executing the clear bgp ipv4 unicast peer command on the router.

Conditions: This symptom occurs with the following conditions:

Router3 ---ebgp--- Router1 ---ibgp--- Router2

ROUTER1:

--------

ROUTER2:

--------

ROUTER3:

-------

Crash reproduce steps are as follows:

1. Traffic travel from ROUTER3 to ROUTER2.

2. “clear bgp ipv4 unicast 10.1.1.1” on ROUTER2.

Workaround: There is no workaround.

• CSCtn99858

Symptoms: Crashinfo is seen.

Conditions: This symptom is observed during an 8k session.

Workaround: There is no workaround.

• CSCto00796

Symptoms: In a rare and still unreproducible case, the RR (also PE) misses sending the RT extended community for one of the redistributed VPNv4 prefix to the PE (also and RR) that is part of a peer-group of PE (+RR).

Conditions: This symptom occurs when a new interface is provisioned inside a VRF and the configuration such that the connected routes are redistributed in the VRF. This redistributed route fails to tag itself with the RT when it reaches the peering PE (+RR)

Workaround: Soft clear the peer that missed getting the RT.

• CSCto07586

Symptoms: An IPv4 static BFD session does not get established on a system which does not have IPv6 enabled.

Conditions: This symptom occurs with the following conditions:

1. Create an IOS image that does not IPv6 enabled.

2. Enable BFD on an interface.

3. Configure an IPv4 static route with BFD routing through the above interface.

The IPv4 BFD session does not get established, so the static route does not get installed.

Workaround: Unconfigure BFD on the interface, and then reconfigure it. Then, the session will come up.

• CSCto11957

Symptoms: PPPoE is terminated on port-channel with ES+ session limit error occurring incorrectly.

Conditions: This symptom is observed when scaled PPPoE sessions are terminated on a port-channel with ES+ ports. Sessions negotiate, disconnect, and attempt to renegotiate the port-channel number other than port-channel 2.

Workaround: Change the port-channel number to port-channel 2. Configure sessions to terminate on standalone ports.

• CSCto31265

Symptoms: ABR does not translate Type7 when primary Type7 is deleted even if another Type7 LSA is available.

Conditions: This symptom occurs with OSPFv3. ABR receives multiple Type7 LSA for the same prefix from Multiple ASBR.

Workaround 1: Delete/readd the static route that generates Type7.

Workaround 2: Execute the clear ipv6 ospf force-spf command on ABR.

Workaround 3: Execute the clear ipv6 ospf redistribution command on ASBR.

• CSCto41165

Symptoms: The standby router reloads when you use the ip extcommunity-list 55 permit|deny command, and then the no ip extcommunity-list 55 permit|deny command.

Conditions: This symptom occurs when the standby router is configured.

Workaround: There is no workaround.

• CSCto46716

Symptoms: Routes over the MPLS TE tunnel are not present in the routing table.

Conditions: This symptom occurs when the MPLS TE tunnel is configured with forwarding adjacency. In “debug ip ospf spf”, when the SPF process link for the TE tunnel is in its own RTR LSA, the “Add path fails: no output interface” message is displayed. Note that not all tunnels are affected. It is unpredictable which tunnel is affected, but the number of affected tunnels grows with the number of configured tunnels.

Workaround: If feasible, use autoroute announce instead of forwarding adjacency. Otherwise, upgrade to the fixed version.

• CSCto52194

Symptoms: With the P2MPTE scenario, some of the met3 entries are not programmed properly.

Conditions: This symptom occurs with shut/no shut of the interface.

Workaround: There is no workaround.

• CSCto55567

Symptoms: The ES+ card goes to a major error state because of fabric CRC errors.

Conditions: This symptom occurs after SSO with multicast traffic flowing through the line card.

Workaround: Soft reload the line card.

• CSCto63720

Symptoms: No traffic passes after a link flap if port-security is configured on the Gigabit Ethernet interface on 6748 LC.

Conditions: This symptom occurs when the Cisco IOS version running is Cisco IOS Release 12.2(33)SRE2. This issue is seen when port-security is configured on a 6748 port and the link flap occurs on this interface.

Workaround: Reconfiguring port-security fixes the problem.

• CSCto70633

Symptoms: Packets get punted to the RP because the default ACL does not get programmed on the Distributed Feature line card (DFC), which causes high RP CPU.

Conditions: This symptom is observed upon removal and reinsertion of the line card when there are VRF-scale configurations on the ES+ card as given below:

Workaround: Reload the router.

• CSCto71075

Symptoms: High CPU usage is seen on changing the root node multiple times in an MLDP setup. Loss of PIM neighborship is also seen when changing the path in a P2MP setup.

Conditions: This symptom occurs when ptcam redirection being enabled for LSPVIF can cause unexpected results. By default, LSPVIF ptcam redirection is disabled. This fix ensures that this is taken care of in scenarios of PIM state change.

Workaround: There is no workaround.

• CSCto72629

Symptoms: A MAXAGE LSA is repeatedly retransmitted bringing down the OSPFv3 adjacency.

Conditions: This symptom occurs when the unadjusted age of the LSA in the OSPFv3 database (as opposed to the advertised age, which includes time spent in the database) is less than MAXAGE. Note that the age of the LSA in the database is not updated once it is installed unless maxaging is initiated by OSPFv3 process.

Workaround: Use the clear ipv6 ospf process command to clear the OSPF state based on the OSPF routing process ID.

• CSCto79174

Symptoms: A Cisco 7600 router crashes with the following logs:

Conditions: The conditions are not known.

Workaround: There is no workaround.

• CSCto88660

Symptoms: Command failure on the RP is causing both protecting and working APS to go to active.

Conditions: This symptom may be caused by switchover during scaled conditions.

Workaround: There is no workaround.

• CSCto99523

Symptoms: Convergence can take more time if there are a lot of VRF/routes and aggregation is configured in many VRFs. Massive route churn happens (for example, session reset with RR).

Conditions: This symptom occurs when convergence takes more time if there are a lot of VRF/routes and aggregation is configured in many VRFs. This issue is seen when massive route churn happens (for example, session reset with RR). There is no functionality impact.

Workaround: There is no workaround.

• CSCtq04117

Symptoms: DUT and RTRA have IBGP-VPNv4 connection that is established via loopback. OSPF provides reachability to the BGP next hop, and BFD is running.

Conditions: This symptom occurs under the following conditions:

1. DUT has learned VPNv4 route from RTRA, and the same RD import is done at DUT.

2. When switchover is performed in RTRA and when GR processing is done, the route is never imported to VRF.

Workaround: Use the clear ip route vrf x * command.

• CSCtq06538

Symptoms: The RP crashes due to bad chunk in MallocLite.

Conditions: This symptom occurs while executing testcase number 4883. The test case 4883 sends an incorrect BGP update to the router to test whether the router is able to handle the problematic update. The incorrect BGP update has the local preference attribute length incorrect:

Workaround: There is no workaround.

• CSCtq09638

Symptoms: The router crashes upon reconfiguration of the existing frame-relay subinterface with applied QoS.

Conditions: This symptom is observed during reconfiguration of the existing frame-relay subinterface with applied QoS.

Workaround: There is no workaround.

• CSCtq21435

Symptoms: Some specific s,g entries do not pass traffic with MLDP during root node redundancy switchover.

Conditions: This symptom occurs in case of MLDP + RNR. This issue is seen when Accept Vlan is programmed as zero in the platform.

Workaround: Clear the mroute.

• CSCtq29547

Symptoms: The router crashes on watchdog timeout while processing the SNMP request for ciscoEigrpMIB.

Conditions: This symptom occurs while processing the SNMP request for ciscoEigrpMIB.

Workaround: Exclude ciscoEigrpMIB from being polled by using the following SNMP view:

Then, apply the view to your SNMP community string:

• CSCtq29554

Symptoms: All multicast routes may be missing from the multicast forwarding information base (MFIB) after SSO and MFIB/MRIB error messages may be generated, indicating failure to connect MFIB tables to the MRIB. The output of the show ipc port | in MRIB command on a failed line card does not display a port.

Conditions: This symptom can occur on a line card of a distributed router such as the Cisco 7600 if an IPC local error has occurred before switchover. The MRIB IPC port to the new RP is not created after switchover and the MFIB tables cannot connect to the MRIB and download multicast routes.

Workaround: Reload the failing line card to recover it.

• CSCtq34807

Symptoms: Service group does not take effect on EVC Xconnect on a port channel.

Conditions: This symptom is observed with a service group configuration on EVC Xconnect existing on a port channel. This issue is seen when EVC is removed and the configuration is reapplied.

Workaround: Remove and reapply the service group.

• CSCtq36241

Symptoms: ISG session setup fails when per-user IPv4 ACLs are used and IPv6 routing is configured.

Conditions: This symptom is observed when both IPv6 routing and per-user IPv4 ACLs are configured.

Workaround: Remove either IPv6 routing or per-user ACLs.

• CSCtq37538

Symptoms: Duplicate traffic is seen during route changes with P2MP TE for multicast or MLDP.

Conditions: This symptom occurs during LSM configuration and route changes.

Workaround: Clear the problematic mroute using the clear ip mroute command.

• CSCtq58383

Symptoms: A crash occurs when modifying or unconfiguring a loopback interface.

Conditions: This symptom occurs while attempting to delete the loopback interface, after unconfiguring the “address-family ipv4 mdt” section in BGP.

Workaround: Unconfiguring BGP may prevent the issue from happening without reloading the router.

• CSCtq62759

Symptoms: The CLNS routing table is not updated when the LAN interface with CLNS router ISIS configured shuts down because ISIS LSP is not regenerated. The CLNS route will be cleared after 10 minutes when ISIS ages out the stale routes.

Conditions: This symptom is seen when only CLNS router ISIS is enabled on the LAN interface. If IPv4/IPv6 ISIS is enabled, ISIS LSP will be updated.

Workaround: Use the clear clns route command or the clear isis * command.

• CSCtq64072

Symptoms: A DHCP release received on a different member link of a PC other than the one on which it was requested is considered as fake and dropped.

Conditions: This symptom occurs with the DHCP client release/decline message. The binding interface must match before the binding entry is removed to prevent someone from faking these messages to delete others’ valid binding.

Workaround: There is no workaround.

Further Problem Description: In case of a port-ch, the stored hwidb for a binding is that of the bridge interface. When a release is received on the other member-links, the hwidb does not match.

• CSCtq66679

Symptoms: After “power cycle/red reload shelf”, the active PW stays in DOWN state forever in the case of a BFD-enabled protected PW. The standby PW is in UP state and traffic flow is through standby.

Conditions: This symptom occurs in the PW redundancy scenario.

Workaround: There is no workaround.

• CSCtq67680

Symptoms: An SPA reload triggers silent LC reload under the following steps:

1. 1. Configure policy-maps as shown below: policy-map mul1 class GOLD priority 1000 class SILVER bandwidth 1000 policy-map mul2 class GOLD priority 7000 class SILVER bandwidth 7000 class class-default random-detect

2. Apply it on multilink interfaces multilink1 and multilink2.

3. Reload the SPA.

Conditions: This symptom is observed only with QoS policy applied on multilink bundle on serial SPA.

Workaround: There is no workaround.

• CSCtq80603

Symptoms: Newly created SVIs are in down/down state.

Conditions: This symptom occurs when SW VLAN RP process is stuck.

Workaround: The following workarounds may work:

1. Set the memory location of l2vlanifmib_access_count to zero after warm restart of the snmp-server.

2. Perform SSO and/or LC OIR.

3. Perform an active reload.

• CSCtq80648

Symptoms: If a user changes the VRF assignment, such as moving to another VRF, removing the VRF assignment, etc., on which a BGP IPv6 link-local peering (neighbor) is based, the BGP IPv6 link-local peering will no longer be able to delete or modify.

For example:

If the user changes the VRF assignment of Ethernet1/0 from vpn1 to vpn2, the IPv6 link-local neighbor, FE80::A8BB:CCFF:FE03:2200%Ethernet1/0, under address-family ipv6 vrf vpn1, will no longer be able to delete or modify.

Rebooting the router will reject this configuration. Also, if a redundant RP system and the release support the config-sync matching feature, it will cause config-sync mismatch and standby continuous reload.

Conditions: This symptom occurs when a user changes the VRF assignment.

Workaround: Remove the BGP IPv6 link-local peering before changing the VRF assignment on the interface.

• CSCtq82715

Symptoms: When the VPLS VC goes up/down, the DHCP snooping LTL has not been updated, resulting in DHCP packet drop.

Conditions: This symptom occurs when the VPLS VC goes up/down, indicating that the DHCP snooping LTL has not been updated.

Workaround 1: Enable/disable snooping.

Workaround 2: Clear the xconnect peer for the newly elected peer.

Further Problem Description: In such an event, the GPI is now passed onto DHCP snooping code to program its LTL.

• CSCtq83629

Symptoms: The error message is associated with a loss in multicast forwarding state on line cards under scaled conditions when an IPC error has occurred.

Conditions: This symptom is observed during router boot or high CPU loading, which can cause IPC timeout errors. This issue is seen on line cards during recovery from an IPC error in the MRIB channel.

Workaround: Line card reload is required to resolve the problem.

• CSCtq86216

Symptoms: Multicast traffic flows over both primary and backup interfaces during TEFRR reopt.

Conditions: This symptom occurs when multicast traffic flows over an MLDP core with TEFRR link protection.

Workaround: Duplicate traffic flows only for a short period of time (20 seconds). So, the issue gets automatically resolved after 20 seconds.

• CSCtq91403

Symptoms: High CPU can be seen during reloads under the MVPN topology.

Conditions: This symptom occurs in an MVPN network with an S,G with an incoming interface over the MDT tunnel, when there are no forwarding interfaces for that S,G.

Workaround: A possible workaround is to create a static join for that S,G to protect the RP CPU. Also, in some case multicast rate-limiters will be useful.

• CSCtq92182

Symptoms: An eBGP session is not established.

Conditions: This issue is observed when IPv6 mapped IPv4 addresses are used, such as ::10.10.10.1.

Workaround: Use an IPv6 neighbor address with bits. Set some higher bits, along with the IPv4 mapped address.

• CSCtq93823

Symptoms: Ping drops with fragment size of 256.

Conditions: This symptom occurs when doing a sweep ping with sizes 500 to 1000.

Workaround: Flap the interfaces.

• CSCtq94418

Symptoms: Adding, deleting, and readding an access subinterface may sometimes lead to loss of data path.

Conditions: The symptom is observed when the configuration sequence involves an add-delete-add sequence.

Workaround: Create dummy access subinterfaces belonging to a new VRF. Do not remove the interface.

• CSCtq96329

Symptoms: The router fails to send withdraws for prefixes, when bgp deterministic-med is configured. This could lead to traffic blackholing and routing loops. Also, it could result in memory corruption/crash in rare conditions.

Conditions: This symptom can happen only when bgp deterministic-med is configured.

The following releases are impacted:

– Cisco IOS Release 15.0(1)S4

– Cisco IOS Release 15.1(2)T4

– Cisco IOS Release 15.1(3)S

– Cisco IOS Release 15.2(1)T

Workaround: Disable deterministic med in the network/AS by issuing the no bgp deterministic-med command and then the clear ip bgp * command or hardreset the BGP session to remove any stale prefixes.

It is further recommended to do an SSO on routers that are running impacted software to eliminate any potential corruption that might have already existed on routers that are running impacted software.

Further Problem Description: If deterministic med is enabled, withdraws are not sent.

• CSCtr14852

Symptoms: A Cisco 7600 router may experience the following error conditions:

1. The router starts displaying ICC WATERMARK messages. (This is expected if it happens for a short duration and is not associated with the second symptom mentioned below). For example:

2. The above symptom would trigger a situation where the flow control mechanism is turned “ON” by the communication infra (ICC). As a result, the communication infra will fail to carry application data from one point to another within the router. This in turn would lead to failure of multiple features that are dependent on the ICC.

For example: The ICC flow control can be verified by the following command:

As shown above, the flow control is turned ON on L3-MGR-INTF, but never turned OFF.

The ICC flow control mechanism is required to manage the ICC. If the flow control is turned on for a genuine reason, it will be turned OFF in a short while. This is expected.

However, in this case, because of a bug in accounting, the flow control is turned ON (when not required), and never gets turned OFF, leading to the above situation.

Conditions: This symptom occurs during “ICC MULTICAST” (not IP multicast) usage. This issue may be caused by heavy route flaps or interface flaps.

Workaround: There is no workaround.

• CSCtr19286

Symptoms: A “no shut” on an administratively down interface may result in overruns on other interfaces that are forwarding traffic. This occurs on ports being no shut for the first time in the same ASIC group. Subsequent shut/no shut on the same port does not cause this issue.

Conditions: This symptom occurs under the following conditions:

– This issue has been seen on Rohini ASIC-based DFC LAN cards such as WS-X6748-GE-TX.

– The ports belong to the same port ASIC.

– This issue is seen only the first time you no shut an interface

Workaround: No shut all the ports in the ASIC group after bootup. Subsequent shut/no shut will not cause the overrun issue.

• CSCtr22007

Symptoms: A Cisco 7600 router that is configured with RSVP crashes.

Conditions: This symptom is observed with MPLS-TE Tunnel Flap.

Workaround: There is no workaround.

• CSCtr28527

Symptoms: After a few minutes of HA cutover, DHCP snooping on a VLAN stops.

Conditions: This symptom occurs after a few minutes of HA cutover.

Workaround: Shut/no shut the port-channel interface.

Further Problem Description: After SSO, the LTL consistency checker starts recomputing fpoe for each LTL. For those from the sw-mcast region, the LTL cc makes a callback to retrieve the gpid list to program the fpoe for the LTL. In this case, the DHCP snooping feature provides an incomplete list because the VPLS VC programming is done directly by the cwan_atom code and the feature is unaware of this gpid list. The VPLS VC gpid programming to LTL is now redirected to the feature itself.

• CSCtr28857

A vulnerability in the Multicast Source Discovery Protocol (MSDP) implementation of Cisco IOS Software and Cisco IOS XE Software could allow a remote, unauthenticated attacker to cause a reload of an affected device. Repeated attempts to exploit this vulnerability could result in a sustained denial of service (DoS) condition.

Cisco has released free software updates that address this vulnerability. Workarounds that mitigate this vulnerability are available. This advisory is available at the following link:

http://tools.cisco.com/security/center/content/CiscoSecurityAdvisory/cisco-sa-20120328-msdp

• CSCtr34793

Symptoms: The router cannot establish mVPN PIM adjacencies over an MDT tunnel. The core PIM still works normally.

Conditions: This symptom may occur after router reload when mVPN with PIM is configured and PIM-hellos from the neighbors are coming to the line card with DFC. Another possible trigger could be removal/recreation of the MDT in a VRF definition.

Workaround: Reload the line card.

• CSCtr37073

Symptoms: WS-X6196-RJ-21 and WS-X6148X2-RJ-45 may fail to come online on the Cisco 7600 router when running SRC or higher images.

Conditions: This symptom occurs when SRC or higher images are run on a Cisco 7600 router.

Workaround: There is no workaround.

Further Problem Description: This issue occurs due to a timing problem in the module initialization routine of the Cisco IOS.

• CSCtr37182

Symptoms: XAUI coding errors are seen on the console.

Conditions: This symptom is observed on a Cisco 7600 series router.

Workaround: There is no workaround.

• CSCtr53677

Symptoms: ARP failure is seen with the following show command:

show arp vrf vrf name

Conditions: This symptom occurs during deletion and readdition of the VRF with the multicast MDT configured and P2P tunnels as the access-facing interface, along with Gigabit subinterfaces.

This issue is seen under the following conditions:

1. Configure the mcast VRF with P2P tunnels as access-facing, along with Gigabit subinterfaces. Use a sip400 line card as the access-facing line card.

2. Delete the VRF, P2P tunnel, and Gigabit subinterface.

3. Add the same VRF again after a 60-second interval.

4. Observe the ARP failure on the Gigabit subinterface.

Workaround: There is no workaround.

• CSCtr53739

Symptoms: The tunnel-encap entry is wrongly programmed. The following show command is used:

show platform software multicast ip cmfib vrf vrf-name tunnel-encap verbose

Conditions: This symptom occurs during deletion and readdition of the VRF with the multicast MDT configured and P2P tunnels as the access-facing interface, along with Gigabit subinterfaces.

This issue is seen under the following conditions:

1. Configure the mcast VRF with P2P tunnels as access-facing, along with Gigabit subinterfaces. Use a sip400 line card as the access-facing line card.

2. Delete the VRF, P2P tunnel, and Gigabit subinterface.

3. Add the same VRF again after a 60-second interval.

4. Observe the tunnel-encap entry wrong programmed on the SP, with corrupt values.

Workaround: There is no workaround.

• CSCtr69937

Symptoms: The POS link flap in the core breaks the IPv4 PIC Core functionality.

Conditions: This symptom occurs on Cisco 7600 routers running Cisco IOS Release 15.1(03)S.

Workaround: Execute the clear ip route command for the affected prefix.

• CSCtr74529

Symptoms: The following error messages are displayed:

Conditions: This symptom is not caused by any specific conditions.

Workaround: There is no workaround.

• CSCtr80366

Symptoms: Relay miscalculates the giaddr from the OFFER packet, and hence cannot find the binding.

Conditions: This symptom occurs while configuring multiple pools on the server and multiple secondary IP addresses on the relay loopback IP address.

Workaround: There is no workaround.

• CSCts15072

Symptoms: Multicast traffic in the MVPN solution is dropped.

Conditions: This symptom is observed on the Cisco 7600 series routers after deletion and (re)creation of a VRF.

Workaround: Do not delete VRFs. All configuration related to a VRF can safely be removed. Only the VRF name should be retained in the configuration.

• CSCto55812

Symptoms: The router may crash.

Conditions: This symptom occurs on entering VLAN mode from a different mode, for example, VFI, without exiting from the previous command mode.

Workaround: Always exit from the current command mode while entering into another command mode.

• CSCtn52529

Symptoms: After the TE tunnels are recovered and resignaled, a timer is started to wait for the RESV to come in. If the RESV does not arrive, the timer is restarted and the PATH is sent out again. In this DDTS, the PATH was not being sent out after the timer expired.

Conditions: This symptom occurs after the TE tunnels are recovered and resignaled and the RESV does not arrive after the timer is started.

Workaround: There is no workaround.

• CSCts16285

Symptoms: The system may experience delays in updating multicast information on the line cards. MFIB/MRIB error messages may be observed when IPC messages from the line card to the RP time out. In the worst case, the line card may become disconnected if timeouts continue for a long period.

Conditions: This symptom occurs when the system has a very heavy IPC load or CPU load.

Workaround: Take necessary actions, if possible, to reduce the IPC load. Sometimes, the IPC load could be due to noncritical processes.

• CSCts51980

Symptoms: STM1-SMI PAs of version 3.0 do not come up.

Conditions: This symptom is observed when the new version of PAs do not come up with enhanced flexwan.

Workaround: There is no workaround. Without the PA, flexwan will come up.

Resolved Caveats—Cisco IOS Release 12.2(33)SRE4

Cisco IOS Release 12.2(33)SRE4 is a rebuild release for Cisco IOS Release 12.2(33)SRE. The caveats in this section are resolved in Cisco IOS Release 12.2(33)SRE4 but may be open in previous Cisco IOS releases.

• CSCsc84683

Symptoms: A crash is observed, along with a HASH_TABLE collision message, when a MAC address table entry for non-IP multicast MAC is removed.

Conditions: This symptom occurs only with non-IP multicast MAC addresses. The crash is seen when the ports belonging to that entry are removed one by one from the MAC address table entry.

Workaround: There is no workaround.

• CSCsl63149

Symptoms: If you repeatedly use the test mcast ltl command on a switch processor (SP), a buffer leak is introduced. The SP may finally run out of buffers and the system crashes.

Conditions: This symptom occurs when the test mcast ltl command is used repeatedly on an SP.

Workaround: There is no workaround.

• CSCsq02771

Symptoms: DHCP relay may hang when a request for an IP address is received from a DHCP client on an unnumbered MPLS and VPN setup.

Conditions: The symptom is observed on a Cisco 7200 router that is running Cisco IOS Interim Release 12.4(19.16)T1.

Workaround: There is no workaround.

• CSCsw32280

Symptoms: Cisco Catalyst 6500 WS-X6148x-RJ-45 line cards report the following error:

Conditions: This symptom is observed when the interface (Po=Port number) reported in the error message receives traffic with the wrong CRC and is reported under “show interface”.

Workaround: The only workaround is to turn off the test as follows:

The fix is available in Cisco IOS Release 12.2(33)SXI1 or later releases.

Further Problem Description: The following are sample logs:

• CSCsz39222

Symptoms: A Cisco router reloads and the crashinfo file indicates a cache error. CPO_ECC has the following value:

This was a hardware corrected cache error that should not result in a router reload.

Conditions: This symptom is observed when register 26/0 contains 0xC0000000.

This issue affects the RAN SVC card, NPE-G1 on a Cisco 7200 platform, NSE-150 on a Cisco 7300 platform, Sup32 for Cisco 6500/7600 platforms, SIP line cards for the Cisco 6500/7600, Cisco 67XX lan line cards for the Cisco 6500/7600 platforms, Cisco AS5400XM, Cisco UBR10K/PRE4, and other platforms using the same memory controller chip. Sup720 is not affected. NPE-G2 is not affected. NSE-100 is not affected. While rare, there is no specific trigger for this failure other than having a single bit parity error on ECC memory.

Workaround: There is no workaround. The router will reload and continue normal operation. The fix prevents a crash after a single bit parity error occurs on ECC memory.

Further Problem Description: This symptom does not cause a parity error or actually cause the crash. This symptom is just to add an error handler for the specific case of a single bit correctable parity error in ECC memory. The crash results from the parity error itself. The following is an example of the beginning of a crashinfo collection for a hardware corrected cache error:

• CSCta62394

Symptoms: A Cisco Catalyst 6000 router that is running Cisco IOS Release 12.2(18)SXF16 with ipsec HA and dynamic crypto maps reloads when the standby tears down all the IPSEC SAs.

Conditions: This symptom occurs with the following conditions:

1. The Cisco Catalyst 6000 router is running Cisco IOS Release 12.2(18)SXF16.

2. IPSEC HA is enabled.

3. Dynamic crypto maps are configured.

4. SAs are torn down on the standby (for example, when using the clear crypto session command or deleting a VLAN with crypto on it).

5. A dynamic crypto map is configured on the router. 6) The user attempts to clear the dynamic crypto map’s SAs using the clear crypto sa spi command.

Workaround: Under conditions 1 through 4 that apply to the Cisco IOS Release 12.2SXF throttle, use static crypto maps instead of dynamic crypto maps. Under conditions 5 and 6 that apply to the Cisco IOS Release 12.2SR throttle, do not use the clear crypto sa spi command or do not use a dynamic crypto map.

• CSCtc49086

Symptoms: When configuration changes are performed within a multicast-enabled VRF that cause the PIM register tunnel interface to go down and up again, spurious memory access appears when traffic is sent at the same time.

Conditions: This symptom occurs when traffic is sent when configuration changes are performed.

Workaround: There is no workaround.

• CSCtd14703

Symptoms: High CPU at Net Background is seen for around 3 minutes after SSO. Unicast protocol such as ospf may be flapping due to high CPU.

Conditions: This symptom is observed when scale P2P GRE tunnels are configured on Cisco Catalyst 6500 LAN switching.

Workaround: There is no workaround.

• CSCtf07365

Symptoms: The vrf oif tunnel interface is missing for several seconds after SSO on the source side PE.

Conditions: This symptom occurs when you configure the Cisco Catalyst 6500 as MVPN PE. The mfi/mrib and HW entry works fine for the fist 3 minutes after SSO and after BGP convergence is completed. The vrf oif tunnel interface is missing for 20 seconds in the mrib/mfib table when the system does reconstruction, causing vrf traffic to be dropped on the source side PE for several seconds.

Workaround: There is no workaround.

• CSCtf08317

Symptoms: Unable to connect to line card after switchover.

Conditions: This symptom occurs while connecting to line card after switchover.

Workaround: There is no workaround.

• CSCtf50944

Symptoms: Active or standby RP crash is seen when configuring dot1q/QinQ non-access subinterfaces.

Conditions: This symptom is observed with the following conditions:

1. Configure around 10 non-access subinterfaces, bring session and delete the subinterfaces.

2. Configure same subinterfaces (subinterface #), with a different encapsulation.

Crash is seen with the following message:

% idbman_if_set_vlan_id: GigabitEthernet12/8.4 has vlan 1022, vlan 1023 not set

The crash occurs very rarely and is triggered by timing idiosyncrasies. The issue is only seen when the configuration is done using scripts.

Workaround: There is no workaround.

• CSCtf91692

Symptoms: When a WS-X6708-10GE module or a WS-X6716-10GE module is inserted into a 6509 or 6513 chassis, it may cause the module in slot N-8 to reload. For example, inserting the module into slot 9 may reset the module in slot 1, inserting the module into slot 10 may reset the module in slot 2, inserting the module into slot 13 may reset the module in slot 5, and so on.

Conditions: This symptom occurs with the following conditions:

1. The 6708/6716 module is inserted in slots 9 to 13.

2. Module insertion is done slowly.

Workaround: This problem has been fixed.

• CSCtg91572

Symptoms: A router with an SSM (S,G) entry consisting of a NULL outgoing list sends a periodic PIM Join message to the upstream RPF neighbor, thereby pulling unnecessary multicast traffic.

Conditions: This symptom is observed when the router has a NULL outgoing list for an SSM (S,G) entry either due to PIM protocol action (Assert) or when the router is not the DR on the downstream access interface receiving IGMPv3 reports.

Workaround: There is no workaround.

• CSCth15353

Symptoms: Incorrect result codes are displayed in vpdn sys logging. The CDN message for admin down was reported in the syslog as “Result Code=2, Error Code=6” instead of “Result Code=3, Error Code=6”.

Conditions: This symptom is observed when a session is cleared by a clear command (for example, clear interface virtual-access 3.1).

Workaround: There is no workaround.

• CSCth37674

Symptoms: A crash happens when the show ip dhcp vrf vrf-name command is executed.

Conditions: This symptom occurs when the show ip dhcp vrf vrf-name command is executed.

Workaround: There is no workaround.

• CSCth52444

Symptoms: When the strict priority is configured in the policy, the policing action allows to configure exceed action as transmit.

Conditions: This symptom is observed with Cisco IOS Release 15.1(0.9)S.

Workaround: There is no workaround.

• CSCth66177

Symptoms: The standby route processor (RP) triggers an active RP crash.

Conditions: This problem is observed when the standby RP crashes due to a memory parity error.

Workaround: There is no workaround.

• CSCth84714

Symptoms: With scaled number of MLP bundles on SIP200 which has Distributed Link Fragmentation and Interleaving (dLFI) enabled, the SIP200 crashes.

Conditions: This symptom occurs with the following conditions:

1. Reload the SPA having MLP bundles.

2. Shut/no shut the controller.

3. Flap the links by any other means.

Workaround: The issue is not seen without high traffic and without dLFI enabled.

• CSCti18615

Symptoms: Reloading a router which has multicast forwarding configured can result in the standby RP being out of sync with the active RP. The A and F flags are missing from the multicast forwarding base entries.

Conditions: This symptom occurs when multicast forwarding is operational and configured in the startup configuration, and when the router is in HA mode SSO and is reloaded from the RP.

Workaround: Perform a shut/no shut of the affected interfaces.

• CSCti81177

Symptoms: Features like Videomon do not work on a routed port.

Conditions: This symptom occurs when an interface is configured as a switch port and reconfigured as a routed port.

Workaround: Reload the line card.

• CSCti92812

Symptoms: After physical interface flap, GRE tunnel for VRF does not come up correctly.

Conditions: This symptom occurs when GRE tunnel is configured for default (global) routing table.

Workaround: There is no workaround.

• CSCtj11497

Symptoms: The Shared Port Adapter (SPA) crashes after receiving the “%INTR_MGR-3- INTR: PL3 RX Sequence” error.

Conditions: This symptom occurs under normal working conditions.

Workaround: The SPA reloads automatically and clears the problem.

• CSCtj20776

Symptoms: Accounting-stop record is sent for radius proxy session when reauthentication happens for that session.

Conditions: This symptom is seen in the following scenarios:

1. The authentication request comes from AP.

2. The accounting request comes from AZR and the session on ISG is associated to AZR.

3. ISG receives a reauthentication request from AP. The Accounting-stop record is sent for Radius-Proxy session and the services under the session, but the radius-proxy session is still active and no stop record is sent for the session on clearing the session. Also, acct-terminate-cause in the stop record is set to none.

Workaround: There is no workaround.

• CSCtj61748

Symptom: Service activation fails occassionally.

Conditions: This symptom occurs with multiple services in the session authentication or authorization response that are configured in the same service-group.

Workaround: Remove fields that are related to “service-group” or “service- type” in service definitions.

• CSCtj65692

Symptoms: The service policy applied to a service instance stops forwarding any traffic. The output of the show policy-map interface x/y command indicates that all packets are hitting the violation queue. The conform counter does not increase at all and all traffic is dropped.

Conditions: This symptom is observed in Cisco 7600 running Cisco IOS Release 12.2SRD. This issue is applicable for the service policy applied for ingress or egress traffic.

Workaround: There is no workaround. To restore the services, the service policy has to be removed from the service instance, and then the condition clears. The service policy can then be reapplied and will work normally.

• CSCtj79676

Symptoms: The router crashes sometimes once CEF is enabled.

Conditions: This symptom occurs when CEF is enabled.

Workaround: There is no workaround.

• CSCtj87846

Symptoms: Performance Routing (PfR) traffic class fails to transition out of the default state.

Conditions: When a subinterface is used as an external interface and the corresponding physical interface goes down and comes up, the PfR master is not notified that the subinterface is back up.

Workaround: Do shut/no shut on PfR master or PfR border.

• CSCtk07240

Symptoms: When a member-link is removed from an L2 port-channel (a port-channel with switchport configured under it), the traffic stops flowing.

Conditions: This symptom occurs when a member link of L2 port-channel that is passing traffic is removed from the port-channel.

Workaround: Remove and add the port-channel configurations again.

• CSCtk35953

Symptoms: The dampening information will not be removed even if dampening is unconfigured in VPNv4 AF.

Conditions: This symptom is observed only if DUT has an eBGP-VPNv4 session with a peer and a same-RD import happens on the DUT for the route learned from the VPNv4 peer.

Workaround: A hard reset of the session will remove the dampening information.

• CSCtk59347

Symptoms: CPU is busy and console is locked up for minutes after entering the clear counter command.

Conditions: This symptom occurs with a large-scale configuration with hundreds of interfaces and service groups configured on the system.

Workaround: Instead of clearing all counters of all interfaces, clear the counters of specific interfaces as needed.

• CSCtk64538

Symptoms: The ip igmp join-group and ipv6 mld join-group commands will not work as expected on the Cisco 7600 platform.

Conditions: This symptom occurs with basic configurations of join group on Cisco 7600 series routers.

Workaround: Use the ip igmp static-group or ipv6 mld static-group commands instead.

Futher Problem Description: The ip igmp join-group and ipv6 mld join-group commands are not normal configurations on the Cisco 7600 router. They cause traffic to be punted to RP CPU and cause problems.

• CSCtk67658

Symptoms: Traceback and infrequent crash of the new active are seen when SSO is performed on a router.

Conditions: This symptom occurs when SSO is performed on a router.

Workaround: There is no workaround.

• CSCtk67768

Symptoms: RP crash is observed in DHCPD receive process.

Conditions: This symptom occurs on the DHCP server that is used on Cisco ASR routers and acting as ISG.

Workaround: There is no workaround.

• CSCtk68890

Symptoms: The router configured with an empty export-map crashes upon sending out refreshed routes using the clear ip bgp * soft command.

Conditions: This symptom occurs in a rare situation when the router is configured as a PE and ASBR, and crashes upon configuring an empty export-map without any ext-community RTs.

Workaround: Do not configure an empty export-map.

• CSCtk76697

Symptoms: Service instances on the line card go to the down state for the approximately first 100 service instances of 4000 service instances after a test crash on the line card, resulting in a complete traffic drop on these service instances.

Conditions: This symptom occurs only during the first test crash on the LC after booting up the router.

Workaround: A shut/no shut on the service instance/interface would resolve this issue.

• CSCtk83376

Symptoms: The subscriber is not able to get authorized after initial access-reject.

Conditions: This symptom occurs when a subscriber supplies wrong credentials and access is rejected. Then, the subscriber provides correct credentials, but the access is rejected. The subscriber is not able to log on even with the correct credentials after initial access-reject.

Workaround: Add the following to reset the session:

• CSCtk95106

Symptoms: CPU 1 of SPA 8XT1E1 goes into a forced reload followed by a software forced reload of line card SIP-200 when a multilink PPP with interleave enabled having fragment size 42 is disabled and enabled. One member of the link is removed.

Conditions: This symptom is observed when traffic is pumped onto the DUT from the remote end. The size could be as low as 800 bytes. Interleave is disabled and enabled on the mulilink interface, and one of the members of the MP is detached from the bundle using the no ppp multlink group <> command.

Workaround: There is no workaround.

• CSCtk97082

Symptoms: IPv6 addresses are not cleared from an interface when vrf forwarding is applied for an IPv4-only VRF defined with the vrf definition CLI.

Conditions: This symptom occurs with the following conditions:

– The VRF is defined using the vrf definition CLI.

– The address-family ipv4 command is configured.

– The address-family ipv6 command is not configured.

– An IPv6 address is present on the interface.

– vrf forwarding is then configured on the interface.

Under these conditions, the IPv6 address will not be removed from the interface when vrf forwarding is applied.

Workaround: Clear IPv6 addresses from the interface before applying vrf forwarding.

• CSCtl00127

Symptoms: The output of show ip int command does not indicate whether the “ip security ignore-cipso” option is configured and/or operational.

Conditions: Configure “ip security ignore-cipso” on an interface. This was not indicated on the show ip interface interface-name output of that interface.

This symptom is observed on the following devices:

– Cisco IOS Catalyst 6500 router that is running Cisco IOS Releases 12.2(33)SXH and 12.2(33)SXI.

– Cisco IOS Catalyst 7600 router that is running Cisco IOS Releases 12.2(33)SRA7, 12.2(33)SRB, 12.2(33)SRC, 12.2(33)SRD, and 12.2(33)SRE.

– Cisco IOS Catalyst 4500 router that is running Cisco IOS Release 12.2(40)SG.

The output is indicated correctly when it is enabled on Cisco IOS Release 12.2(18)SXF17a.

Workaround: There is no workaround.

• CSCtl04285

Symptoms: After provisioning a new BGP session, a BGP route reflector may not advertise IPv4 MDT routes to PEs.

Conditions: This symptom is observed on a router running BGP, configured with new style IPv4 MDT and peering with an old style IPv4 MDT peer. Affected releases are Cisco IOS Release 12.2(33)SRE, 15.0M, and 12.2(33)XNE and later releases.

Workaround: There is no workaround.

• CSCtl05785

Symptoms: Connectivity is broken on Cisco 7600 L3 subinterfaces upon reconfiguration of the assigned VRF. Directly connected devices are no longer reachable. Input path is broken (packets are seen in netdr but do not reach the RP).

Conditions: This symptom is observed on Cisco 7600 routers that are running Cisco IOS Release 12.2(33)SRE2. This issue is seen on Sip-400 subinterfaces.

Workaround: Reload the router.

• CSCtl21695

Symptoms: An LNS configured for PPTP aggregation might stop accepting new PPTP connections after PPTP tunnels exceed one million. Debug vpdn l2x ev/er shows:

Conditions: This symptom occurs when LNS is configured for PPTP aggregation and over one millions tunnels have been accepted (on VPDN level).

Workaround: Reload LNS.

• CSCtl54033

Symptoms: Resignaling sub-LSPs for P2MP TE tunnels may take up to 10 seconds, after the sub-LSP has been pruned or torn down.

Conditions: This symptom occurs when a P2MP TE tunnel is configured to request FRR protection, but for the physical link down the path on the tunnel headend, there is no backup tunnel configured at the failure point (TE tunnel headend) to protect the sub-LSP. The TE tunnel headend will take 10 seconds for sub-LSP resignaling.

Workaround: Configure FRR backup tunnels at the TE tunnel headend to provide link protection for P2MP TE tunnels for the physical link that is connected to the TE tunnel headend in the TE tunnel path.

• CSCtl67195

Symptoms: The following three BGP debug commands are not allowed to enable:

– debug ip bgp vpnv4 unicast

– debug ip bgp vpnv6 unicast

– debug ip bgp ipv6 unicast

Conditions: This symptom is observed with the above BGP debug commands.

Workaround: There is no workaround.

• CSCtl71478

Symptoms: In an HA system, the following error message is displayed on the standby RP and LC:

Conditions: This symptom is observed on standby/LC modules when you bring up both the RP and standby/LC routers with or without any configuration.

Workaround: There is no workaround.

• CSCtl88066

Symptoms: A router reloads (seen with a Cisco ASR 1000 Series Aggregation Services router) or produces a spurious memory access (seen with most other platforms).

Conditions: This symptom is observed when BGP is configured and you issue one of the following commands:

– show ip bgp all attr nexthop

– show ip bgp all attr nexthop rib-filter

Workaround: Do not issue either of these commands with the all keyword. Instead, issue the address family-specific version of the command for the address family you are interested in.

For example, the following are safe:

– show ip bgp ipv4 unicast attr nexthop

– show ip bgp attr nexthop

– show ip bgp vpnv4 vrf vrfname attr nexthop

Further Problem Description: While the show ip bgp all attr nexthop has never done anything that show ip bgp attr nexthop did not do, the reload bug was introduced during the development of multi-topology routing. All versions of Cisco IOS which include multitopology routing or which are derived from versions which included multitopology routing, and where this fix is not integrated are impacted.

The fix prevents the issuing of commands beginning with show ip bgp all attr.

• CSCtl93514

Symptoms: QoS configurations do not get applied on the interfaces when the router is upgraded from ES20 to ES+.

Conditions: This issue happens when the ES20 is replaced with ES+. Remove the ES20 LC and insert the ES+ LC on the same slot.

Workaround: Remove all QoS policies applied on the ES20 interfaces. Insert ES+ and reapply all QoS policies once the ES+ interfaces are up.

• CSCtl98132

Symptoms: XDR CPU hog may cause system crash.

Conditions: This symptom occurs when a double failure, such as SSO switch and FRR cutover, causes XDR CPU hog and crashes the system.

Workaround: There is no workaround.

Further Problem Description: The crash can be avoided if the system has no double failure.

• CSCtn01832

Symptoms: The following command sequence crashes the router at check syntax mode:

– config check syntax

– route-map hello

– match local-preference

– no match local-preference

Conditions: This symptom is observed with the commands mentioned above.

Workaround: There is no workaround.

• CSCtn10922

Symptoms: A router configured with “atm route-bridged ip” on an ATM subinterface may drop multicast traffic, and in some cases, may undergo a software initiated reload due to memory corruption. This issue is also evidenced by the presence of an incomplete multicast adjacency on the ATM subinterface.

Conditions: This symptom is observed on ATM subinterfaces that are configured with “atm route-bridged ip” and forwarding multicast traffic.

Workaround: Configure the ip pim nbma-mode command on the point-to-point ATM subinterfaces.

• CSCtn15317

Symptoms: Traffic on MPLS VPN is dropped. When you check LFIB information on the P router, the entry has an instruction to TAG all packets that are destined to the PE router instead of a POP instruction which is expected on a directly connected P router.

Conditions: This symptom occurs with the following conditions:

– The ISIS protocol is running as IGP on MPLS infrastructure.

– ISIS on the PE router is summarizing network that includes BGP vpnv4 update-source

– The P router is running an MFI-based image.

Workaround 1: Remove the summary-address command in ISIS on PE.

Workaround 2: Change the BGP update source.

• CSCtn16840

Symptoms: VPLS imposition traffic does not go through for some of the VCs when the core is a port channel on ES20.

Conditions: This symptom is observed when core facing is a port channel on ES20.

Workaround: Do a shut/no shut on the port channel.

• CSCtn17680

Symptoms: When performing an OIR on a Cisco WS-X6708 module, the router may crash. When inserting the card, the following message is displayed:

Conditions: This symptom is observed on Cisco 7600 series routers with either a single or dual RSP720 supervisor. In the case of dual supervisors, both supervisors crash. The cause of the crash is unknown. However, after the router reloads, the affected module is installed again without further issue in a couple of instances.

Workaround: There is no workaround.

• CSCtn19178

Symptoms: If you are running an Inter-AS MPLS design across two autonomous systems, the router may clear the local label for a working VRF “A” and a new local label will not be reassigned.

Conditions: This symptom occurs on the MPLS Edge LSR when you remove the configuration of an unused VRF “B”, including:

– The vrf interface, for example, no interface Gi1/0/1.430

– The same vrf process, for example, no router ospf process id vrf vrf name

Run the following commands to verify whether you are facing this issue:

– show ip bgp vpnv4 vrf A subnet (this is for the working vrf)

– show mpls forwarding-table labels local label

Workaround: To reprogram a new local label on the PE router, clear the MP-BGP session by using either of the following commands:

– clear ip bgp mp-bgp neighbor soft in

– clear ip bgp mp-bgp neighbor soft out

• CSCtn21561

Symptoms: A crash occurs following DNS translation by NAT.

Conditions: NAT is configured.

Workaround: Identify the responding server of the DyDNS updates and block traffic destined to it to prevent it from being NATted. This server may be the IANA “blackhole” server address.

• CSCtn22728

Symptoms: See the following:

Conditions: This symptom is seen when using unsupported interface CLI option with destination keyword in ERSPAN source session configuration, which may result in Config-Sync failure between active and standby-RP, therefore reloading standby-RP.

Workaround: Do not issue not applicable commands.

• CSCtn26307

Symptoms: In a scaled setup, a new subinterface will behave as expected during the first 20 minutes, and then will stop working.

Conditions: This symptom is observed in a scaled setup after deleting and recreating subinterfaces. Though the sequence of commands to reproduce this is not yet clearly identified, it seems to be triggered by deleting interfaces and some timing issues. There should be an entry for the interface/VLAN in the hidden VLANs section of the show platform vlan command. If the entry is missing, then this is probably the defect that is being encountered. There will probably be an entry in the recycled VLANs for that interface instead.

Workaround: A reboot will resolve this issue.

• CSCtn41245

Symptoms: Subinterface ingress stats do not work for access subinterfaces.

Conditions: This symptom is observed only for the access subinterface. Interface stats and regular subinterface stats work as expected.

Workaround: There is no workaround.

• CSCtn41662

Symptoms: Standby RP crashes sometimes when policymap configuration is done. This crash happens randomly with the following crash decode:

Conditions: This symptom occurs randomly when policy-map, class-map is modified and applied on different interfaces. This does not happen consistently.

Workaround: There is no workaround.

• CSCtn45777

Symptoms: Align messages are seen when enabling the debug cwan atom debug command.

Conditions: This symptom is observed when the cwan atom debug command is enabled. Spurious memory access messages are seen on the router console.

Workaround: There is no workaround.

• CSCtn53094

Symptoms: The router crashes or generates the following error:

Conditions: This symptom is observed when toggling very fast between the ip pim <mode> and no ip pim commands on an interface when that interface is the only one where PIM is being enabled. The most common way this can happen in a production network is through the use of “config replace”, which results in the toggling of the command from ON to OFF and then ON on a different interface.

Workaround: Avoid fast toggling of the pim <mode> command if possible when it is only present on a single interface.

• CSCtn53222

Symptoms: The reals are stuck in READY_TO_TEST state and they never come to OPERATIONAL state. The only way to make them operational is to make them OUTOFSERVICE and INSERVICE again.

Conditions: This symptom occurs when the real moves to FAILED state because of real failure that is detected by the inband failure mechanism. After the retry timeout, the real will be moved to READY_TO_TEST state.

Workaround: There is no workaround.

• CSCtn59698

Symptoms: When MLP bundle comes up on LNS with conditional debugging based on username enabled, certain attributes like IDB description and IP-VRF are not applied on the MLP bundle Virtual-Access.

Conditions: This symptom is observed with the following conditions:

1. Only for MLP sessions on LNS.

2. When you configure per-user attributes in the user’s Radius profile such as “ip:vrf-id” and “ip:description”.

3. When you bring up the session.

4. When you run “show interfaces <Virtual-Access intf> configuration” for both the member-link VA and bundle VA.

5. When the VRF and IDB description sent by Radius is applied only on member link VA and not on bundle VA.

Workaround: Do not enable conditional debugs like “debug condition username <user-name>”.

• CSCtn62250

Symptoms: After upgrade on the Cisco 7600 router from Cisco IOS Release 12.2(33)SRD5 to Cisco IOS Release 12.2(33)SRE3, there may be a problem with PIM MDT neighbors, which do not get brought up, though the configuration is not changed.

Conditions: This symptom is observed after upgrade on the Cisco 7600 router from Cisco IOS Release 12.2(33)SRD5 to Cisco IOS Release 12.2(33)SRE3.

Workaround: Remove/reinsert the mdt default command in ip vrf configuration mode.

• CSCtn68329

Symptoms: When source and receivers are in the same VLAN, receivers are unable to receive multicast traffic unless IGMP snooping is disabled for the VLAN.

Conditions: This issue is not seen when VLAN is in global routing table (no MVPN).

Workaround: Disable IGMP snooping for the VLAN.

• CSCtn73941

Symptoms: After doing an OIR for an ES+ card having EVC configuration with the module clear-config command enabled, restoring the old configuration does not work anymore, indicating that traffic will not be forwarded over those service instances. The VLANs used in the previous configuration cannot be effectively used on those ports, not even by changing the service instance numbers. It is observed that the Cisco IOS software still believes that the port is configured though there is no configuration yet.

Conditions: This symptom occurs only with module clear-config configured.

Workaround: There is no workaround. A complete reload would probably resolve this issue.

• CSCtn74673

Symptoms: After reload, incoming mcast traffic is punted into the CPU before MFIB is downloaded into line cards. Due to the CPU rate being high, the line cards are stuck in a continual loop of failing to complete MFIB download.

Conditions: This symptom is observed when high CPU utilization is caused by multicast traffic and the show mfib linecard command does not show cards in sync and tables are in “connecting” state. The clear mfib linecard command does not correct the line card table states.

Workaround: There is no workaround other than line card reload.

• CSCtn80120

Symptoms: Vlan translation in ES+ line cards is not working when ports are configured as Layer 2 switch ports (as in LAN cards).

Conditions: This symptom is observed when you configure vlan translation in ES+ line cards.

Workaround: There is no workaround

• CSCtn89179

Symptoms: Output drops are observed when traffic is sent beyond 64k rate with single E1 when E1 is configured as unframed. Issue is seen rarely with using time-slots 1-31. After LC OIR, this symptom is not observed. If the channel is removed and attached, this issue reappears.

Conditions: This symptom occurs on the following hardware and software:

– Hardware: SIP: 7600-SIP-400, SPA: 7/1 8xCHT1/E1 SPA

– Software: Cisco IOS c7600rsp72043-adventerprisek9-mz.122-33.SRD or later releases

Workaround:

1. Apply a service policy similar to below:

2. Reload the LC.

• CSCtn90664

Symptoms: On a Cisco 7600 router, which has globally configured “mls qos protocol arp police <value>”, packets which are received on an ES+ switchport/SVI interface bypass the policer and cause high CPU.

Conditions: This symptom is observed on an ES+ switchport/SVI interface with “mls qos protocol arp police <>” enabled on the router.

Workaround 1: Broadcast storm control could be used to rate-limit arp broadcast packets.

Workaround 2: The following policy can be configured on the interfaces (applicable only after Cisco IOS Release 12.2(33)SRE3, 15.0(01)S2, 15.1(01)S01, and 15.1(2)S onwards):

Policy-map ingress_policy-map

And then dscp-transparency enabled using the following CLI:

• CSCtn95344

Symptoms: After RPR downgrade from SRE2 CCO to SRE1 CCO, the standby RSP gets stuck in cold bulk and reboots every 50 minutes.

Conditions: This symptom occurs after RPR downgrade from SRE2 CCO to SRE1 CCO.

Workaround: Perform reload on the router.

• CSCtn98966

Symptoms: In the following topology, the port-channel link on the standby PoA may forward packets unexpected to DHD. The issue is observed in both Cu’s environment and the test lab:

Topology:

-------------------------------POA-1(active)

| |

DHD |(L3 ICC link and L2 Trunk)

| |

----------------------------POA-2(standby)

In Cu’s environment: When DHD sends an ARP request to ask for MAC of an HSRP virtual IP, it will receive the ARP reply from the standby PoA, causing MAC flapping on DHD. In the lab test environment: When you configure static ARP on PoAs to bind an IP address with a nonexistent MAC address, ping this IP, so it will do unicast flooding within VLAN. When you ping, POA-2(standby) also sends out the unicast packet to DHD via its port-channel link.

Conditions: This symptom occurs both on SRE2 and SRE3 with MLACP deployment.

Workaround: There is no workaround.

• CSCto02448

Symptoms: On doing an inbound route refresh, the AS-PATH attribute is lost.

Conditions: This symptom is observed with the following conditions:

1. The neighbor is configured with soft-reconfiguration inbound.

2. The inbound routemap is not configured for the neighbor

3. The non-routemap inbound policy (filter-list) allows the path.

Workaround: Instead of using the non-routemap inbound policy, use the routemap inbound policy to filter the prefixes.

• CSCto04593

Symptoms: StatID leak in line card is observed while churning pppoe sessions when using “show plat npc xlif 0 statid-usage”. The statid leak results in high LC CPU, when it runs out of stat ids.

Conditions: This symptom is seen only with scale.

Workaround: There is no workaround.

• CSCto04953

Symptoms: On a multilink with strict priority, after several hours of traffic load, the traffic gets dropped to 50 to 70 percent. At the same time, there is a 100 percent drop on default class. At worst, 100 percent of the strict priority queue can be dropped.

Conditions: This symptom is observed when the traffic load is close to the given bandwidth.

Workaround 1: Stop the traffic; the issue goes away for several hours.

Workaround 2: Remove and add the policy back; the issue goes away for several hours.

Workaround 3: Unconfigure and reconfigure the impacted interface multilink.

• CSCto07919

Cisco IOS Software is affected by two vulnerabilities that cause a Cisco IOS device to reload when processing IP version 6 (IPv6) packets over a Multiprotocol Label Switching (MPLS) domain. These vulnerabilities are:

– Crafted IPv6 Packet May Cause MPLS-Configured Device to Reload

– ICMPv6 Packet May Cause MPLS-Configured Device to Reload

Cisco has released free software updates that address these vulnerabilities.

Workarounds that mitigate these vulnerabilities are available.

This advisory is posted at http://www.cisco.com/warp/public/707/cisco-sa-20110928-ipv6mpls.shtml.

• CSCto10336

Symptoms: The LNS router hangs up at the interrupt level and goes into an infinite loop.

Conditions: This symptom occurs during control channel cleanup.

Workaround: There is no workaround. This symptom can be only removed through power cycle.

• CSCto15040

Symptoms: When configuring a service instance under the physical interface, the service instance may not be programmed properly on the Switch Processor or the line card, leading to loss of connectivity.

Conditions: This symptom is observed when configuring the service instance under the physical interface of an ES+/ES20/SIP-600 card. This issue is seen with Cisco IOS Release 12.2(33)SRE or later releases.

Workaround: Configure the port in a channel-group and move the service instance configuration under the port-channel interface.

• CSCto16106

Symptoms: Address not assigned when “ip dhcp use class aaa” is configured.

Conditions: When the DHCP server is configured to download a class name from RADIUS using “ip dhcp use class aaa” and lease an IP address from that class, the IP address is not assigned to the client.

Workaround: There is no workaround.

• CSCto29720

Symptoms: Packets drop in the LLQ queue without any congestion on the link when the line card is SIP-400.

Conditions: This symptom occurs when LLQ is configured under Shaper on the physical interface and the line card is SIP-400.

Workaround: There is no workaround.

• CSCto43154

Symptoms: A Cisco device running Cisco IOS may reload unexpectedly with the following message:

Conditions: This symptom is observed on a Cisco device running Cisco IOS.

Workaround: There is no workaround.

• CSCto55643

Symptoms: High CPU loading conditions can result in delayed download of multicast routes to line cards, resulting in multicast forwarding (MFIB) state on line cards out-of-sync with the RP. The show mfib linecard command shows line cards in sync fail state with many in LOADED state.

Conditions: This symptom occurs during high CPU loading due to router reload or line card OIR events in a highly scaled multicast environment with high line rates of multicast traffic and unrestricted processed switched packets, before HW forwarding can be programmed.

Workaround: There is no workaround. Ensure that mls rate limits are properly configured.

Further Problem Description: IPC errors may be reported in the MRIB Proxy communications channel that downloads multicast routes to line cards.

• CSCto55983

Symptoms: After reload, incoming mcast traffic is punted into the CPU before MFIB is downloaded into line cards. Due to the high CPU rate, line cards are stuck in a continual loop of failing to complete MFIB download and retrying.

Conditions: This symptom occurs during high CPU utilization caused by multicast traffic. The show mfib line summary does not show cards in sync.

Workaround: There is no workaround.

• CSCto64240

Symptoms: Port-channel access sub-interface with three memberlinks cannot be configured.

Conditions: This symptom occurs when the port-channel has more than two members.

Workaround: There is no workaround.

• CSCto70972

Symptoms: Multicast traffic drops and does not reach the corresponding entries like (*,G/m) or (*,G).

Conditions: This symptom occurs when multicast traffic drops and does not reach the corresponding entries like (*,G/m) or (*,G).

Workaround: There is no workaround.

• CSCto74038

Symptoms: After an upgrade to SRE3, the CESoPSN (clock) pseudowire stays down due to payload size value mismatch.

Conditions: This symptom occurs when, before the upgrade to SRE3, the payload size is configured to 80 and dejitter value is the default (5). After the upgrade, the payload size 80 and dejitter 5 combination is not accepted anymore as it is not the recommended value, so the payload size is removed from the configuration. The pseudowire is therefore configured with the default payload size. The default value is not accepted by the remote end of the pseudowire, thus leading to payload size mismatch.

Workaround: Configure an acceptable dejitter value, and then reconfigure the payload size.

• CSCto80714

Symptoms: Prowler SPA goes out of service with heartbeat failures when traffic flows through the MLPPP (multilink) interface. This issue is seen only in the Cisco IOS Release 12.2SRE throttle and not in mcp_dev. Some optimizations and a microcode reload-related fix is also included as part of this DDTS.

Conditions: This symptom is observed when traffic flows through the MLPPP interface on Prowler. Microcode and SPA reload is required to recover.

Workaround: There is no workaround.

• CSCto83073

Symptoms: A Cisco 7600 router running Cisco IOS Release 12.2(33)SRE3 crashes when installing a new certificate or when authenticating a trustpoint.

Conditions: This symptom occurs with the following conditions:

1. If the router loads the CA certificate. If the trustpoint has been already authenticated, this symptom will occur usually after a reload.

2. After you configure a new trustpoint and try to authenticate it.

Workaround: There is no workaround.

• CSCtq10019

Symptoms: After router reload, rate-limiters for multicast do not come into effect and packets are punted.

Conditions: This symptom occurs during high CPU load when mfib is unable to distribute into lc and SP.

Workaround: There is no workaround.

• CSCtq09088

Symptoms: The router crashes while trying to unconfigure “ip rsvp sender-host 10.0.0.5 10.0.0.1 UDP 11 11 10 10 identity bogusID”.

Conditions: This symptom is observed on the Cisco 7200 router running the c7200-adventerprisek9-mz.122-33.3.13.SRE image.

Workaround: There is no workaround.

Resolved Caveats—Cisco IOS Release 12.2(33)SRE3

Cisco IOS Release 12.2(33)SRE3 is a rebuild release for Cisco IOS Release 12.2(33)SRE. The caveats in this section are resolved in Cisco IOS Release 12.2(33)SRE3 but may be open in previous Cisco IOS releases.

• CSCsb87956

Symptoms: A software loop enters the IKE_CONFIG_MODE state.

Conditions: This symptom is observed when the following events occur:

1. A user is normally connected to the IPSec gateway trough a NAT-T device.

2. The physical connection goes down on the user side, the interface is shut down, and SafeNet disconnects, but the connection is still up on the IPSec gateway.

3. The physical connection goes up on the user side and a new connection is attempted before DPD causes the old connection to go down on the IPSec gateway.

Workaround: After a disconnection, wait for DPD to tear down the IPSec/ISAKMP sessions. In a production network, there is no workaround.

• CSCsj70622

Symptoms: Cisco router running Cisco IOS Release 12.2(33)SRD or 12.2(33)SRE may experience a memory leak due to crypto processes using MallocLite.

Conditions: The symptom is observed when crypto is configured.

Workaround: There is no workaround.

• CSCsl18054

Symptoms: A local user created with a one-time keyword is removed after unsuccessful login attempts. A one-time user should be removed automatically after the first successful login, not after failed logins.

Symptoms: This symptom occurs on a router running Cisco IOS Release 12.4.

Workaround: There is no workaround.

• CSCso20810

Symptoms: A buffer leak may occur when a router is configured with the combination of NAT, multicast and encryption. This sympotm occurs when multicast subsequently flows out a crypto-enabled interface.

Conditions: This symptom will effect only those users whose routers are part of a multicast group. They must also have NAT and crypto configured on one or more of the interfaces in the multicast group.

Workaround: Multicast traffic can be forwarded via a GRE tunnel instead of in the clear.

• CSCsq40621

Symptoms: Accounting fails due to No Acct-Input/output-Octets/packets field in the STOP accounting records. Many sub-test cases are failing for the same reason in the ipsec accounting script. A session startID and stopID exists, but it shows as “0” and does not have the following fields:

The traffic is flowing successfully but the outputs are not available in the stop accounting records stored in the radius server.

Conditions: This symptom does not affect IKE/ipsec functionality.

Workaround: There is no workaround.

• CSCsv30540

Symptoms: The error message %SYS-2-CHUNKBOUNDSIB and a traceback are seen.

Conditions: These symptoms are observed when the show running-config/write memory command is entered.

Workaround: There is no workaround.

• CSCsv70157

Symptoms: On a Cisco 7609 router running Cisco IOS Release 12.2(33)SRD, after configuring any interface with any carrier-delay value other than 0 (the default), upon entering wr mem you get an unexpected warning message:

Conditions: No special conditions are required to reproduce this defect. Simply configure any interface with a carrier delay and execute wr mem:

Workaround: There is no workaround.

Further Problem Description: This warning should come when Asymmetric Carrier Delay (ACD) is configured on an interface and you then attempt to configure Standard Carrier Delay via the carrier-delay delay interface command. However, the message is coming when ACD is not configured.

• CSCsw21000

Symptoms: Active-RP crash always occurs with core/crashinfo by an abnormal DHCPv4 sequence. This problem is seen soon after the abnormal sequence starts.

Conditions: This symptom is observed under the following conditions:

– ASR is set as a dhcp relay agent

– 8kvlan configuration

– 8port-channel configuration

– Not any other traffic and stress

Workaround: There is no workaround.

Further Problem Description:

– Problem may be memory corruption issue.

– DHCP request packet is abnormal in this case.

• CSCsy16092

Symptoms: A router running Cisco IOS or Cisco IOS XE may unexpectedly reload due to watchdog timeout when there is a negotiation problem between crypto peers. The following error will appear repeatedly in the log leading up to the crash:

Conditions: When a malformed payload (Transform payload with vpi length =0) is received and “debug crypto isakmp” is enabled, the error messages are repeatedly seen leading up to the crash.

Workaround: Remove this debug command.

• CSCsy61302

Symptoms: A chunk header corruption and a router crash with BADMAGIC error message is seen for either a free or in-use chunk.

Conditions: The symptom is observed when the following SNMP commands are configured:

snmp-server community public ro snmp-server packetsize 17940

The crash is seen upon doing a show run and doing a grep for some keyword (e.g.: show run | inc mem). Memory checks need to be enabled. To see this issue reasonably fast, the interval of memory checks needs to be in the order of 3-4 seconds.

Workaround: Do not configure “snmp-server packetsize more than 2048".

Further Problem Description: This crash is seen because of the snmp-server packetsize 17940. There is a local variable in one of SNMP functions with the configured packet size and when we run the CLI show run, the exec process stack overflows and corrupts the subsequent malloced block. This causes the memory corruption.

• CSCsz82950

Symptoms: A peer RP reloads.

Conditions: If any configurations are done using NMS for DCTM MIB, this symptom occurs when unconfiguring the configuration that is created by DCTM MIB configuration.

Workaround: There is no workaround.

Further Problem Description: DCTM was not HA supported before. HA is supported now. If configurations are not done by using NMS, there will not be any issues.

• CSCta26520

Symptoms: The following traceback is seen:

Conditions: This symptom is observed when numerous tunnel interfaces are rapidly added and removed.

Workaround: There is no workaround

• CSCta33011

Symptoms: You may not be able to terminate PPPoE sessions on a Cisco ASR P2. The issue starts after days of normal working operation.

Conditions: The symptom is observed on a Cisco ASR RP2 configured as an LNS.

Workaround: There is no workaround.

Further Problem Description: Except PPP sessions, other functionality works fine. Only PPP is in a stuck state and reload is the only option to recover from this state.

• CSCta43825

Symptoms: A CMTS walk of the ARP table causes high cpu usage. This symptom is also seen with an SNMP walk of the ARP table.

Conditions: This symptom is observed in Cisco IOS Release 12.2S.

Workaround: To prevent high cpu usage due to SNMP walk, implement SNMP view to prevent SNMP walk of the ARP table:

Further Problem Description: This symptom is widely observed in Cisco IOS Release 12.2S since the arp redesign in 2004. It is not an efficient way to do next search/tree walk. When there are a lot of arp entries, the CPU utilization can reach as high as 99% when polling ipNetToMediaTable or atTable (they share the same logic).

• CSCta53372

Symptoms: A VPN static route is not seen in the RIB after an interface is shut down and brought back up (shut/no shut).

Conditions: Configure the crypto client and server routers in such a way that the session is up and RRI installs a static route on the server that is pointing to the client IP address. Now shut down the interface on the server router that is facing the client. The RRI static route disappears from the RIB and never reappears.

Workaround: Reset the RRI session.

• CSCtb07984

Symptoms: A Cisco ASR router acting as LNS fails to apply D2 QoS on the first few sessions after every new reboot and configures the D2 QoS on all subsequent sessions.

Conditions: The symptom is observed when multiple routes exist on an LNS router to reach LAC router. PPPoX sessions are brought on LNS with D2 QoS model after new reboot of router.

Workaround 1: LNS router configures D2 QoS on all subsequent sessions in Cisco IOS Release 12.2XND images.

Workaround 2: In Cisco IOS Release 12.2XNE images, LNS router should have a single route to reach LAC router.

Workaround 3: Wait until CEF is converged before bringing up a second session on the LNS router.

• CSCtb32043

Symptoms: CPUHOG messages may be displayed or Cisco IOS might crash when executing no ipv6 multicast-routing in a configuration with more than 20,000 IPv6 multicast-enabled interfaces or sub-interfaces.

Conditions: This symptom is observed only rarely when an alternate software path is taken. It is not known what causes this alternate path to be taken.

Workaround: There is no workaround.

• CSCtb32892

Symptoms: Tracebacks such as:

may be seen on a router console when loading an image or during an RP SSO.

Conditions: The symptom is observed upon reloading a Provider Edge (PE) router with an mVPN configuration or during a simple SSO. It is observed on the standby RP.

Workaround: There is no workaround.

• CSCtb47647

Symptoms: Active RP crashes at pim_send_join_prune.

Conditions: The symptom is observed when performing some PIM-related testing with specific configurations and after carrying out an SSO. When you attempt to debug memory leak issue using a memory traceback recording command, the router crashes while executing the command show memory traceback exclusive.

Workaround: There is no workaround.

• CSCtb78752

Symptoms: A Cisco device crashes when module 1 is reset.

Conditions: This symptom is observed when the adjacency for the destination IPv6 address lookup is not formed for packets coming into the tunnel interface (leaving the tunnel).

Workaround:There is no workaround.

Further Problem Description: After an IPv6 automatic tunnel is established between 2 UUT, when reflexive acl is applied in one of the uuts, and after telneting the ping device configured with the IPv6 address attached to each UUT (ping--uut1--uut2--ping), temporary entries are created in one of the UUTs if the session-closed entries are removed as expected. When module 1 is reset and the telnet session is established, the device crashes.

• CSCtb99914

Symptoms: The EIGRP VRF configuration is missing in named mode.

Conditions: This symptom happens only after a power cycle.

Workaround: Use other modes instead of naming modes. For example:

• CSCtc20254

Symptoms: When the PTA sessions are disconnected from a Cisco router, the active RP logs display an error message and subsequent tracebacks.

Conditions: This symptom is seen in a prepaid scaling scenario.

Workaround: There is no workaround.

• CSCtc33679

Symptoms: Routes are not being controlled properly when PIRO is used.

Conditions: If more than one exit per BR is configured and PIRO is used to control the routes, the nexthop is not being calculated correctly. As a result, traffic for these traffic classes is not taking the correct route.

Workaround: There is no workaround.

• CSCtc55897

Symptoms: R2 will not advertise the routes.

Conditions: The symptom is observed under the following conditions:

1. R2 has two IBDG neighbors in the same update-group, one neighbor with 4BAS and the other with 2BAS capability.

2. The locally originated routes or routes without any AS_PATH will not be advertised to this kind of group.

Workaround: Try to make the 2BAS and 4BAS neighbors fall into different update-groups by configuring dummy route-maps.

• CSCtc78966

Symptoms: IOSD crash is seen while sending traffic through user-defined IP sessions.

Conditions: This issue is seen on a Cisco ASR 1000 router with RP2 processor.

Workaround: There is no workaround.

• CSCtc84758

Symptoms: On a router configured for ISG that is running postpaid Web-Logon users with SESM as the external portal, a memory leak may occur in RADIUS LOCAL SERVER.

Conditions: The symptom is observed on a Cisco 10000 series router with a PRE-3 and running Cisco IOS Release 12.2(33)SB7 using SESM as a captive portal. The issue can be triggered with this sequence of events:

1. Postpaid user is redirected to SESM.

2. SESM sends Access-Request to router after captivating user/pass from postpaid user.

3. RADIUS LOCAL SERVER creates AAA request and sends it to ISG.

4. ISG creates another AAA request to send an Access-Request to authenticate the postpaid user.

5. AAA receives a response from external AAA.

6. AAA passes the response to RADIUS LOCAL SERVER which transmits an Access-Accept or Access-Reject to SESM.

If the processing delay of sum (C,D,E,F) is greater than the SESM timeout, SESM will send another Access-Request with the same credentials for the Account logon postpaid user in B.

If this occurs, policy/AAA will now use this second Account-Logon request from SESM for this user’s Account Login and the policy will not free the AAA request from the former Account Logon request, hence the memory leak will present as RADIUS LOCAL SERVER.

Workaround: Make sure SESM Account Logon Timeout > RADIUS timeout.

Alternate workaround: Decrease load on external AAA (RADIUS) machines.

• CSCtd10712

The Cisco IOS Software network address translation (NAT) feature contains multiple denial of service (DoS) vulnerabilities in the translation of the following protocols:

– NetMeeting Directory (Lightweight Directory Access Protocol, LDAP)

– Session Initiation Protocol (Multiple vulnerabilities)

– H.323 protocol

All the vulnerabilities described in this document are caused by packets in transit on the affected devices when those packets require application layer translation.

Cisco has released free software updates that address these vulnerabilities.

This advisory is posted at http://www.cisco.com/warp/public/707/cisco-sa-20110928-nat.shtml.

• CSCtd33166

Symptoms: A Cisco router may crash at “parse_call_action_func.”

Conditions: This symptom occurs in “before and after” mode when configuring the Call Home feature.

Workaround: Turn off “before and after” mode.

• CSCtd45085

Symptoms: Stale RPs are shown as BIDIR PIM Designated Forwarders.

Conditions: This symptom is observed after a configuration change, where the same RP is configured as BIDIR mode first, then sparse-dense mode.

Workaround: Reconfigure the RP in BIDIR mode.

• CSCtd57788

Symptoms: A dynamic IP ACL is created when a session comes up and is together with the policy private route created according to the “Ascend-Private-Route” downloaded from the user profile. When the session goes down, the route is cleared but the dynamic ACL is not cleared:

Conditions: The symptom is observed with routes downloaded from the radius server.

Workaround: There is no workaround.

• CSCtd57876

Symptoms: A Cisco router may experience process switching and high CPU utilization.

Conditions: This symptom is triggered by the removal of ip nbar protocol-discovery.

Workaround: Reconfigure ip nbar protocol-discovery.

• CSCtd78587

Symptoms: A Cisco Catalyst 6000 switch running Cisco IOS Release 12.2SX software might crash under rare conditions when err-disable recovery tries to recover a port. The following messages are seen in the logs before the switch resets itself: %CPU_MONITOR-6-NOT_HEARD

Conditions: This symptom may be observed after the following sequence of events:

1. An interface on the switch gets err-disabled as expected due to a certain feature; for example, due to BPDU Guard

2. Shortly after, before BPDU Guard err-disable recovery kicks in, the same port gets err-disabled for a different reason; for example, because a diagnostic error is detected on the already err-disabled port

3. Err-disable recovery (BPDU Guard) tries to recover the port and this leads to the crash.

Workaround: Disable err-disable recovery.

• CSCte01606

Symptoms: When Bidirectional Forward Detection (BFD) is enabled, issuing certain CLI commands that are not preemption safe may cause the device to restart. This condition has been seen when issuing commands such as show mem or show mem frag detail.

Conditions: The issue may occur if BFD is enabled on a device that utilizes Pseudo Preemption to implement this feature. The device must be running an affected software build.

Workaround: Disable BFD

PSIRT Evaluation: The Cisco PSIRT has assigned this bug the following CVSS version 2 score. The Base and Temporal CVSS scores as of the time of evaluation are 4.4/3.8:

https://intellishield.cisco.com/security/alertmanager/cvssCalculator.do?dispatch=1&version=2&vector=AV:L/AC:M/Au:S/C:N/I:N/A:C/E:H/RL:OF/RC:C

CVE ID CVE-2010-3049 has been assigned to document this issue.

Additional information on Cisco’s security vulnerability policy can be found at the following URL:

http://www.cisco.com/en/US/products/products_security_vulnerability_policy.html

• CSCte15193

Symptoms: The no spanning-tree vlan vlanno command is not removed on standby alone.

Conditions: The symptom is observed under the following conditions:

– the no spanning-tree vlan vlanno command is configured first

– the default spanning-tree vlan vlan-range command is entered next

– the vlanno falls within the designated range, but the last vlan number in the range does not have “no spanning-tree vlan <>” configured for that.

Workaround: Enter the default spanning-tree vlan vlanno command to remove it.

• CSCte39707

Symptoms: Supervisor crashes unexpectedly when IPv6 multicast is configured.

Conditions: This symptom is observed when IPv6 multicast is configured.

Workaround: There is no workaround.

• CSCte52419

Symptoms: When an SSS session is created and the command show sss session identifier authenticated-username <string> is executed, a memory leak is seen.

Conditions: The symptom is observed when an SSS session exists in the router.

Workaround: There is no workaround.

• CSCte56437

Symptoms: NAT programming on a Cisco Catalyst 6500 may become corrupted; the source and/or destination IP addresses of traffic passing through the NAT box are changed to the wrong IP addresses.

Conditions: This symptom is observed when the NAT configuration is changed during a high-volume traffic session.

Workaround: There is no workaround.

• CSCte60787

Symptoms: A 528-byte memory leak is observed for every radius-proxy session.

Conditions: This symptom is observed upon bringing up and clearing radius- proxy sessions.

Workaround: There is no workaround.

• CSCte65688

Symptoms: Easy VPN server prints “Client_type=UNKNOWN” in “%CRYPTO-6-EZVPN_CONNECTION_UP: (Server)” log when Software VPN Client establishes an IPSec session.

Conditions: The symptom is observed when:

– Easy VPN is configured between a Cisco VPN Client and an IOS router

– crypto logging ezvpn is configured.

Workaround: There is no workaround.

Further Problem Description: This is simply a cosmetic issue. Currently, this message can identify hardware VPN clients (IOS/PIX/VPN3002) only.

• CSCte74705

Symptoms: A Cisco 7600 router that is running Cisco IOS Release 12.2(33)SRD may generate L2 Queue Error (%L2-SP-4-QUEER) messages when a link down/up event occurs across multiple interfaces in a short period of time.

Conditions: This symptom may occur when a large number of attachment circuits are configured with EVC style configuration, and a large number of MAC addresses are known to the system.

Workaround: There is no workaround.

• CSCte78406

Symptoms: The following error message is logged at the new standby RP when PTA sessions are established:

Conditions: The symptom is observed when PTA sessions are established, then an RP switchover is performed. After both RPs sync up, flap the sessions. The error messages are logged at the new standby RP.

Workaround: There is no workaround.

• CSCte85961

Symptoms: The router crashes while doing a shut command followed by the no shut command to the main interface.

Conditions: The issue is seen with scale configuration and giving the shut command followed by the no shut command in the main ATM interface.

Workaround: There is no workaround.

• CSCte89292

Symptoms: There are ping failures on a Cisco 7600 series router with a linecard.

Conditions: The symptom is observed on any linecard having a subinterface connection. The following steps trigger the issue:

1. Have a back-to-back subinterface connection

2. The local interface should be down/down. (i.e.: admin shutdown on the remote interface)

3. Do two consecutive redundancy force-switchover(s)

4. “No shutdown” on remote interface

5. Then ping the remote end from local router on the subinterface.

Workaround:

1. Do admin shutdown followed by no shutdown on remote interface

2. Then ping the remote end from local router on the subinterface. Ping works.

• CSCte95396

Symptoms: A subscriber cannot enable the SSS session due to DPM not finding the binding in the DPM table although the DHCP binding exists as shown by performing the show ip dhcp server binding command.

Debug sss policy event/err would show “SG-DPM: DHCP Binding does not exist to query session”.

Conditions:

– Subscriber has dhcp binding when doing “show ip binding...” note: also check the vrf (if any)

– Subscriber has no entry in the dpm policy

– Session trigger needs to be l2-connect dhcp

Workarounds:

– If this is a "low lease time and relay dhcp case”, make sure subscriber does not send a DHCP packet: waiting for the DHCP binding to disappear (i.e., expire), re-enable the user’s dhcp forwarding path

– If this is a “dhcp server” case, clear dhcp binding on the ISG

– Reload the router

• CSCtf05827

Symptoms: A Cisco 10000 router crashes with chunk error.

Conditions: This symptom occurs due to memory corruption longevity and stress test.

Workaround: There is no workaround.

• CSCtf12072

Symptoms: The expected behavior after a failed authorization action does not get applied onto the session when authorization is based on option 82 information. The FSOL does not contain option 82 information.

Conditions: The symptom is observed when the ISG policy is to provide authorization based on the subscriber’s option 82 information, such as remote-id and/or circuit-id. However, the option 82 is missing in the DHCPDISCOVER packet. The subscriber session comes up in unauthenticate as expected, but the expected actions (i.e.: applying L4R service) do not get applied onto the session.

Workaround: There is no workaround.

• CSCtf19902

Symptoms: For some clients, relaying of DHCP Discover packets is not triggered following session authentication. For a single ISG, this results in the client never receiving an address. For redundant ISGs, where one is affected by this issue and one is not, this results in the affected ISG never clearing the session, even though it sees the request from the client accepting the other ISGs offer.

Conditions: This symptom is seen when service-start event under control- policy is configured to unapply all possible services (including the desired service), then apply the new service.

Workaround: Change the service-start event configuration to only unapply other services, then apply the new service. However, this will require a separate event configuration for each service type.

• CSCtf34720

Symptoms: DR will not send a periodic join for an SSM group with a “static- group” configuration on the RPF interface. This will result in the S,G states expiring in the upstream routers and may result in traffic loss.

Conditions: The symptom is observed when the static-group join is configured on the RPF interfaces and the output interface list of the mroute is NULL.

Workaround: Add a local join by using ip igmp join-group for the same group and source, so that it adds a local interested receiver and sends a periodic join upstream.

• CSCtf36402

Symptoms: A Cisco router crashes when the user telnets and Transmission Control Block is cleared for that session before entering the password.

Conditions: This symptom is observed when aaa authentication protocol is set to TACACS.

Workaround: Do not clear the Transmission Control Block for a session before entering the password.

• CSCtf52106

Symptoms: There is a failure of EEM TCL scripts when using the “exit_comb” keyword for the Interface Event Detector.

Conditions: The symptom is observed when using the “exit_comb” keyword in an EEM TCL script.

Workaround: There is no workaround.

• CSCtf54561

Symptoms: A MPLS TE FRR enabled router can encounter a crash if the show ip cef vrf vrf-name command is issued.

Conditions: This symptom occurs when the VRF contains many entries (17k) in which the outgoing interface changes due to a topology change.

Workaround: Command should not be issued when many topology changes occur on interface flaps.

• CSCtf64375

Symptoms: Memory corruption and router crash are seen with overlapping mac addresses.

Conditions: This symptom is seen when bringing up Cisco 10000 router sessions with overlapping mac addresses at 40CPS each set of 10 sessions having the same mac address.

Workaround: There is no workaround.

• CSCtf71990

Symptoms: An alert message is not sent if “source-ip-address” is configured in the call-home configuration. The following message is shown:

Conditions: The symptom is observed when “source-ip-address” is configured.

Workaround: Remove “source-ip-address”.

• CSCtf72328

Symptoms: BFD IPv4 Static does not fully support AdminDown.

Conditions: The symptom is observed with the following setup and configuration:

Though the BFD state is DOWN the static has the route active. If the BFD peer signals AdminDown on a session being used to monitor the gateway for a static route, no action will be taken.

Workaround: Perform a shut/no shut the interface on which the BFD session is configured.

• CSCtf77047

Symptoms: Ping ATM subinterface peer IP address has packet loss from Cisco 7206.

Conditions: This symptom occurs with the following:

1. NPE-G2+PA-MC-STM-1SMI+PA-A6-OC3SML

2. Enable EIGRP on ATM subinterface

Workaround: There is no workaround.

• CSCtf80105

Symptoms: When basic SIP-SIP calls are placed using automation scripts, calls start failing due to UDP socket connection error

Conditions: The symptom is observed when the router is configured with a dial peer and with SNMP. A dial peer is most likely required to reproduce the issue, but it is possible that a different UDP protocol other than SNMP could also cause the symptom. Once a call failure occurs, all the calls placed later will fail with a UDP socket connection error.

Workaround: Use the following steps:

1. Under sip-ua, configure connection-reuse (which is a hidden command)

2. Configure the use of TCP.

• CSCtf98801

Symptoms: Tracebacks are seen while doing ISSU:

Conditions: This symptom is observed with ISSU from SRD4 to SRE1

Workaround: There is no workaround.

• CSCtg07201

Symptoms: DHCP sessions with FSOL that do not contain option 82 information may get stuck in “Attempting” state during authorization.

Conditions: The symptom is observed when the keepalive feature is applied prior to the authorization action, i.e:

Workaround: While removing the action “5 service-policy type service name KA-SERVICE” in the above example would avoid this problem, keepalive feature is highly recommended when provisioning IP sessions with ISG, or else there is no way for the ISG router to detect and clean-up inactive subscriber sessions.

• CSCtg13269

Symptoms: On peers of Route Reflectors (RR), the received prefixes counter shows an incorrect number when session flaps occur during a network churn.

Conditions: The symptom is observed with BGP RRs.

Workaround: Use the clear ip bgp * command.

• CSCtg18555

Symptoms: A memory leak is observed with process_online_diag_pak.

Conditions: This symptom is observed on a card supporting TestNonDisruptiveLoopback and TestFabricChHealth tests.

Workaround: Disable the HM tests TestNonDisruptiveLoopback and TestFabricChHealth on LCs to stop the leak.

• CSCtg25798

Symptoms: The issue is associated with the two labels imposition for the next-hop address. If there is no label bind for the destination prefix and in order to reach next-hop address the router imposes two labels, only one label is imposed for the final prefix.

Conditions: The symptom occurs when all of the following conditions are met:

1. The prefix does not have a label bind (BGP prefixes for example)

2. There is a static route for the next-hop address pointing to the tunnel only

3. The router imposes two labels for the next-hop address.

Workaround: There are three potential workarounds:

1. Explicit next hop avoiding recursive research: “ip route 192.168.4.4 255.255.255.255 Tu1 192.168.4.4" (i.e.: breaking rule 2)

2. Use “neighbor 192.168.1.1 send-label” on both PEs (i.e.: breaking rule 1)

3. Use “mpls traffic-eng signaling interpret explicit-null verbatim” on P (i.e.: breaking rule 3).

Further Problem Description: In the following example 192.168.200.200 is the final destination. There is no label bind for this prefix and it is recursive to 192.168.100.100:

The next-hop 192.168.100.100 has a static route pointing to the tunnel and is double tagged:

So the traffic to 192.168.200.200 should also be double tagged as shown below:

However traffic is leaving the router only with the tunnel label:

• CSCtg26324

Symptoms: Router acting as a DHCP relay crashes with a CPUHOG.

Conditions: The symptom is observed when there are several thousand DHCP bindings at the time of issuing the no service dhcp command.

Workaround: Remove the DHCP bindings on the router before issuing no service dhcp.

• CSCtg29219

Symptoms: QoS not working when switching to backup pseudowire.

Conditions: This symptom is observed on a multi-chassis LAG set up. When switching over from primary to backup pseudowire, QoS applied on the evc is not working. This is because of a mismatch in QoS programming on the member link and the actual member link used for evc flow.

Workaround: There is no workaround.

• CSCtg35298

Symptoms: Traffic drops are seen between two PEs after re-optimization.

Conditions: The symptom is observed with 16k VPLS VC, 4k scalable EoMPLS, 1K software EoMPLS, 600 primary tunnels to nPE1 and one tunnel to nPE2 from nPE3.

Workaround: There is no workaround.

• CSCtg37885

Symptoms: A Cisco router crashes.

Conditions: This symptom is observed with 25k VCs when the RSVP and OSPF neighbors are cleared in succession.

Workaround: Do not clear the RSVP and OSPF neighbors in succession when a high number of VCs exist.

• CSCtg45099

Symptoms: Router crashes.

Conditions: The symptom is observed when the show cca command is issued.

Workaround: There is no workaround.

• CSCtg49331

Symptoms: Multicast streams may not be forwarded to some interfaces, even though they are forwarded to other interfaces on the device without issues.

Conditions: The symptom is observed on a Cisco 7600 series router that is running Cisco IOS Release 12.2(33)SRD4 with egress multicast replication mode.

Workaround: Use ingress replication mode. If egress replication mode is used and the issue is present, service can be restored by using this command:

clear ip mroute A.B.C.D

Or perform a shut/no shut on the affected interface.

• CSCtg58786

Symptoms: When an external interface on the BR is shut down, the BR could be crashed.

Conditions: If more than one thousand Application Traffic Classes are configured on MC, and if that traffic is traversing through an external interface on a BR, and if the external interface is shut down, this could result in a crash.

Workaround: There is no workaround.

• CSCtg60065

Symptoms: DLFIoATM back-to-back ping fails for AAL5MUX encapsulation.

Conditions: The symptom is observed when you configure DLFIoATM with encapsulation AAL5MUX.

Workaround: There is no workaround.

Further Problem Description: Issue is not seen when you configure for AAL5SNAP.

• CSCtg60201

Symptoms: Unconfiguring the maximum-path command does not trigger a backup path calculation.

Conditions: This symptom is observed if addition-path install is configured along with the maximum-path command.

Workaround: Reconfigure “bgp additional-path install.”

• CSCtg64175

Symptoms: The ISIS route is missing the P2P link, it is mistakenly marked as “parallel p2p adjacency suppressed”.

Conditions: The symptom is observed when the ISIS neighbor is up and multiple topologies are enabled on P2P interfaces. It is seen if you enable a topology on a P2P interface of the remote router and send out the serial IIH packet with the new MTID to the local router where the topology has not been enabled on the local P2P interface yet.

Workaround: Do a shut and no shut on the local P2P interface.

• CSCtg65989

Symptoms: Only the first user is able to get authenticated successfully and browse the internet. All subsequent users are constantly redirected to the web portal after successful authentication. The show sss sess uid xxx command shows that the internet service is not applied to the account even though the session is authenticated.

Conditions: The symptom is observed with customers using web logon with a session applying an auto service.

Workaround: Remove “autoservice” and apply static service at account logon.

• CSCtg73456

Symptoms: Bulk sync fails due to applying the tx-ring- limit command on the main interface, which is not supported.

Conditions: This symptom occurs when applying the tx-ring- limit command on the main interface and doing an SSO.

Workaround: There is no workaround.

• CSCtg74946

Symptoms: QoS counters are stuck at “0.”

Conditions: This symptom is observed with SSO and when dlfi is configured over ATM.

Workaround: Detach and re-attach the service-policy.

• CSCtg76688

Symptoms: An active Cisco route processor reloads in a scale scenario (16k - 24k sessions) when the clear subscriber session all command is entered.

Conditions: This symptom is observed only when there are 16k-24k sessions and the clear subscriber session all command is entered.

Workaround: Do not enter the clear subscriber session all command when more than 16k sessions are up on the router.

• CSCtg79881

Symptoms: Subscriber cannot enable the SSS session due to DPM not finding the binding in the table although the binding exist when performing the show ip dhcp server binding... command. If you use debug sss policy event/err the following message shows:

SG-DPM: DHCP Binding does not exist to query session

Conditions: The symptom is observed under the following conditions:

– Subscriber has DHCP binding when doing show ip dhcp server binding (note: also check the VRF, if any)

– Subscriber has no entry in the DPM policy

– Session trigger needs to be L2-connect DHCP.

Workaround 1: If “low lease time and relay dhcp case”:

– Make sure subscriber does not send a DHCP packet

– Wait for the binding to disappear

– Re-enable the user DHCP forwarding.

Workaround 2: If “dhcp server case”, then clear DHCP binding.

Workaround 3: Reload the router.

• CSCtg89555

Symptoms: There is no forwarding interface seen in the mfib output on a DFC.

Conditions: This symptom is observed when configuring an ip address after multicast has been configured on a dot1Q interface.

Workaround: Performing a shut/no shut of the interface will fix the problem.

• CSCtg94250

Symptoms: Removing address-family ipv4 vrf <vrf> (in router BGP) followed by no ip vrf <vrf> (where “vrf” is the same) could result in a crash.

Conditions: The symptom is observed in a large VPNv4 scale setup, when applying the following commands to the same VRF back-to-back:

1. no address-family ipv4 vrf <vrf>

2. no ip vrf <vrf>

3. ip vrf <vrf>

The trigger of the BGP crash is a result of a racing condition between event 1 and event 2.

Workaround: Since this is a racing condition, the workarounds are:

1. Not applying (1) before (2).

2. Give sufficient time for (1) to complete before applying (2).

• CSCtg98116

Symptoms: An ES-20 crashes on performing a config copy from startup-config to running-config.

Conditions: The symptom is observed with a 4k EVC and QoS policy attached to the EVC when a config copy is performed from startup-config to running-config.

Workaround: There is no workaround.

Further Problem Description: ES-20 recovers and works fine after the crash.

• CSCth00317

Symptoms: When a large number of service groups are configured with multiple service instances on a port-channel, the following anomaly is observed: on addition of a new member-link, not all the policies applied to the port- channel will be configured in the linecard.

Conditions: The symptom is observed upon adding a new member-link (having large policies) to the EVC port-channel.

Workaround: Do a shut/no-shut of the member link.

Alternate Workaround: Reset the linecard on configuration of the port-channel.

• CSCth01394

Symptoms: On a Cisco 7606 router that is running Cisco IOS Release 12.2(33) SRD3 with SIP200/SPA-4XCT3/DS0, when you have ppp multilink interface(s) configured with member links from same SPA (software based multilink) and you physically remove SPA, you will see that upon executing the show ppp multilink command, the multilink interface still has reference for member links. If you do the sh run int serialx/y command, you will get message interface not found.

Conditions: This issue is consistently reproducible.

Workaround: There is no workaround.

• CSCth01526

Symptoms: MDT interface deactivated and activated after an SSO.

Conditions: After an SSO switchover, the PIM register tunnel or MDT tunnel may go down briefly on switching to the standby RP.

Workaround: There is no workaround.

• CSCth02725

Symptoms: There is an interoperability issue between a third-party vendor’s routers and Cisco routers with severe IPTV service failure in Prune-Overriding environment.

Conditions: The symptom is observed in the following scenario:

1. Router A is Cisco 7609 router (IP address 10.1.1.1) and connects to Router B (third-party vendor’s router; IP address 10.1.1.3) and Router C (IP address 10.1.1.2).

2. If subscriber under Router C disappears, Router A receives "Prune" message from Router C.

3. Router A does not change “source IP of PruneEcho message (10.1.1.2)” and sends it to Router B.

4. At this time, Router B should send overriding-join to Router A because Router B still has subscribers. But Router B drops the PruneEcho message because source IP (10.1.1.2) is not from PIM neighbor. Router B cannot send overriding-join to Router A.

5. As a result, multicast traffic (IPTV stream) to Router B stops.

Workaround: Connect C and B to become PIM neighbors. However, this cannot always be considered a recommended workaround because of potential high cost and/or other (sometimes third-party) limitations.

• CSCth02812

Symptoms: A prolonged unicast flood can be seen on an ingress path after a TCN event. The flood will last until entries in the arp table are refreshed.

Conditions: This symptom is observed in Cisco IOS Release 12.2(33)SXH3a (issue has been tracked back to Cisco IOS Release 12.2(18)SXF in an L2 asymmetric environment. The flood is only seen if there is no bi-directional flow on the switch. This issue can be seen in all STP modes.

Workaround: Clearing ip arp will correct this issue. Lowering the arp timeout will also minimize the impact of the flood.

• CSCth05476

Symptoms: On router bootup, the SIP200 linecard is flooded with “%CWSLC-3- DIAGFAIL: Failed to handle diag” messages.

Conditions: The symptom is observed on a Cisco 7600 series router.

Workaround: There is no workaround.

• CSCth05778

Symptoms: Router is showing memory leaks.

Conditions: The symptom is observed when the remote end is sending LCP conf_req messages to a Cisco 10000 series router a lot frequently (1 per 4 msec) than the normal scenario (1 per 2 seconds).

Workaround: Shut down the PPP link that is flapping.

• CSCth11747

Symptoms: When a switchover occurs with GR enabled, sometimes the NSF states are not preserved and the forwarding entries are lost, leading to packet loss for a few seconds.

Conditions: This symptom is observed only with single sessions with GR configured when the restarting neighbor does a passive open. Chances of hitting this are low since this issue occurs because we receive a new open message before the old tcp session has a chance to reset.

Workaround: Configuring multi-session capability on the neighbor sessions or restricting the restarting neighbors connection to active mode would prevent this issue.

Further Problem Description: When an established session already exists between the GR-enabled routers, and the tcp has not yet notified of reset due to neighbor SSO, if the receiving router gets a new open from the restarting router, as per the RFC it is supposed to tear down the old session and accept the new connection. The old session was being torn down properly but it would take the service reset walker to completely free the session. In case of multi-sessions there was no problem in accepting the new session since multiple sessions are allowed. But in case of a single session that already exists, the new sessions are not allowed until the old session is completely freed. Hence, the new session was getting rejected and notification was sent to the restarting neighbor. The restarting neighbor, upon reception of this notification, would clear the NSF preserve bits and further opens would clear the NSF states on the receiving neighbor and hence the problem. The solution would be to accept the new connections in single session support neighbors when the GR reopen has marked the session for reset and de-linked the topologies. The topologies would be added to the new session and the connection accepted. The old session would be freed when service reset walker is invoked. So, for a transient period of time between the session mark reset and the session free, there would be multiple sessions established on the neighbor even though the neighbor was configured as single session. Dependent DDTS CSCtd99802 and CSCth90239 need to be committed along with this fix to ensure complete working of this functionality.

• CSCth13105

Symptoms: Traceback is seen at policymgr_handle_get_context.

Conditions: The symptom is observed while creating a session with many policies attached.

Workaround: There is no workaround.

• CSCth15105

Symptoms: BFD sessions flap after unplanned SSO (test crash).

Conditions: The symptom is observed on a UUT up with unicast/multicast along with BGP and BFD configurations. For BFD timers of 1*5, 500*8, after doing a test crash (option C followed by 6), we see BFD sessions flap.

Workaround: There is no workaround.

• CSCth16011

Symptoms: After a network event is introduced in the network, such as a 3- percent loss, MOS policy will detect the OOP condition. But PfR will let the prefix stay in the OOP condition for some time and then switch over to an alternative exit.

Conditions: Introduce loss to network.

Workaround: There is no workaround.

• CSCth18571

Symptoms: An ES+ module may reload when a SPAN session is configured for a source VLAN.

Conditions: The symptom is observed with a Cisco 7600 ES+ and with Cisco IOS Release 12.2(33)SRD4. VFI configuration needs to be present.

Workaround: There is no workaround.

• CSCth23787

Symptom: A Cisco router crashes at mcast_aaa_send_stop_acct_event.

Conditions: This symptom is observed while unconfiguring “ipv6 mld join-group FF1E:7777:7777::1” in the client after configuring within 15-20 seconds.

Workaround: Unconfigure, if required, after multicast start record is sent.

• CSCth29393

Symptoms: Downstream traffic (to the subscriber) is not forwarded. Only upstream counters are increasing.

Conditions: The symptom is observed with the show sss session detail command with PXF output.

Workaround: Clear the affected SSS session.

• CSCth31271

Symptoms: A Cisco ASR router crashes with next-hop recursive.

Conditions: This symptom is observed after the following tasks are executed:

1. Configure a route-map with recursive next-hop clause for ip address (for example, 1.2.3.4)

2. Change the recursive next-hop to ip address (for example, 5.6.7.8)

3. Apply PBR with this route-map to an interface

4. Delete the route-map

5. Shut the interface.

Workaround: There is no workaround.

• CSCth33500

Symptoms: NAS port is reported as zero on LNS.

Conditions: This symptom occurs when “vpdn aaa attribute nas-port vpdn-nas” is configured.

Workaround: There is no workaround.

• CSCth35356

Symptoms: DHCP relay/server stops working and multiple DHCP processes are created.

Conditions: The symptom is observed when fragmented DHCP packets are received.

Workaround: There is no workaround.

• CSCth37580

Symptoms: Dampening route is present even after removing “bgp dampening”.

Conditions: The symptom is observed under the following conditions:

– DUT connects to RTRA with eBGP + VPNv4. - eBGP + VPNv4 peer session is established and DUT.

– Also DUT has VRF (same RD) as route advertised by RTRA.

In this scenario, when DUT learns the route it will do same RD import and the net’s topology will be changed from VPNv4 to VRF. When dampening is unconfigured, we do not clear damp info.

Workaround: There is no workaround.

• CSCth37998

Symptoms: WEB logon always fails after one failed account logon.

Conditions: The symptom is observed after a failed account logon.

Workaround: There is no workaround.

• CSCth38699

Symptoms: Cisco IOS platforms configured for Auto-RP in a multicast environment lose the RP-to-group mappings.

Conditions: This symptom is observed in Cisco IOS Release 12.2(18)SXF7, Release 12.2(33)SXH4, and Release 12.2(33)SRC4, but it is believed to affect other releases. This symptom occurs when the length of the RP-Discovery packet reaches its limit. If the Mapping Agent receives RP-Announce packets, increasing the number of multicast groups, and that number makes the limit of the packet size, then an empty RP-Discovery packet is triggered that clears the RP-to-Group mapping tables in all the routers receiving such a packet.

Workaround: Configure static RP-to-Group mappings.

• CSCth42594

Symptoms: Remote standby router crashes when you configure and remove “ppp multilink mrru local” under a multilink interface.

Conditions: The symptom is observed with the following conditions:

1. When multilink is bundled with more than one serial interfaces (not seeing this issue with only one serial interface).

2. Seeing this issue from 1500 and above (not seeing this issue when configure and remove “ppp multilink mrru local 1499").

Workaround: There is no workaround.

• CSCth42798

Symptoms: In a very corner case, when BGP is in read-only mode and attributes are deleted before the networks, memory can be corrupted.

Conditions: The device should be in read-only mode, and attributes should be deleted before networks.

Workaround: There is no workaround.

• CSCth46888

Symptoms: When the ARP entry is refreshed due to timeout or use of the clear arp command, the router sends ARP request for cached MAC address. However, the request message does not use virtual MAC for Source (Sender) MAC.

Conditions: The symptom is observed when the router is VRRP master and VRRP IP is configured the same as the interface IP.

Workaround: There is no workaround.

• CSCth47888

Symptoms: In a Hot-Standby psuedowire redundancy setup, traffic is forwarded on the Standby psuedowire instead of the Active psuedowire which is in up/up state.

Conditions: This symptom is seen in a Cisco 7600 router that is running Cisco IOS Release 15.0(1)S with hot psuedowire redundancy configuration.

Workaround: There is no workaround.

• CSCth55383

Symptoms: When entering the show tech command on RP, the line card with DFC may display SWITCH_BUS_IDLE message.

Conditions: This symptom occurs when entering the show tech command on RP.

Workaround: There is no workaround.

• CSCth55689

Symptoms: If you do a clear xconnect before the primary VC has come up, the system erroneously brings up PW on the backup VC.

Conditions: The symptom is observed if you do a clear xconnect before PW establishment.

Workaround: Use the following command:

clear xconnect peer peer-ip vcid vcid-id.

• CSCth60232

Symptoms: The port-channel interface may flap when adding or removing a VLAN from the trunk on a port-channel interface when one or more interfaces are in a state other than P or D.

Conditions: This symptom is observed only when the port-channel interface has interfaces in states other than P or D.

Workaround: Shut down the non-P members and make the vlan changes.

• CSCth62854

Symptoms: A Cisco router crashes with traceback ospfv3_intfc_ipsec_cmd.

Conditions: This symptom is observed when the interface is configured with ospfv3, null authentication/encryption, and non-null encryption/authentication.

Workaround: Remove the ospfv3 area command, then remove the null authentication/encryption.

• CSCth65072

Symptom: A memory leak occurs in the big buffer pool while using the service reflect feature.

Conditions: This symptom is observed when the service reflection feature is enabled. A packet is generated from service reflection and is blocked by an ACL on the outgoing interface. This will cause the buffer leak.

Workaround: Remove the ACL on the outgoing interface or permit the packets generated from service reflect on the ACL.

• CSCth66246

Symptoms: A configuration sync issue and stand-by reload occur when configuring and removing the ip slb sticky feature.

Conditions: This symptom is observed in the Cisco IOS Release 15.1(00.10)S image.

Workaround: There is no workaround.

• CSCth69469

Symptoms: ICMP filtering is not working in an SACL configuration.

Conditions: This symptom is observed when ACE is configured with icmp options.

Workaround: There is no workaround.

• CSCth69504

Symptoms: A Cisco 7600 series router may experience a small buffer leak in the small buffer pool on SP.

Conditions: The symptom is observed on a Cisco 7600 series router that is running Cisco IOS Release 12.2(33)SRD configured with IGMP snooping.

Workaround: Disable IGMP snooping either globally or per VLAN.

• CSCth69364

Cisco IOS Software contains a memory leak vulnerability in the Data-Link Switching (DLSw) feature that could result in a device reload when processing crafted IP Protocol 91 packets.

Cisco has released free software updates that address this vulnerability.

This advisory is posted at http://www.cisco.com/warp/public/707/cisco-sa-20110928-dlsw.shtml.

• CSCth69588

Symptoms: Disposition traffic does not flow across ES+ cards. Imposition traffic for sceompls works fine.

Conditions: This symptom is observed with ES+ card mainly on switchover, reprovisioning of circuits, etc.

Workaround: Enter the clear xconnect all command on the device and wait for reprovision. Once all circuits are up, traffic will flow across fine.

• CSCth71095

Symptoms: DHCP binding table is not completely synced to standby RP.

Conditions: This symptom occurs when box is acting as DHCP Relay and ISG is configured.

Workaround: Use unnumbered multiservice interface instead of numbered one.

• CSCth71349

Symptoms: Some SSS sessions are staying in “attempting” state for a while when using ISG Static Session Creation.

Conditions: The symptom is observed when using ISG Static Session Creation.

Workaround: Stop incoming traffic from subscribers and wait until the sessions recover, then re-apply the traffic.

• CSCth72290

Symptoms: Traffic over PPPoMPLS drops continuously on a SIP400 with microcode reload.

Conditions: The symptom is observed when PPPoMPLS is configured over any interface on a channelized SPA on SIP400 and following a microcode reload of the linecard.

Workaround: Reload the SPA.

Further Problem Description: When PPPoMPLS is configured on an channelized interface the encapsulation for that interface on the linecard is PPP where as the SPA encapsulation is HDLC as the SPA just tunnels the packets through. After the microcode reload, the linecard queries the SPA for a connection ID for the channelized interface with encapsulation as PPP but as the SPA encapsulation set is HDLC the SPA does not give any connection ID to the SIP400 and hence the traffic starts to drop after a reload.

• CSCth72565

Symptoms: The leachability of the PE2 router’s loopback is lost from PE1 after an interface flap in the core. The LSP toward PE2 “breaks” due to data plane programing error (wrong labels).

Conditions: The symptom is observed with MPLS with the presence of ECMP. The PE1 has two uplinks to core routers. In a steady state there is no ECMP between PE1 and PE2. When a link is lost in the P-core (link flap or shut/no shut) there is ECMP between PE1 and PE2. After the link flap between the two P routers in the core, PE1 is losing connectivity to PE2.

Workaround 1: Use the clear ip route on the affected IP address.

Workaround 2: Avoid ECMP by altering link cost.

• CSCth72765

Symptoms: Configuring “mls qos protocol hsrp police <rate>” does not enable policing of HSRPv2 packets.

Conditions: The symptom is observed on a Cisco 7600 series router when “mls qos protocol hsrp police <rate>” is configured.

Workaround: There is no workaround.

• CSCth73173

Symptoms: ASR may crash if a QoS policy applied using CoA through Service-Template is more than 256 characters in length.

Conditions: This symptom is observed when a QoS Policy string length exceeds 256 characters.

Workaround: Ensure that the QoS policy string length is less than 256 characters.

• CSCth75354

Symptoms: There is an intermittent problem when a SPAN source is set to be a VLAN. The destination in the SPAN session does not receive the data.

Conditions: The symptom is observed on a Cisco 7600 series router with an ES+20 card that is running Cisco IOS Release is 12.2(33)SRE0.

Workaround: Reload the module.

• CSCth77531

Symptoms: A Cisco ASR 1000 Series Aggregation Services router with hundreds of IPv4 and IPv6 BGP neighbors shows high CPU utilization in the BGP-related processes for several hours (more than 2.5).

Conditions: The symptom is observed with Cisco IOS Release 12.2(33)XNF. The BGP task process uses the most CPU; also, the number of routemap-cache entries should be very high.

Workaround: Use “no bgp route-map-cache.” This will not cache the route-map cache results, and the issue will not be observed.

• CSCth78148

Symptom: Cannot attach to 8XCHT1/E1 SPA console.

Conditions: This issue is observed after configuring SSO.

Workaround: Reload the SPA.

• CSCth79336

Symptoms: Ingress QoS policy, when applied on an EVC configured on port- channel, is getting configured for only one of the NPs, instead of all of the NPs that have member-links belonging to the port-channel. If traffic enters the port-channel via multiple member-links, provided the member-links belong to different NPs, ingress QoS will be applied on only one of the member-links.

Conditions: This symptom is observed when EVC is configured on port-channel.

Workaround: There is no workaround.

• CSCth80166

Symptoms: EVCs belonging to the same group are mapped to different member links, and the packets are forwarded over different member links.

Conditions: This symptom is observed when dynamic changes such as adding an EVC to a group or removing it are made; all the group members are not mapped to the same member link.

Workaround: Perform a shut/no shut on the device.

• CSCth81950

Symptoms: A memory leak occurs in ES+ cards.

Conditions: This symptom is observed on a Cisco 7600 with the cac enabled.

Workaround: There is no workaround.

• CSCth82486

Symptoms: A SIP600 crashes.

Conditions: The symptom is observed following an OIR of the active supervisor.

Workaround: There is no workaround.

• CSCth84995

Symptoms: Router may reload when performing an ISSU upgrade or downgrade.

Conditions: This symptom occurs when performing an ISSU upgrade or downgrade.

Workaround: There is no workaround.

• CSCth85294

Symptoms: A PIM neighborship not established with the remote PE and RP for the MVRFs.

Conditions: This symptom is observed with traffic, after removal and restoration of mvrfs. Traffic does not flow properly since the PIM neighborship is not established with the remote PE and RP for those MVRFs.

Workaround: There is no workaround.

• CSCth85449

Symptoms: Interfaces on SIP-400 may experience.9% packet loss in the LLQ when priority percent is configured.

Conditions: This symptom is observed on a Cisco 7600 series router running Cisco IOS Release 12.2(33)SRE when the low latency queue traffic is allowed to grow past its configured percentage. This condition only occurs when the interface has greater than 85% utilization.

Workaround: There is no workaround.

• CSCth86402

Symptoms: When flapping a WAN interface, the PIM tunnel disappears.

Conditions: This happens when flapping a WAN interface after a few hours of working.

Workaround: Disable multicast routing, then enable it again.

• CSCth87132

Symptoms: Diagnostic tests may fail on an ES+ linecard with the following message:

Conditions: This symptom is observed when 802.1 TAP MIB is used to tap based on an if_index belonging to an “access” interface.

Workaround: IP-TAP mib may be used instead of 802.1 TAP MIB.

• CSCth87195

Symptoms: Flexwan ATM interface goes down.

Conditions: This symptom is observed while configuring “mac-address” or “atm bridge-enable.”

Workaround: Perform a shut/no shut on the interface.

• CSCth87587

Symptoms: Spurious memory access or a crash is seen upon entering or modifying a prefix-list.

Conditions: The primary way to see this issue is to have “neighbor <neighbor address> prefix-list out” configured under “address-family nsap” under “router bgp” when configuring/modifying a prefix-list.

Workaround: There is no workaround.

Further Problem Description: The issue is only specific to certain scenarios when prefix-lists are used in conjunction with “nsap address-family”.

• CSCth90001

Symptoms: Packets egressing interfaces of ES+ line cards are not received on the other side of L2 link when SVI plus switchport configuration is used. It is random and does not occur on every ES+ line card.

Conditions: This symptom is observed when ES+ line card is egress line card and SVI plus L2 switchport configuration is used. When this issue is seen the CFI bit in vlan tag header for such packets is set by X40g egress-intf causing the peer router to drop such packets.

Workaround: Use L3 802.1q subinterface configuration.

• CSCth92171

Symptoms: The serial interface stays down longer if a switchover is done while flapping the multilink interface from the far end.

Conditions: This symptom is observed when switching over to the standby while flapping the multilink interface from the far end.

Workaround: Shut the flapping links, then perform the switchover.

• CSCth93218

Symptoms: The error message “%OER_BR-4-WARNING: No sequence available” displays on PfR BR.

Conditions: The symptom is observed in a scale setup with many PfR application prefixes and when PfR optimizes the application prefixes.

Workaround: There is no workaround.

• CSCth94814

Symptoms: Crash is seen in static route component.

Conditions: The symptom is observed when changing IVRF on a virtual-template when there are about 100 active sessions.

Workaround: There is no workaround.

• CSCth94827

Symptoms: IDBINDEX_SYNC-STDBY tracebacks are seen when unconfiguring ima-group on a SONET-ACR controller.

Conditions: This symptom is observed on a standby supervisor when unconfiguring and configuring ima-group on a SONET-ACR controller.

Workaround: There is no workaround.

• CSCth99237

Symptoms: LNS does not respond to an LCP echo reply when waiting for a response from the AAA server. As a result, the peer may close the session.

Conditions: The symptom is observed under the following conditions:

1. If the client starts to send LCP echo requests during the PPP Authentication phase

2. If the primary AAA server is unreachable and/or the authentication response is otherwise delayed.

Workaround: There is no workaround.

• CSCth99786

Symptoms: A Cisco ASR1000 acting as an ISG crashes.

Conditions: This symptom is observed when subscriber policy debugging is enabled; for example:

Workaround: Disable subscriber policy debugging.

• CSCti01036

Symptoms: A Cisco ASR1000 series router crashes on the Radius Process.

Conditions: This symptom is observed on a Cisco ASR 1000 series router with Radius AAA services enabled. When the Radius server sends attributes with no information (empty VSA strings), it produces an unexpected reload on the router.

Workaround: Prevent the AAA server from sending empty VSA strings.

• CSCti01971

Symptoms: The active router crashes during a switchover in a scaled BFD IPv6 setup.

Conditions: The router is configured with a larger number of IPv6 routes with BFD sessions configured. (The test was done with 500 BFD IPv6 sessions.)

Workaround: There is no workaround.

• CSCti04678

Symptom: A Cisco router crashes with redzone corruption.

Conditions: This symptom is observed when a router is configured for any subscribers and someone tries to execute some of the show CLI while clearing the sessions.

Workaround: There is no workaround.

• CSCti04754

Symptoms: PPPoE sessions are stuck at attempting state forever.

Conditions: This symptom is seen when sessions are triggered during SSO time, which get stuck at attempting state.

Workaround: Clear attempting state sessions by the clear command from box.

• CSCti05663

Symptoms: A DHCP ACK which is sent out in response to a renew gets dropped at relay.

Conditions: The symptom is observed in the case of an numbered relay.

Workaround: There is no workaround.

• CSCti08115

Symptoms: The removal of a port-channel interface associated with mpls ldp advertise-labels interface Port-channelN can cause a “config sync” error upon an SSO.

Conditions: The symptom is observed after doing an SSO following the removal of the port-channel interface.

Workaround: Before the SSO, remove the offending advertise-labels command when removing the port-channel command with:

no interface Port-channelN no mpls ldp advertise-labels interface Port-channelN

• CSCti08336

Symptoms: PfR moves traffic-class back and forth between primary and fallback links the when PfR Link group feature is used.

Conditions: The symptoms are most likely to occur when there is one exit in the primary link-group and utilization is one of the criteria. The issue can also occur when there are two links in the primary. A traffic-class is moved from the primary link to the fallback link when the primary link is OOP. After the move, the primary link and the fallback link are “IN” policy. At that time, PfR moves the traffic-class back to primary causing the primary link to go “Out” of policy.

Workaround: There is no workaround.

• CSCti10518

Symptoms: Under very rare circumstances, EIGRP could exhibit a memory leak of NDB structures in the RIB.

Conditions: If redistribution is occurring into EIGRP and the route ownership is changing in the middle of the redistribution process, EIGRP may leak the NDB in process.

Workaround: There is no workaround.

• CSCti13286

Symptoms: Putting this configuration on a router:

and doing a reload causes the “no validate-update-source” statement to disappear from the VRF configuration (the one under the global RIP configuration remains). This affects functionality, preventing the RIP updates in VRF from being accepted.

Conditions: The symptom has been observed using Cisco IOS Release 15.0(1)M3 and Release 15.1(2)T.

Workaround: There is no workaround.

• CSCti14290

Symptoms: A Cisco 7600 series router acting as the PE router in an MPLS network may stop forwarding traffic for certain IP prefixes within a VRF. This symptom may occur after a router reload, upgrade or crash due to corrupted hardware-forwarding information on the ingress module for the VPN label of the affected IP prefix.

The problem can be identified by comparing the output of the following commands:

1. Determine the BGP VPN Label for the prefix

show ip bgp vpnv4 all vrf vrf name prefix

2. Determine the hardware forwarding for the prefix on the Supervisor

show mls cef mpls label label detail

show mls cef adjacency entry entry id

3. attach to the ingress module and use the same commands as step 2 and compare the values, if the destination mac address is not the same there is hardware forwarding corruption. Note: The adjacency index will be a different number on the module dfc.

remote login module module number

Conditions: The Cisco 7600 must have a distributed forwarding card installed on the ingress module and be configured as an MPLS PE router. The problem is only observed after a router reload, upgrade or crash.

Workaround: Reloading the ingress module will resolve the hardware forwarding corruption on the module:

hw-module module <module number> reset

• CSCti22091

Symptoms: Traceback will occur after a period of use and when the show oer master command is used a few times. The traceback is always followed by the message “learning writing data”. The traceback causes the OER system to disable. Manually reenabling PfR will not work. A reboot is required.

Conditions: The symptom is observed when PfR is configured with the following conditions:

1. list > application > filter > prefix-list

2. Learn > traffic-class: keys

3. Learn > traffic-class: filter > ACL

Workaround: There is no workaround.

• CSCti24657

Symptoms: Fabric CRC errors, fabric sync errors, fabric minor/major errors, Signal Integrity issues over the fabric channel, as well as Blackwater queue stall. Sample error messages:

Conditions: These symptoms are observed on ES20 linecards with Cisco 7600 series routers.

Workaround: There is no workaround.

• CSCti25339

Symptoms: Cisco IOS device may experience a device reload when receiving certain SNMP packets from an authenticated user. Successful exploitation may cause the affected device to reload. This vulnerability could be exploited repeatedly to cause an extended DoS condition.

Conditions: Cisco IOS device configured for SNMP. Authentication is required to cause a device reload.

Workaround: There is no workaround.

PSIRT Evaluation: The Cisco PSIRT has assigned this bug the following CVSS version 2 score. The Base and Temporal CVSS scores as of the time of evaluation are 6.8/5.6:

https://intellishield.cisco.com/security/alertmanager/cvssCalculator.do?dispatch=1&version=2&vector=AV:N/AC:L/Au:S/C:N/I:N/A:C/E:F/RL:OF/RC:C

CVE ID CVE-2010-3050 has been assigned to document this issue.

Additional information on Cisco’s security vulnerability policy can be found at the following URL:

http://www.cisco.com/en/US/products/products_security_vulnerability_policy.html

• CSCti26540

Symptoms: A memory leak in both SSS Manager and AAA Attribute list can be created when multiple services are downloaded and one of the services fails.

Conditions: This symptom is observed when a failure in the finishing application of all services leads to a memory leak in the cleanup code.

Workaround: Proper service profiles should avoid the memory leak.

• CSCti26768

Symptoms: A Cisco 6500 router configured with a PKI trustpoint could reload due to bus error.

Conditions: This symptom is observed when following the steps for a trustpoint configuration: removal, reconfiguration, authentication, enrollment.

Workaround: There is no workaround.

• CSCti32641

Symptoms: A Cisco ASR 1004 (RP2) router is not able to establish an LDP session to a 3rd-party device and receives an Error Notification (0x07) Bad TLV Length message from that device.

Conditions: This symptom is observed on a Cisco ASR 1004 with Cisco IOS Release 15.0(1)S when LDP ICCP capability TLV (0x405) is supported on the router.

Workaround: There is no workaround.

Further Problem Description: It seems that Cisco ASR 1004 routers send to the peer a malformed ICCP capability TLV (0x405).

• CSCti34396

Symptoms: The router distributes an unreachable nexthop for a VPNv4 or VPNv6 address as an MVPN tunnel endpoint.

Conditions: The symptom is seen when “next-hop-unchanged allpaths” is configured for an external neighbor of the VPNv4 or VPNv6 tunnel endpoint, and the previous hop is an unreachable.

Workaround 1: Configure a route-map to rewrite routes so that the tunnel endpoint is an address reachable from both inside the VRF and outside of it. For example, to rewrite statically configured routes so that the nexthop is set to a visible address, you would configure:

Alternate Workaround: Instead of configuring static routes with a next-hop, specify an interface name.

For example, if you had: ip route x.x.x.x 255.255.255.0 y.y.y.y And y.y.y.y was on the other end of the interface serial2/0, you would replace this configuration with: ip route x.x.x.x 255.255.255.0 interface serial2/0

Further Problem Description: You may also need to override the standard behavior of next-hop-unchanged allpaths in a generic manner with a single standard configuration which could be applied to all the routers. In order to solve this problem, the configuration “set ip next-hop self” is added to route-maps.

When used in conjunction with the newly added configuration:

router bgp <asn> address-family vpnv4 unicast bgp route-map priority

The “set ip next-hop self” will override “next-hop unchanged allpaths” for the routes which match the route-map where it is configured, allowing the selective setting of the next-hop.

• CSCti34462

Symptoms: After FPD upgrade, a shut on the active shows no shut on the standby.

Conditions: The symptom is observed after an FPD upgrade.

Workaround: Perform a no shut then shut the interface on the active to sync it properly.

• CSCti34627

Symptoms: This bug is caused by a problem with the fix for CSCth18982. When a neighbor in multiple topologies is enabled, the open sent for the base topology clears the nonbase topology session for the same neighbor.

Conditions: A GR-enabled neighbor exists in different topologies, one of them being the base topology.

Workaround: Disable GR.

• CSCti35170

Symptoms: With REP over EVC configured and a high volume of traffic, REP could flap due to REP Hellos getting dropped.

Conditions: This condition is observed only when a high volume of traffic (mostly priority traffic) is sent on the interface.

Workaround: There is no workaround.

• CSCti37533

Symptoms: A Cisco router may crash.

Conditions: This symptom is observed with a half-inserted standby card that generates an internal stall; if this stall continues for more than 30 seconds, a crash occurs.

Workaround: Remove or fully insert the standby card.

• CSCti41910

Symptoms: Changes to the spanning-tree mst instance configuration are not synced to a standby SP. Hence, after a switchover, the new Active will have old MST instance configuration on SP.

Conditions: The symptom is observed after completing the following steps:

1. Configure MST instance configuration

(config)#spanning-tree mst configuration (config-mst)#instance 1 vlan 102

2. Do HA switchover

3. “Show spanning-tree mst configuration” on active SP will not show ’instance 1’ in configuration.

Workaround: Reload the standby.

Further Problems Description: Bulk sync for MST configuration works fine. Only the incremental configuration sync is broken.

• CSCti43395

Symptoms: Tracebacks are seen during DHCP message exchange. Crash may also be seen with the tracebacks.

Conditions: This symptom is seen when DHCP relay agent is configured with “ip dhcp relay information option vpn” and clients with duplicate MAC address are coming in at the same time.

Workaround: Unconfigure “ip dhcp relay information option vpn”. Or, disallow clients with duplicate MAC.

• CSCti45732

Symptoms: Upon a reload, a Cisco 7600 series router configured as VTP server may lose some VLANs from its VLAN database.

Conditions: The VLANs lost do not have any access ports in the device. All other switches in the network should be in VTP transparent mode. This issue is seen on a Cisco 7600 series router that is running Cisco IOS 12.2 (33)SRE1 and SRE2 Releases.

Workaround: Configure the Cisco 7600 as VTP transparent instead of VTP server.

• CSCti47158

Symptoms: When configuring the T1 channels on the SPA-2CHT3-CE-ATM/CHOC3-ATM on a SIP-400 on a Cisco 7600 series router, the T1 remains in a “Down” state.

Conditions: The symptom is observed under the following conditions:

– Seen on a Cisco 7600 series router with SIP-400 and SPA-2CHT3-CE-ATM/CHOC3- ATM.

– Seen when the configuration is done using copy and paste, or using a script.

Workaround: Perform a shut/no shut on the CEM.

• CSCti47550

Symptoms: With a scaled L3 ACL on EVC on ES+ linecards, some of the ACEs do not work, while others work as normal.

Conditions: The symptom is observed when the linecard or router is reloaded with the ACL configuration present.

Workaround: Remove and add ACL on the EVC.

• CSCti49508

Symptoms: The command show platform isg session all displays stale entries on a Cisco 7600 series router for ISG sessions that are not on the router.

Conditions: This symptom is observed under the following conditions:

1. A number of port channel subinterfaces are configured with ISG

2. ISG sessions are active on the subinterfaces

3. The main port channel is removed without removing the sessions or ISG configuration from the individual port channel subinterfaces, using the “no interface port-channel <>” command

Workaround: There is no workaround.

To avoid this symptom, Delete the session/ISG configuration from the individual port channel subinterface, then delete the port channel.

• CSCti50419

Symptoms: For PPPoMPLS/HDLCoMPLS pseudowires, after you perform the switchover, traffic loss is seen and CE interfaces stay down.

Conditions: The symptom is observed on performing an SSO switchover with PPPoMPLS and HDLCoMPLS pseudowires. The control word gets programmed incorrectly on the linecard leading to traffic loss.

Workaround: Unprovision and provision the pseudowire.

Alternate Workaround: Perform a SPA OIR.

• CSCti50607

Symptoms: A Cisco 7200 SRE1 router drops GRE packet size 36-45.

Conditions: The symptom is observed on a Cisco 7200 series router with SRE1 code.

Workaround: There is no workaround.

• CSCti51145

Symptoms: After a reload of one router, some or all of the BGP address families do not come up. The output of show ip bgp all summary will show the address family in NoNeg or idle state, and it will remain in that state.

Conditions: In order to see this problem, ALL of the following conditions must be met:

– The non-reloading device must have a “neighbor x.x.x.x transport connection- mode passive” configuration, or there must be an ip access list or packet filter which permits connections initiated by the reloading device, but not by the non-reloading device. In Cisco IOS, such ip access-lists typically use the keyword ’established’ or “eq bgp”

– It must be configured with a BGP hold time which is less than the time required for the neighbor x.x.x.x to reload

– When the neighbor x.x.x.x reloads, no keepalives or updates must be sent on the stale session during the interval between when the interface comes up and when the neighbor x.x.x.x exchanges BGP open messages

– Both peers must be multisession capable

– “transport multi-session” must not be configured on either device, or enabled by default on either device

– “graceful restart” must not be configured.

Workarounds:

1. Remove the configuration “neighbor x.x.x.x transport connection-mode passive” or edit the corresponding filter or ip access list to permit the active TCP opens in both directions

2. Configure “neighbor x.x.x.x transport multi-session” on either the device or its neighbor

3. Configure a very short keepalive interval (such as one second) on the non-reloading device using the neighbor x.x.x.x timers 1 holdtime command

4. Configure graceful restart using the command neighbor x.x.x.x ha- mode graceful-restart

5. 5. If the issue occurs, use the clear ip bgp * command to cause all sessions stuck in the NoNeg state to restart. You can also use clear ip bgp x.x.x.x addressFamily to bring up individual stuck sessions without resetting everything else.

Further Problem Description: This is a day one problem in the Cisco IOS multisession implementation which impacts single-session capable peers. CSCsv29530 fixes a similar problem for some (but not all) situations where “neighbor x.x.x.x transport single-session” is configured and NSF is not configured.

The effect of this fix is as follows: when the neighbor is in single-session mode, AND the router sees an OPEN message for a neighbor which is in the ESTABLISHED state, then the router will send a CEASE notification on the new session and close it (per section 6.8 of RFC 4271). Additionally, it will send a keepalive on the ESTABLISHED session. The keepalive is not required, but will cause the established session to be torn down if appropriate.

Note that the fix does not solve the problem when interacting with Cisco IOS 12.2(33)SB-based releases if the 12.2(33)SB router is the one not reloading.

• CSCti53664

Symptom: CoPP hardware counters not incrementing when sh policy-map control-plane command is run for traffic coming on ES20+ cards.

Conditions: This symptom is observed when CoPP is configured and traffic is coming in on ES20+, which is destined to switch the ip address.

Workaround: Move the l3 interface from the switch for the traffic coming in on ES20+ line cards.

• CSCti56980

Symptoms: Applying a service-policy under an interface or subinterface on an ES+ card in a Cisco 7600 series router may fail with the following error:

random-detect aggregate is not supported in output direction for this interface Configuration failed!

Conditions: The symptom only occurs when a SIP400 is being replaced by an ES+ card on which the QoS configuration will be applied.

Workaround: Reload the router with the ES+ card installed.

• CSCti61949

Symptoms: Unexpected reload with a “SYS-2-CHUNKBADMAGIC: Bad magic number in chunk header” and “chunk name is BGP (3) update” messages.

Conditions: The symptom is observed when receiving BGP updates from a speaker for a multicast-enabled VRF.

Workaround: Disable multicast routing on VRFs participating in BGP or reduce the number of extended communities used as route-target export.

• CSCti62125

Symptoms: When a Cisco 67XX card is inserted in slot 2 of a Cisco 7606-S chassis, then other cards (such as ES+, ES, and SIP) in the other slot face fabric CRC errors. The ES+ in the other slot gets hung and leads to a crash.

Conditions: The symptom is observed when a Cisco 67XX card is inserted in slot 2 of a Cisco 7606-S chassis.

Workaround: There is no workaround.

• CSCti65716

Symptoms: The access interface connecting to the client is on global routing domain. If a service logon profile on a VRF is downloaded to the client, the client could potentially stay on a VRF even when a service logoff is performed later. The client traffic has to return to global domain when a service logoff is performed.

Conditions: This symptom is see when access interface is on global routing domain. Service logon is on a VRF.

Workaround: There is no workaround.

• CSCti66076

Symptoms: A standby HSRP router could be unknown after reloading the ES20 module that configured HSRP.

Condition: This symptom is observed under the following conditions:

– HSRP version 1 is the protocol that must be used

– Use HSRP with sub-interfaces on ES20 module

– Reload the ES20 module

Workaround: Change to HSRPv2, which is not exposed to the issue.

Alternate Workarounds:

1. Reconfigure HSRP on all subinterfaces

2. Configure multicast or igmp configuration on the interface where HSRP is configured (like ip pim sparse-mode).

• CSCti67429

Symptoms: A REP segment configured on 7600-ES+20G3CXL interfaces on a Cisco 7600 series router that is running Cisco IOS Release 15.0(1)S is not recovering as expected upon link failure recovery of the edge port configured on the 7600. A traffic storm triggered by ISIS protocol configured between 7600 and the MWR 2941s in the REP ring is occurring when the failed REP edge port becomes operational again.

Conditions: The symptom is observed with a REP ring including two Cisco 7600 series routers equipped 7600-ES+20G3CXL and running Cisco IOS Release 15.0(1) S configured with ISIS and MPLS LDP. The problem is also present in Cisco IOS Release 12.2(33)SRE1.

Workaround: Configure static routes between the 7600 routers and the MWR 2941s instead of ISIS.

• CSCti67447

Symptoms: During an SSO, an 8 to 12 second packet drop may occur on EoMPLS VCs.

Conditions: The symptom is observed under the following conditions:

1. EoMPLS port-based or VLAN-based configuration; VC between PE1 and PE2

2. Enable MPLS LDP GR.

Workaround: There is no workaround.

• CSCti68153

Symptoms: QinQ packet header is modified, breaking QinQ.

Conditions: This occurs when you use an ES+ card on a Cisco 7600 series router where the pseudowire encapsulation begins.

Workaround: Use double VLAN tagging before the packet gets to the ES+ card so that when it pops the first VLAN tag it will modify the second dummy tag and not the actual payload.

Further Problem Description: This issue is not seen on ES20 card.

• CSCti68721

Symptoms: The output of show performance monitor history interval <all | given #> will appear to have an extra column part way through the output.

Conditions: This symptom is observed sporadically while traffic is running on a performance monitor policy at the time when a user initiates the CLI show command.

Workaround: If the symptom occurs, repeat the command.

• CSCti69008

Symptoms: When dampening is configured for many VRFs, doing full vpnv4 radix tree walk and the proposed fix improves convergence by doing subtree walk based on VRF/RD.

Conditions: Dampening configuration changes for VRFs.

Workaround: There is no workaround.

• CSCti72498

Symptom: A crash occurs on a device acting as DHCP Server.

Conditions: This symptom is observed when a requested IP address option is present in DHCP requests.

Workaround: Disable the DHCP ping check with the help of CLI “ip dhcp ping packets 0."

• CSCti74736

Symptoms: A traffic drop might appear on a GREoMPLS tunnel after an SSO switchover in an egress direction. If an ingress interface is located on a SIP400 series linecard, the following error message will be continuously printed:

Conditions: The presence of “mls mpls tunnel-recir” is required for the GREoMPLS feature to work. After the second SSO switchover since bootup, the command will be inactive and the feature broken. The issue is applicable to Cisco IOS Release 12.2(33)SRE2, but not to Release 12.2(33)SRE1.

Workaround: Reload the router.

• CSCti74962

Symptoms: “%PM-SP-4-PORT_BOUNCED: bounced by Consistency Check IDBS UP” message seen on A3-1 new active router after linecard OIR followed by an SSO switchover.

Conditions: This symptom will occur only with a linecard OIR followed by an SSO switchover.

Workaround: There is no workaround.

• CSCti77521

Symptoms: Policy-map is not attached to a DLFIoATM interface after a SPA OIR.

Conditions: The symptom is observed upon performing a SPA OIR. The issue is seen with ATM SPA on a SIP400.

Workaround: Perform a shut/no shut of the ATM interface.

• CSCti81444

Symptoms: Traffic does not flow in egress direction over VPLS PW on router reload.

Conditions: The symptom is observed after a router reload. POE bits for the imposition interface are not getting programmed on the egress linecard.

Workaround: There is no workaround.

• CSCti83705

Symptoms: IPv4 unicast traffic not forwarded out of a Cisco 7600 series router’s GREoMPLS in VRF tunnel.

Conditions: The symptom is observed with an IPv6 Address Family (AF) configured under VRF. If the IPv6 AF is in the startup configuration then the feature is broken straight after boot up. If the IPv6 AF is configured after boot up, then feature gets broken after this configuration.

Workaround: Remove IPv6 AF from the tunnel’s VRF.

• CSCti83737

Symptom: SIP-600 will crash with a software-forced crash:

Conditions: This symptom is observed on SIP-600.

Workaround: There is no workaround.

• CSCti85446

Symptoms: A nexthop static route is not added to RIB even though the nexthop IP address is reachable.

Conditions: The symptom is observed with the following conditions:

1. Configure a nexthop static route with permanent keyword

2. Make the nexthop IP address unreachable (e.g.: by shutting the corresponding interface)

3. Change the configuration in such a way that nexthop is reachable

4. Configure a new static route through the same nexthop IP address used in step 1.

Workaround: Delete all the static routes through the affected nexthop and add them back.

• CSCti86169

Symptoms: A device that is acting as a DHCP relay or server crashes.

Conditions: This symptom is observed when the “no service dhcp” command is configured.

Workaround: There is no workaround.

• CSCti88062

Symptoms: Traffic stops flowing through ports configured with REP over EVC BD when an ES20 linecard is replaced by an ES+ in the same slot.

Conditions: The symptom is observed on a router running MST, having an ES20 card configured with EVC BD which is replaced by an ES+ in the same slot with an EVC BD configuration. MST puts the BD VLAN in a disabled state and the traffic on that VLAN stops flowing.

Workaround: Reload the router.

• CSCti97759

Symptoms: IPSG configuration with DHCP snooping entry configuration causes the RP to crash.

Conditions: This is seen when DHCP static entry is configured.

Workaround: There is no workaround.

• CSCtj00039

Symptoms: Some prefixes are in PE router EIGRP topology although those routes are not being passed to the CE router.

Conditions: The symptom is observed when EIGRP is configured as a routing protocol between PE and CE routers.

Workaround: Clear the route on the PE router using clear ip route vrf xxx x.x.x.x.

• CSCtj01623

Symptoms: REP topology stays incomplete after manual preemption. When the issue occurs, REP preemption will not take effect.

Conditions: The symptom can be observed for EVC or switchport.

Workaround: There is no workaround.

• CSCtj05198

Symptoms: When there are two EIGRP router processes (router eigrp 7 and router eigrp 80), PfR is unable to find the parent route. The problem occurs only if one of the processes has the parent route and other one does not. As a result, probe and route control fail.

Conditions: This symptom is observed when there are two EIGRP router processes.

Workaround: Use one EIGRP process. There is no workaround if two processes are used.

• CSCtj05591

Symptoms: Memory corruption and SP crash seen.

Conditions: The symptom is observed when creating 600 subinterfaces as OIF for Mroute entries.

Workaround: There is no workaround.

• CSCtj08533

Symptoms: QoS classification fails on egress PE if the route is learnt via BGP.

Conditions: The symptom is observed when there are redundant paths to the CPE.

Workaround: Use only one path between PE and CPE.

• CSCtj15265

Symptoms: The event trace buffer gets wrapped around resulting in met cc log not being available.

Conditions: This symptom is seen when there is a lot of OIF churn.

Workaround: There is no workaround.

• CSCtj15805

Symptoms: Keepalive functionality not working. An ICMP echo reply coming back from a client is ignored by ISG.

Conditions: The symptom is observed when a VRF mapping service is used.

Workaround: There is no workaround.

• CSCtj17545

Symptoms: Immediately after a switchover, the restarting speaker sends TCP- FIN to the receiving speaker, when receiving speaker tries to establish (Active open). It can cause packet drops after a switchover.

Conditions: The symptom can occur when a lot of BGP peers are established on different interfaces.

Workaround: When the receiving speaker is configured to accept passive connections, the issue will not be observed:

template peer-session ce-v4 transport connection-mode passive

• CSCtj17561

Symptoms: Description for T1 broken in Prowler/Chopper SDH > C-11 mode. This might lead to sync issues while switching over.

Conditions: The symptom is observed in SDH > C-11 mode.

Workaround: There is no workaround.

• CSCtj18622

Symptoms: The “disperser” feature is not working when it is pushed through AV_Pair by a RADIUS server; either Cisco ACS or a third party vendor’s server.

Conditions: The symptom is observed with non-multilink PPP users.

Workaround: There is no workaround.

• CSCtj24453

Symptoms: The following traceback is observed when clear ip bgp * is done:

Conditions: It is rarely observed, when clear ip bgp * is done with lot of routes and route-map-cache entries.

The clear ip bgp * command is not a very common operation in production network.

Workaround: Use no bgp route-map-cache. This will not cache the route-map cache results and the issue will not be observed.

• CSCtj25243

Symptoms: If non-LLQ or parent (logical) is rate-limited and oversubscribed, this can cause some policer drops in the LLQ queue, if LLQ exceeds the bandwidth allocated to it.

Conditions: The symptom is observed if non-LLQ or parent (logical) is rate- limited and oversubscribed and if LLQ exceeds the bandwidth allocated to it.

Workaround: There is no workaround.

Further Problem Description: This issue is caused by CSCth85449. That caveat was intended to detect congestion on the physical interface and police LLQ traffic if it exceeds the configured bandwidth and the physical link is congested.

• CSCtj28696

Symptoms: Session QoS will not get applied after an OIR of the linecard.

Conditions: The symptom is observed with sessions (with QoS) on a port- channel subinterface.

Workaround: Clear session and bring up again.

• CSCtj28747

Symptoms: Route control of prefix and application are out-of-order thereby making application control ineffective. As a result, an “Exit Mismatch” message will be logged on the MC and the application will be uncontrolled for a few seconds after it is controlled.

Conditions: The symptom is observed only if PIRO control is used where prefixes are also controlled using dynamic PBR. PIRO control is used when the routing protocol is not BGP, STATIC, or EIGRP, or when two BRs have different routing protocol, i.e.: one has BGP and the other has EIGRP.

Workaround: There is no workaround.

• CSCtj30155

Cisco IOS Software is affected by two vulnerabilities that cause a Cisco IOS device to reload when processing IP version 6 (IPv6) packets over a Multiprotocol Label Switching (MPLS) domain. These vulnerabilities are:

– Crafted IPv6 Packet May Cause MPLS-Configured Device to Reload

– ICMPv6 Packet May Cause MPLS-Configured Device to Reload

Cisco has released free software updates that address these vulnerabilities.

Workarounds that mitigate these vulnerabilities are available.

This advisory is posted at http://www.cisco.com/warp/public/707/cisco-sa-20110928-ipv6mpls.shtml.

• CSCtj32574

Symptoms: Deleting the redistribute command into EIGRP does not get synchronized to the standby. For example:

router eigrp 1 redistribute connected no redistribute connected

The no redistribute connected command is not being backed up to the standby.

Conditions: The symptom is observed with any redistribute-related commands.

Workaround: There is no workaround.

• CSCtj35573

Symptoms: When an interface is configured as an access interface, back-to-back ping will fail.

Conditions: The ping failure is seen only for access interfaces intermittently. This issue is observed with the SRE2 image with SUP720 and ES+ card, in a situation when the ping packet coming from source has the BPDU bit set.

Workaround: There is no workaround.

• CSCtj36521

Symptoms: IPv4 MFIB stays enabled on interfaces even when IPv4 CEF is disabled. The output of the show ip mfib interface command shows the interface as configured and available, when it should be disabled.

Conditions: The symptom is observed only if IPv6 CEF is enabled at the same time.

Workaround: Make sure IPv6 CEF is always disabled when running only IPv4 multicast. There is no workaround if running a mixed IPv4/IPv6 environment.

• CSCtj38606

Symptoms: The following error message is seen:

%SYSTEM_CONTROLLER-3-MISTRAL_RESET: System Controller is reset:Normal Operation continues

The show ibc exec command reports increments of the following counter:

Hazard Illegal packet length = 7580

Conditions: The symptom is observed on a Cisco 7600 series router.

Workaround: There is no workaround.

• CSCtj41215

Symptoms: On an ES+, a service instance configuration is rejected with following error:

Service instance configuration Failed. Service-Policy has already been configured on this interface

Conditions: The symptom is observed when an ES+ is inserted in the same slot where an ES20 was previously present.

Workaround: Unconfigure service-policy from the interface and then create a service instance.

• CSCtj44237

Symptom: High CPU observed in RP.

Conditions: The symptom is observed with MVPN configurations.

Workaround: There is no workaround.

• CSCtj45571

Symptoms: If OAM VC state reaches to “AIS/RDI” after PVC is flapping, then OAM Loopback status gets stuck in “OAM failed” state. Loopback cell is not generated until shut/no shut is performed on the subinterface.

Conditions: The symptom is observed when the OAM VC state reaches “AIS/RDI”.

Workaround: Perform a shut/no shut on the subinterface.

• CSCtj46297

Symptoms: Ping fails when performing a shut/no shut on the outgoing interface in an FRR setup.

Conditions: The symptom is observed in an FRR setup when performing a shut/no shut on the outgoing interface.

Workaround: Perform a shut/no shut on the tunnel interface.

• CSCtj50072

Symptoms: High CPU interrupt level caused by IPv4 unicast or multicast traffic received via GREoIP or GREoMPLS tunnel if rate is high. If ingress interface is tunnel and egress is tunnel (MDT included) as well, then outer IP ToS of egress packet will be reset to 0x0.

Conditions: The symptom is observed after a reload (under 10% probability), GRE tunnel must be in VRF:

To confirm hit:

Workaround: Reload the router.

• CSCtj52865

Symptoms: Unable to utilize 16 queues per lowq port.

Conditions: If you remove any QoS policy on subtargets of lowq port, because of stale lowq count on the show platform lowq command, we will not be able to use maximum number of queues per lowq port.

Workaround: Only reloading the router resolves the issue.

• CSCtj53299

Symptoms: Met corruption issue is observed.

Conditions: This symptom occurs during OIF churn.

Workaround: Use the clear ip mroute command for the problematic entry.

• CSCtj58943

Symptoms: Standby RP reloads due to line by line sync failure for encapsulation dot1q 1381 command:

Config Sync: Line-by-Line sync verifying failure on command: encap dot1Q 1381 due to parser return error

Conditions: Symptom occurs when issuing a configuration command under a sub-interface mode.

Workaround: There is no workaround.

• CSCtj59254

Symptoms: Data to default MDT switchover fails in highly scaled scenarios.

Conditions: The symptom is observed during a default to data MDT switchover.

Workaround: There is no workaround.

• CSCtj67133

Symptoms: Standby gets stuck at “in progress to standby cold-config” state in ISSU. The following message will be seen:

Conditions: The symptom is observed when doing an ISSU with Cisco IOS Release 12.2(33)SRE or Release 15.0(1)S.

Workaround: There is no workaround.

• CSCtj70271

Symptoms: Non-local replications are programmed as local replications in the MET3 (i.e.: if the replications are on slot 3 DFC module, then the supervisor is programmed with the sublsps of slot 3 as local replications). This causes a waste of TCAM resources and can cause traffic outage.

Conditions: The symptom is observed with LSM/MLDP configurations.

Workaround: Use this command: clear ip mroute source group.

• CSCtj72148

Symptoms: A Cisco 7600 router might face an SP crash upon first reload after upgrade from Cisco IOS Release 12.2(33)SRC5 to Release 12.2(33)SRE2. After successive reloads, the system functionality is restored.

Conditions: This symptom is observed when upgrading from Cisco IOS Release 12.2(33)SRC5 to Release 12.2(33)SRE2.

Workaround: There is no workaround.

• CSCtj74611

Symptoms: Active supervisor in the Cisco 7600 series router reloads.

Conditions: The symptom is observed after a linecard is powered off due to keepalive failures. Possible sequence of syslog messages:

Workaround: There is no workaround.

• CSCtj77004

Symptoms: Archive log configuration size impacts CPU utilization during PPPoE establishment. Also, only some configuration lines from the virtual-template are copied to archive (some lines missing).

Conditions: The symptom is observed when “archive log config” is configured.

Workaround: There is no workaround.

• CSCtj79085

Symptoms: MFIB entries struck in NP HW_ERR MET-FULL:5, NP HW_ERR MET-ALLOC:6

Conditions: This symptom is observed during slot 7 reload, UUT-CE1 interface Flap and UUT reload with traffic.

Workaround: There is no workaround.

• CSCtj79349

Symptoms: After swapping GE-T SFP with GE-SX, if port is at 100m, it will retain the same old speed.

Conditions: Issue is seen only after swapping with GE-T which was at 100m with any other SFP on an ES+.

Workaround: There is no workaround.

• CSCtj86514

Symptoms: An SNMP walk on Cisco AAL5 MIB may not return information for all PVCs configured on the device.

Conditions: An SNMP walk query on the Cisco AAL5 MIB may fail to return information of bundled PVCs that are in down state. Information about PVCs in UP state is returned correctly.

Workaround: To get information of bundled PVCs in down state, you need to poll with more specific OIDs. Instead of doing an snmpwalk on “1.3.6.1.4.1.9.9.66.1.1.1.1.3”, do an snmpget on “1.3.6.1.4.1.9.9.66.1.1.1.1.3.<IfIndex>.<VPI>.<VCI>”.

• CSCtj87180

Symptoms: A LAC router running VPDN may crash when it receives an invalid redirect from the peer with a CDN error message of “SSS Manager Disconnected Session”.

Conditions: The symptom is observed when the LAC router receives an incorrect “Error code(9): Try another directed and Optional msg: SSS Manager disconnected session <<<< INVALID” from the multihop peer.

Workaround: There is no workaround.

• CSCtj89941

Symptoms: IOSd crash when using the command clear crypto session on an EzVPN client.

Conditions: Testbed setup:

1. RP2+ESP20 worked as the EzVPN simulator, which is configured with over 1000 clients. Then simulator is connected to Cisco ASR 1004-RP1/ESP10 (UUT) with DVTI configured

2. Use IXIA to generate 1Gbps traffic

3. Wait until all the SAs have been established and traffic is stable

4. Use CLI clear crypto session on EzVPN simulator.

Workaround: There is no workaround.

• CSCtj92563

Symptoms: Interfaces take a longer time to come up.

Conditions: The symptom is observed with an ES+ linecard.

Workaround: There is no workaround.

• CSCtj94358

Symptoms: SIP400 will pass the traffic through a previously configured VLAN on reconfiguring the bridge-domain command.

Conditions: This symptom is seen with the egress interface that is a SIP400 with MPB configured.

Workaround: Remove the “bridge-domain” configuration and then add the new “bridge-domain”.

• CSCtj94835

Symptoms: Spurious memory access and tracebacks are seen on router reload.

Conditions: The symptom is observed when the router is reloaded.

Workaround: There is no workaround.

• CSCtj95032

Symptoms: PIM packets are dropped at SIP400. As a result PIM neighborship is not formed between the CEs.

Conditions: This symptom is seen when the egress interface is on SIP400 with bridging configured on it.

Workaround: There is no workaround.

• CSCtj96489

Symptoms: In a CISCO 7600 router, a freshly provisioned interface, or an interface which has been administratively no shut, belonging to non-default VRF, may fail to forward traffic.

Conditions: This is a race condition and hence timing sensitive.

Workaround: Another interface shut/no shut may help restore service.

• CSCtj96915

Symptoms: LNS rounter hangs up at interrupt level and goes into an infinite loop.

Conditions: Unknown at this moment, See Further Problem Description below.

Workaround: There is no workaround; power-cycle to remove the symptom.

Further Problem Description: This is a hypothesis based on analysis of the data provided for the failures experienced by the customer, together with an extensive code review. The issue can happen during L2TP session creation and removal, specifically where a session removal/addition is prevented from completion by an interrupt that is raised. We believe this is a timing issue; while this is a rare event, the probability of it occurring increases with load and number of sessions.

• CSCtj97360

Symptoms: Punted datapaths are multicast flows GREoIP->DefaultMDT & GREoMPLS->Default MDT.

Conditions: This symptom is observed with device bootup with ipv4-only VRF. After bootup IPv6 was enabled for VRF, which has triggered the problem.

Workaround: Do not have IPv6 AF and the mcast configs in the same VRF.

• CSCtj99415

Symptoms: Traffic is not dropped when the packet size is more than the egress interface MTU.

Conditions: This symptom is observed when the egress interface is on SIP400. When the outgoing interface is on ES20, the packets are dropped at RP with an error message.

Workaround: There is no workaround

• CSCtk00976

Symptoms: File descriptor reaches the maximum threshold limit. You will be unable to save the configuration or do any file system related operation as file descriptors are exhausted. You will get “File table overflow” error.

Conditions: The symptom is observed when running the dir/recursive <> command periodically using the ANA tool.

Workaround: Do not run dir/recursive <> command if leaks are detected. Also, if it is running through ANA server polling, disable it.

• CSCtk02155

Symptom: Attachment to the CHOC3 SPA console fails after seeing VC configuration command failures.

Conditions: This symptom is seen with CHOC3 SPA on SIP200 or SIP400.

Workaround: Reset the line card.

Further Problem Description: The periodic process resyncs the IPC between the host and CHOC3 SPA. As this is not happening, we are not able to attach to the SPA console.

• CSCtk02661

Symptoms: Bundles stop forwarding any traffic.

Conditions: The symptom is observed when you move the SPA to a different bay on a SIP-400 and apply configurations on the new bay.

Workaround: Reload spa on both ends.

Alternate workaround: Unconfigure multilink before moving the SPA out.

• CSCtk02666

Symptoms: During a graceful restart event, the peer undergoes reconfiguration. This may result in stale labels on the RRP.

Conditions: The symptom is observed with GR + SSO + peer reprovisioning.

Workaround: Perform a clear xconnect or flap the local VC.

• CSCtk05652

Symptoms: UDLD used end-to-end across an AToM link causes the CE link on one side to be put into err-disabled state.

This is the topology:

We are seeing UDLD errdisabling the port on SW2, though the link is not unidirectional.

Conditions: This symptom is observed on Cisco IOS Release 12.2(33)SRE2.

Workaround: Running Cisco IOS Release 12.2(333)SRD5 instead of Release SRE2 fixes the problem.

• CSCtk07369

Symptoms: The buginf statement “draco2_fastsend: PAK_BUF_ON_OBL processing vlan” appears on the console.

Conditions: This is displayed in certain cases, such as multicast replication.

Workaround: There is no workaround.

• CSCtk07632

Symptoms: Even with the filter option, traffic on a different VLAN on trunk port is getting spanned.

Conditions: The symptom is observed when the filter vlan specified is not configured on the box.

Workaround: Configure the vlan on the box, then configure it as SPAN filter vlan.

• CSCtk12608

Symptoms: Route watch fails to notify client when a RIB resolution loop changes. This causes unresolved routes to stay in the routing table.

Conditions: The symptoms are observed using Cisco IOS Release 15.0(1)M, Release 15.1 (2)T and Release 15.1(01)S and with the following configurations:

Workaround: Use static routes tied to a specific interfaces instead of using “floating static routes”.

• CSCtk13364

Symptoms: Traffic is blackholed over EVC bridge domain interfaces on the port.

Conditions: The symptom is observed when a subinterface is deleted and an EVC with the same encapsulation dot1q is created and configured with a bridge domain. The traffic over all the other EVCs on the interface is blackholed.

Workaround: After the configuration, perform a shut/no shut on the interface.

• CSCtk19108

Symptoms: MVPN traffic failing.

Conditions: The symptom is observed after an SSO switchover.

Workaround: There is no workaround.

• CSCtk31515

Symptoms: Router or linecards crash upon removing VLAN interfaces that are in the OIF list.

Conditions: The symptom is observed with a series of VLAN interfaces in the access and with the hosts joining groups. Configured is “ssm-mapping”. Access facing linecards can be DFC or CFC.

Workaround: There is no workaround.

• CSCtk31615

Symptoms: Standby crashes upon an ISSU upgrade from Cisco IOS Release 12.2 (33)SRD to Release 12.2(33)SRE.

Conditions: The symptom is observed with an ISSU between Cisco IOS Release 12.2(33)SRD and Release 12.2(33)SRE.

Workaround: There is no workaround.

• CSCtk33682

Symptoms: Storm control stops working.

Conditions: The symptom is observed after a shut/no shut of the interface on an ES-20.

Workaround: Remove/add the storm control command on the interface.

• CSCtk33784

Symptoms: After ISSU from SRE1 to SRE3, %CONST_MFIB_LC-SP-6-MET_MCAST_ALLOC_FAILURE is displayed continuously for a particular group.

Conditions: This symptom is observed after configuring 10 groups and 32 OIFs.

Workaround: There is no workaround.

• CSCtk34026

Symptoms: Adding, deleting and re-adding an access subinterface may sometimes cause loss of data path.

Conditions: The symptom is observed when the configuration sequence involves an add-delete-add sequence.

Workaround: Create access subinterfaces from scratch.

• CSCtk36059

Symptoms: Active SRE does a silent reload while undergoing an ISSU from Cisco IOS Release 12.2(33)SRD to Release 12.2(33)SRE.

Conditions: The symptom is observed with scaled configurations.

Workaround: There is no workaround.

• CSCtk36064

Symptoms: The cos value is not copied to exp when adding a label in l3vpn/vpls scenarios.

Conditions: This symptom is observed On A Cisco 7600 router, ES+ LC, QoS policy-map with set cos applied on switchport interface in ingress.

Workaround: There is no workaround.

• CSCtk36090

Symptoms: Router crash at draco2_inband_dma_pak after a router reload with the following SRE image:

Conditions: The symptom is observed following a router reload.

Workaround: There is no workaround.

• CSCtk36377

Symptoms: VRF ping fails for some of the VRFs after deleting and adding MVRFs.

Conditions: This symptom is seen when adding and deleting MVRFs using a script.

Workaround: Delete VRF and add it back.

• CSCtk37068

Symptoms: Policing is not happening.

Conditions: This symptom is observed when copp is enabled.

Workaround: There is no workaround.

• CSCtk39301

Symptoms: Tracebacks such as the following can appear on the RP:

Conditions: The symptom is observed if there are more than eight or 10 ECMP paths for any prefix (i.e.: when there is a load balanced object in the forwarding OCE chain).

Workaround: Reduce the number of paths and do a clear ip route to re-initiate hardware programming.

• CSCtk47739

Symptoms: IPSec tunnels can be established even though the certificate of remote peer has been revoked and the current router’s CRL has been updated.

Conditions: The symptom is observed on a Cisco 7600 series router running Cisco IOS Release 12.2(33)SRD4.

• CSCtk47891

Symptoms: Traffic might be blackholed on lc reset if FRR is in use.

Conditions: This symptom is observed with an FRR configuration in ACTIVE state when the LC is reset.

Workaround: There is no workaround.

• CSCtk47960

Symptoms: Large CLNP packets may be dropped when forwarded over SIP- 200/Flexwan2 module. Header Syntax errors may be recorded on receiving host.

Remote side will generate the following:

Conditions: This symptom is seen on Cisco 7600 switch with SIP-200 line card that is running Cisco IOS 12.2(33)SRD3 and later releases.

Issue is seen when packets larger than 911 bytes are sent (Payload and Header).

Workaround: If CLNS is only used for ISIS neighborships “no isis hello padding” can be configured to establish ISIS neighborship. For the LSP packets, configure lns-mtu 903 under router isis on the Cisco 7600 to make this work.

• CSCtk53463

Symptoms: When configuring "shape average <cir value> <bc value>" currently across all platforms, <bc value> is limited by 4ms * <cir value>. 4ms here represents the minimum interval time for bursts. ES+ LC, however can support an interval value that is faster (smaller) than 4ms. This has been an expected behavior with the exception of ES+ LC.

Conditions: This symptom is observed on all platforms.

Workaround: There is no workaround.

• CSCtk53763

Symptoms: Traffic for some of the SubLSPs is not flowing with P2MP TE or MLDP.

Conditions: The symptom is observed with LSM and MLDP configurations with multiple SubLSPs.

Workaround: Use the following command: clear ip mroute *.

• CSCtk54318

Symptoms: VC creation fails on disabling and re-enabling the card for SIP-400 with 4XT3E3 SPA with below messages on console:

Condition: This issue is seen when the below commands are executed on a T3 serial interface of the SPA 4XT3E3 configured as DTE with frame relay encapsulation:

no card type t3 slot bay card type t3 slot bay

Then unconfigure and reconfigure frame relay enacapsulation.

Workaround: Reload the SPA.

• CSCtk55382

Symptoms: A SPA-OC192POS-VSR or SPA-OC192POS-XFP may fail boot diagnostic test.

Conditions: The symptom is observed when Control Plane Policing (CoPP) is configured on the system. The diagnostic test that fails is the “TestACLPermit” test displayed in “show diagnostic result”. The output of “show module” will indicate a “Minor error” on the subslot.

Workaround: Before a system reload or module reset, disable the CoPP feature. After the module is booted, CoPP can be enabled again.

• CSCtk55573

Symptoms: On SIP-600 and SIP-400 linecards, disposition drops are observed for software EoMPLS or VPLS traffic.

Conditions: The symptom is observed with SIP-400/600 linecards.

Workaround: Perform a shut/no shut on the VLAN where xconnect is configured.

• CSCtk57049

Symptoms: Traffic is not sent over data mdt even though the vrf selects the data mdt for encapsulation. After access interface flap on encap PE in MVPN setup, the traffic is not sent over data mdt even though the vrf selects the data mdt for encap.

Conditions: This symptom is observed after access interface flap on encap PE in an MVPN setup.

Workaround: There is no workaround.

• CSCtk62680

Symptoms: A Cisco 7600 series router may experience a sudden failure upon attempting to upgrade the software from Cisco IOS Release 12.2(33)SRD3 to Release 12.2 (33)SRE3 using the ISSU procedure.

Conditions: The symptom is observed with a multicast configuration with multiple hosts distributed across subinterfaces, main interfaces, EVC service instances, and port-channel interfaces on CFC and DFC cards.

Workaround: None for ISSU. Upgrade using a reload.

• CSCtk64020

Symptoms: A Cisco 7600 may crash.

Conditions: This symptom is observed when the clear ip subsriber command is entered.

Workaround: There is no workaround.

• CSCtk69810

Symptoms: The rw index is set to “invalid” in the catch-all entry. Because of this, the PIM neighbor on MDT is not up and traffic is not flowing.

Conditions: This symptom is observed after ISSU from SRE1 image to SRE3.

Workaround: Delete and add MDT under vrf.

• CSCtk76190

Symptoms: The RSP/SUP fails to switchover automatically when the “TestSPRPInbandPing” fails for more than 10 instances.

Conditions: The symptom is observed when the “TestSPRPInbandPing” fails for more than 10 instances.

Workaround: There is no workaround.

• CSCtk95904

Symptoms: RP crash at oce_to_sw_obj_type.

Conditions: The symptom is observed with a route flap with 2K P2MP_TE sLSPs.

Workaround: There is no workaround.

• CSCtk98030

Symptoms: After replacing an ES20 linecard with an ES+ linecard or vice versa in the same slot, some service groups reject new members to join if the old linecard had ethernet service instances in these groups. Similarly, a named EVC rejects new ethernet service instances if it had association with the old linecard. The named EVC cannot be deleted, complaining that it still has service instances.

Conditions: The symptom is observed if an ES20 linecard has been replaced with an ES+ linecard or vice versa in the same slot. The old linecard had ethernet service instance members in some service groups and/or named EVCs. The old associations between ethernet service instances and service groups or named EVCs are not cleaned up properly, blocking new association to these groups and EVCs.

Workaround: Configure new service groups and named EVCs with same configuration as the problematic ones. Abandon the use of the old groups and EVCs. Assign ethernet service instances from the new linecard to the new groups and EVCs.

• CSCtl05684

Symptoms: Xauth user information remains in “show crypto session summary” output.

Conditions: This symptom is observed when running EzVPN and if Xauth is performed by different username during P1 rekey.

Workaround: Use save-password feature (without interactive Xauth mode) to avoid sending the different username and password during P1 rekey.

• CSCtl05926

Symptoms: Packets of size exceeding MTU are dropped with the following error messages.

Conditions: This symptom is observed when the outgoing interface is on SIP400

Workaround: There is no workaround.

• CSCtl07955

Symptoms: BFD neighbor goes down and never comes up again

Conditions: This is symptom is observed when we power down an unrelated LC by "no power enable module X."

Workaround: There is no workaround.

• CSCtl10395

Symptoms: CoPP stops dropping packets in hardware on a Cisco 7600 after double switchover.

Conditions: This symptom is observed on a Cisco 7600 platform with CoPP configured on the router. SSO (HA Switchover) is done twice.

Workaround: Remove and reconfigure the CoPP policy.

• CSCtl18652

Symptoms: After replacing an ES20 with an ES+ linecard on the same slot, or vice versa, adding ethernet service instance members from the new linecard to an existing service group that was associated with the old linecard may cause a reload of the standby RP in SSO mode. This is due to stale configuration on the standby RP.

Conditions: An ES20 linecard has been replaced by a different type of linecard or vice versa, on the same slot. New members are assigned to a service group that had members from the old linecard. There is a standby RP in SSO mode.

Workaround: Create a new service group with the same configuration as the existing group and assign new members to the new group. Abandon the use of the old group.

• CSCtl19347

Symptoms: On configuring more bundles, LC crashes with sip-400.

Conditions: This is symptom is observed upon trying to copy the dlfi configs from disk to running config to make bundle up the LC crashes.

Workaround: There is no workaround.

• CSCtl22871

Symptoms: Cos value (applied from setcos policy) is not copied to EXP while adding label in one of the VPLS cases (VPLS cfgd on EVC BD Vlan).

Conditions: This symptom is observed On ES+, QoS policy-map with set cos applied on EVC BD with VPLS configured on BD Vlan.

Workaround: There is no workaround.

• CSCtl41921

Symptoms: Traffic duplication occurs.

Conditions: This symptom is observed with a boot-up scale of 2000 slsps.

Workaround: Perform a shut/no shut on the tunnel.

• CSCtl43925

Symptoms: Having P2P GRE tunnels in vrf on the access side causes multicast traffic for the vrf to be dropped.

Conditions: This symptom is observed after removal of tje GRE header and encapsulation changes from Tunnel vlan to QoS vlan. The next entry to be hit has incoming vlan as VPN QoS vlan. Here CR=1 is necessary. However when tunnels are suddenly brought up in vrf, CR=0 gets programmed and packets get bridged and dropped.

Workaround: There is no workaround.

• CSCtl46903

Symptoms: The VLAN mapping/translation feature does not work.

Conditions: This symptom is observed on ES+ when the port is configured as an L2 switchport.

Workaround: The feature works properly only if it is configured under EVC framework or L2 switchport in the LAN cards.

• CSCtl82922

Symptoms: Fast memory leak occurs on the Standby Switch Processor (SP) of Supervisor in the “mfib-const-lc” process. Once this process depletes memory, the system will generate “MALLOC” errors for any other process requesting memory at that time. Eventually, the standby SP will crash and system operation will recover.

To identify observe “Holding” number over time in Standby SP, can grow with speed of 60kB/s:

Conditions: This symptom is observed with multicast stream timeout & restart in a mVPN environment. Stream S,G entry might not be installed in HW, and following MFIB Platform flags error might be seen for this stream along with memory leak:

Workaround: There is no workaround.