Switch software can be corrupted during an upgrade by downloading the incorrect file to the switch, and by deleting the image file. In all of these cases, the switch does not pass the power-on self-test (POST), and there is no connectivity.

The default configuration for the device allows an end user with physical access to the device to recover from a lost password by interrupting the boot process during power-on and by entering a new password. These recovery procedures require that you have physical access to the device.

Note

On these devices, a system administrator can disable some of the functionality of this feature by allowing an end user to reset a password only by agreeing to return to the default configuration. If you are an end user trying to reset a password when password recovery has been disabled, a status message reminds you to return to the default configuration during the recovery process.

Note	You cannot recover encryption password key, when Cisco WLC configuration is copied from one Cisco WLC to another (in case of an RMA).

A Power over Ethernet (PoE) switch port automatically supplies power to one of these connected devices if the switch detects that there is no power on the circuit:

• a Cisco pre-standard powered device (such as a Cisco IP Phone or a Cisco Aironet Access Point)

• an IEEE 802.3af-compliant powered device

• an IEEE 802.3at-compliant powered device

A powered device can receive redundant power when it is connected to a PoE switch port and to an AC power source. The device does not receive redundant power when it is only connected to the PoE port.

After the switch detects a powered device, the switch determines the device power requirements and then grants or denies power to the device. The switch can also detect the real-time power consumption of the device by monitoring and policing the power usage.

For more information, see the "Configuring PoE" chapter in the Interface and Hardware Component Configuration Guide (Catalyst 3650 Switches) .

The Device supports IP ping, which you can use to test connectivity to remote hosts. Ping sends an echo request packet to an address and waits for a reply. Ping returns one of these responses:

• Normal response—The normal response (hostname is alive) occurs in 1 to 10 seconds, depending on network traffic.

• Destination does not respond—If the host does not respond, a no-answer message is returned.

• Unknown host—If the host does not exist, an unknown host message is returned.

• Destination unreachable—If the default gateway cannot reach the specified network, a destination-unreachable message is returned.

• Network or host unreachable—If there is no entry in the route table for the host or network, a network or host unreachable message is returned.

The Layer 2 traceroute feature allows the switch to identify the physical path that a packet takes from a source device to a destination device. Layer 2 traceroute supports only unicast source and destination MAC addresses. Traceroute finds the path by using the MAC address tables of the Device in the path. When the Device detects a device in the path that does not support Layer 2 traceroute, the Device continues to send Layer 2 trace queries and lets them time out.

The Device can only identify the path from the source device to the destination device. It cannot identify the path that a packet takes from source host to the source device or from the destination device to the destination host.

You can use IP traceroute to identify the path that packets take through the network on a hop-by-hop basis. The command output displays all network layer (Layer 3) devices, such as routers, that the traffic passes through on the way to the destination.

Your Device can participate as the source or destination of the traceroute privileged EXEC command and might or might not appear as a hop in the traceroute command output. If the Device is the destination of the traceroute, it is displayed as the final destination in the traceroute output. Intermediate Device do not show up in the traceroute output if they are only bridging the packet from one port to another within the same VLAN. However, if the intermediate Device is a multilayer Device that is routing a particular packet, this Device shows up as a hop in the traceroute output.

The traceroute privileged EXEC command uses the Time To Live (TTL) field in the IP header to cause routers and servers to generate specific return messages. Traceroute starts by sending a User Datagram Protocol (UDP) datagram to the destination host with the TTL field set to 1. If a router finds a TTL value of 1 or 0, it drops the datagram and sends an Internet Control Message Protocol (ICMP) time-to-live-exceeded message to the sender. Traceroute finds the address of the first hop by examining the source address field of the ICMP time-to-live-exceeded message.

To identify the next hop, traceroute sends a UDP packet with a TTL value of 2. The first router decrements the TTL field by 1 and sends the datagram to the next router. The second router sees a TTL value of 1, discards the datagram, and returns the time-to-live-exceeded message to the source. This process continues until the TTL is incremented to a value large enough for the datagram to reach the destination host (or until the maximum TTL is reached).

To learn when a datagram reaches its destination, traceroute sets the UDP destination port number in the datagram to a very large value that the destination host is unlikely to be using. When a host receives a datagram destined to itself containing a destination port number that is unused locally, it sends an ICMP port-unreachable error to the source. Because all errors except port-unreachable errors come from intermediate hops, the receipt of a port-unreachable error means that this message was sent by the destination port.

Caution

Because debugging output is assigned high priority in the CPU process, it can render the system unusable. For this reason, use debug commands only to troubleshoot specific problems or during troubleshooting sessions with Cisco technical support staff. It is best to use debug commands during periods of lower network traffic and fewer users. Debugging during these periods decreases the likelihood that increased debug command processing overhead will affect system use.

All debug commands are entered in privileged EXEC mode, and most debug commands take no arguments.

In a switch stack, when you enable debugging, it is enabled only on the active switch. To enable debugging on a stack member, you must start a session from the active switch by using the session switch-number privileged EXEC command. Then, enter the debug command at the command-line prompt of the stack member.

System reports or crashinfo files save information that helps Cisco technical support representatives to debug problems that caused the Cisco IOS image to fail (crash). It is necessary to quickly and reliably collect critical crash information with high fidelity and integrity. Further, it is necessary to collect this information and bundle it in a way that it can be associated or identified with a specific crash occurrence.

System reports are generated in these situations:

• In case of a switch failure—A system report is generated on the member that failed; reports are not generated on other members in the stack.

• In case of a switchover—System reports are generated only on high availability (HA) member switches. reports are not generated for non-HA members.

The system does not generate reports in case of a reload.

During a process crash, the following is collected locally from the switch:

1. Full process core

2. Tracelogs

3. IOS syslogs (not guaranteed in case of non-active crashes)

4. System process information

5. Bootup logs

6. Reload logs

7. Certain types of /proc information

This information is stored in separate files which are then archived and compressed into one bundle. This makes it convenient to get a crash snapshot in one place, and can be then moved off the box for analysis. This report is generated before the switch goes down to rommon/bootloader.

Except for the full core and tracelogs, everything else is a text file.

Crashinfo Files

By default the system report file will be generated and saved into the /crashinfo directory. Ifit cannot be saved to the crashinfo partition for lack of space, then it will be saved to the /flash directory.

To display the files, enter the dir crashinfo: command. The following is sample output of a crashinfo directory:

Switch#dir crashinfo:
Directory of crashinfo:/
46553 drwx 1024 Jun 29 2015 14:52:09 +00:00 ap_crash
12 -rw- 0 Jan 1 1970 00:00:11 +00:00 koops.dat
11 -rw- 0 Mar 22 2013 07:50:30 +00:00 deleted_crash_files
13 -rwx 594269 Mar 22 2013 07:50:30 +00:00 crashinfo_platform_mgr_20130322-075017-UTC
14 -rw- 44 Sep 9 2015 09:28:47 +00:00 last_crashinfo
15 -rw- 355 Sep 9 2015 09:29:31 +00:00 last_systemreport_log
16 -rw- 105753 Mar 22 2013 07:50:47 +00:00 system-report_1_20130322-075017-UTC.gz
17 -rw- 39 Sep 9 2015 09:29:31 +00:00 last_systemreport
18 -rwx 585996 Mar 22 2013 08:01:58 +00:00 crashinfo_platform_mgr_20130322-080144-UTC
19 -rw- 105065 Mar 22 2013 08:02:15 +00:00 system-report_1_20130322-080144-UTC.gz
20 -rwx 3426209 Sep 9 2015 06:49:12 +00:00 crashinfo_iosd_20150909-064754-UTC
21 -rwx 9540376 Sep 9 2015 06:49:13 +00:00 fullcore_iosd_20150909-064754-UTC
22 -rw- 469476 Sep 9 2015 06:49:56 +00:00 system-report_1_20150909-064754-UTC.gz
23 -rwx 3425350 Sep 9 2015 09:28:47 +00:00 crashinfo_iosd_20150909-092728-UTC
24 -rwx 9535535 Sep 9 2015 09:28:47 +00:00 fullcore_iosd_20150909-092728-UTC
25 -rw- 459709 Sep 9 2015 09:29:28 +00:00 system-report_1_20150909-092728-UTC.gz
26 -rw- 0 Sep 22 2015 11:11:33 +00:00 tracelogs.J8C

50601 drwx 10240 Oct 28 2015 22:42:50 +00:00 tracelogs

248354816 bytes total (204800000 bytes free)

System reports are located in the crashinfo directory in the following format:

system-report_[switch number]_[date]-[timestamp]-UTC.gz

After a switch crashes, check for a system report file. The name of the most recently generated system report file is stored in the last_systemreport file under the crashinfo directory. The system report and crashinfo files assist TAC while troubleshooting the issue.

The system report generated can be further copied using TFTP, HTTP and few other options.

Switch#copy crashinfo: ?
crashinfo:     Copy to crashinfo: file system 
flash:         Copy to flash: file system
ftp:           Copy to ftp: file system 
http:          Copy to http: file system 
https:         Copy to https: file system 
null:          Copy to null: file system
nvram:         Copy to nvram: file system
rcp:           Copy to rcp: file system
running-config Update (merge with) current system configuration
scp:           Copy to scp: file system
startup-config Copy to startup configuration
syslog:        Copy to syslog: file system
system:        Copy to system: file system
tftp:          Copy to tftp: file system
tmpsys:        Copy to tmpsys: file system

The general syntax for copying onto TFTP server is as follows:

Switch#copy crashinfo: tftp:
Source filename [system-report_1_20150909-092728-UTC.gz]?
Address or name of remote host []? 1.1.1.1
Destination filename [system-report_1_20150909-092728-UTC.gz]?

The tracelogs from all members in the stack can be collected by issuing a trace archive command. This command provides time period options. The command syntax is as follows:

Switch#request platform software trace archive ?
last     Archive trace files of last x days
target   Location and name for the archive file

The tracelogs stored in crashinfo: or flash: directory from within the last 3650 days can be collected.

Switch# request platform software trace archive last ?
<1-3650> Number of days (1-3650)
Switch#request platform software trace archive last 3650 days target ?
crashinfo: Archive file name and location
flash:     Archive file name and location

Note	It is important to clear the system reports or trace archives from flash or crashinfo directory once they are copied out, in order to have space available for tracelogs and other purposes.

You can use the onboard failure logging (OBFL) feature to collect information about the device. The information includes uptime, temperature, and voltage information and helps Cisco technical support representatives to troubleshoot device problems. We recommend that you keep OBFL enabled and do not erase the data stored in the flash memory.

By default, OBFL is enabled. It collects information about the device and small form-factor pluggable (SFP) modules. The device stores this information in the flash memory:

• CLI commands—Record of the OBFL CLI commands that are entered on a standalone device or a switch stack member.

• Message—Record of the hardware-related system messages generated by a standalone device or a switch stack member.

• Power over Ethernet (PoE)—Record of the power consumption of PoE ports on a standalone device or a switch stack member.

• Temperature—Temperature of a standalone deicev or a switch stack member.

• Uptime data—Time when a standalone device starts, the reason the device restarts, and the length of time the device has been running since it last restarted.

• Voltage—System voltages of a standalone device.

You should manually set the system clock or configure it by using Network Time Protocol (NTP).

When the device is running, you can retrieve the OBFL data by using the show logging onboard privileged EXEC commands. If the device fails, contact your Cisco technical support representative to find out how to retrieve the data.

When an OBFL-enabled device is restarted, there is a 10-minute delay before logging of new data begins.

By default, the feature is disabled. When more than one of the fans fails in a field-replaceable unit (FRU) or in a power supply, the Device does not shut down, and this error message appears:

Multiple fan(FRU/PS) failure detected. System may get overheated. Change fan quickly.

The Device might overheat and shut down.

To enable the fan failures feature, enter the system env fan-fail-action shut privileged EXEC command. If more than one fan in the Device fails, the Device automatically shuts down, and this error message appears:

Faulty (FRU/PS) fans detected, shutting down system! 

After the first fan shuts down, if the Device detects a second fan failure, the Device waits for 20 seconds before it shuts down.

To restart the Device, it must be power cycled.

Excessive CPU utilization might result in these symptoms, but the symptoms might also result from other causes:

• Spanning tree topology changes

• EtherChannel links brought down due to loss of communication

• Failure to respond to management requests (ICMP ping, SNMP timeouts, slow Telnet or SSH sessions)

• UDLD flapping

• IP SLAs failures because of SLAs responses beyond an acceptable threshold

• DHCP or IEEE 802.1x failures if the switch does not forward or respond to requests

The high usage of memory in systems can be due to various reasons:

• Memory leak in some processes.

• A combination of certain configurations with maximum limits.

The backend monitoring process checks memory usage every 10 seconds, and generates 2 levels of warning (syslog) messages, if the usage exceeds the configured threshold.

When a system reaches 99+% of memory usage, the out-of-memory (OOM) Killer of the Linux kernel starts killing processes that are running on the system. If these processes are noncritical, they are restarted by the Process Manager; if these are critical, the Process Manager does a clean up, and reloads the box. In these circumstances, no core file for the killed processes are generated; however, a system report is generated when the box is reloaded.

The IEEE 802.3ab autonegotiation protocol manages the Device settings for speed (10 Mb/s, 100 Mb/s, and 1000 Mb/s, excluding SFP module ports) and duplex (half or full). There are situations when this protocol can incorrectly align these settings, reducing performance. A mismatch occurs under these circumstances:

• A manually set speed or duplex parameter is different from the manually set speed or duplex parameter on the connected port.

• A port is set to autonegotiate, and the connected port is set to full duplex with no autonegotiation.

To maximize Device performance and ensure a link, follow one of these guidelines when changing the settings for duplex and speed:

• Let both ports autonegotiate both speed and duplex.

• Manually set the speed and duplex parameters for the ports on both ends of the connection.

Note	If a remote device does not autonegotiate, configure the duplex settings on the two ports to match. The speed parameter can adjust itself even if the connected port does not autonegotiate.

Cisco small form-factor pluggable (SFP) modules have a serial EEPROM that contains the module serial number, the vendor name and ID, a unique security code, and cyclic redundancy check (CRC). When an SFP module is inserted in the Device, the Device software reads the EEPROM to verify the serial number, vendor name and vendor ID, and recompute the security code and CRC. If the serial number, the vendor name or vendor ID, the security code, or CRC is invalid, the software generates a security error message and places the interface in an error-disabled state.

Note	The security error message references the GBIC_SECURITY facility. The Device supports SFP modules and does not support GBIC modules. Although the error message text refers to GBIC interfaces and modules, the security messages actually refer to the SFP modules and module interfaces.

If you are using a non-Cisco SFP module, remove the SFP module from the Device, and replace it with a Cisco module. After inserting a Cisco SFP module, use the errdisable recovery cause gbic-invalid global configuration command to verify the port status, and enter a time interval for recovering from the error-disabled state. After the elapsed interval, the Device brings the interface out of the error-disabled state and retries the operation. For more information about the errdisable recovery command, see the command reference for this release.

If the module is identified as a Cisco SFP module, but the system is unable to read vendor-data information to verify its accuracy, an SFP module error message is generated. In this case, you should remove and reinsert the SFP module. If it continues to fail, the SFP module might be defective.

The Device monitors the temperature conditions and uses the temperature information to control the fans.

Use the show env temperature status privileged EXEC command to display the temperature value, state, and thresholds. The temperature value is the temperature in the Device (not the external temperature). You can configure only the yellow threshold level (in Celsius) by using the system env temperature threshold yellow value global configuration command to set the difference between the yellow and red thresholds. You cannot configure the green or red thresholds. For more information, see the command reference for this release.

You can monitor the physical path that a packet takes from a source device to a destination device by using one of these privileged EXEC commands:

To run TDR, enter the test cable-diagnostics tdr interface interface-id privileged EXEC command.

To display the results, enter the show cable-diagnostics tdr interface interface-id privileged EXEC command.

The output from the show platform forward privileged EXEC command provides some useful information about the forwarding results if a packet entering an interface is sent through the system. Depending upon the parameters entered about the packet, the output provides lookup table results and port maps used to calculate forwarding destinations, bitmaps, and egress information.

Most of the information in the output from the command is useful mainly for technical support personnel, who have access to detailed information about the Device application-specific integrated circuits (ASICs). However, packet forwarding information can also be helpful in troubleshooting.

The show debug command is entered in privileged EXEC mode. This command displays all debug options available on the switch.

To view all conditional debug options run the command show debug condition The commands can be listed by selecting either a condition identifier <1-1000> or all conditions.

To disable debugging, use the no debug all command.

Caution

Because debugging output is assigned high priority in the CPU process, it can render the system unusable. For this reason, use debug commands only to troubleshoot specific problems or during troubleshooting sessions with Cisco technical support staff. Moreover, it is best to use debug commands during periods of lower network traffic and fewer users. Debugging during these periods decreases the likelihood that increased debug command processing overhead will affect system use.