Purpose of IP Access Lists

• Filter incoming or outgoing packets on an interface.

• Filter packets for mirroring.

• Redirect traffic as required.

• Restrict the contents of routing updates.

• Limit debug output based on an address or protocol.

• Control vty access.

• Identify or classify traffic for advanced features, such as congestion avoidance, congestion management, and priority and custom queueing.

How an IP Access List Works

An access list is a sequential list consisting of permit and deny statements that apply to IP addresses and possibly upper-layer IP protocols. The access list has a name by which it is referenced. Many software commands accept an access list as part of their syntax.

An access list can be configured and named, but it is not in effect until the access list is referenced by a command that accepts an access list. Multiple commands can reference the same access list. An access list can control traffic arriving at the router or leaving the router, but not traffic originating at the router.

Source address and destination addresses are two of the most typical fields in an IP packet on which to base an access list. Specify source addresses to control packets from certain networking devices or hosts. Specify destination addresses to control packets being sent to certain networking devices or hosts.

You can also filter packets on the basis of transport layer information, such as whether the packet is a TCP, UDP, ICMP, or IGMP packet.

ACL Workflow

The following image illustrates the workflow of an ACL.

ACL Protocols

The following ACL protocol types are supported on NCS5500:

1. IPv4

2. IPv6

3. L2 (Ethernet-Services)

When an ACL of a particular protocol type is attached to an interface in a given direction, that interface must be configured for the specific protocol and the ACL only applies to traffic of that protocol type flowing on the interface.

IP Access List Process and Rules

• The software tests the source or destination address or the protocol of each packet being filtered against the conditions in the access list, one condition (permit or deny statement) at a time.

• The packet is matched with ACE within and ACL in the order of the sequence number.

• If a packet and an access list statement match, the remaining statements in the list are skipped and the packet is permitted or denied as specified in the matched statement. The first entry that the packet matches determines whether the software permits or denies the packet. That is, after the first match, no subsequent entries are considered.

• If the access list denies the address or protocol, the software discards the packet.

• If no conditions match, the software drops the packet because each access list ends with an unwritten or implicit deny statement. That is, if the packet has not been permitted or denied by the time it was tested against each statement, it is denied.

• The access list should contain at least one permit statement or else all packets are denied.

• Because the software stops testing conditions after the first match, the order of the conditions is critical. The same permit or deny statements specified in a different order could result in a packet being passed under one circumstance and denied in another circumstance.

• Only one access list per interface, per protocol, per direction is allowed.

• Inbound access lists process packets arriving at the router. Incoming packets are processed before being routed to an outbound interface. An inbound access list is efficient because it saves the overhead of routing lookups if the packet is to be discarded because it is denied by the filtering tests. If the packet is permitted by the tests, it is then processed for routing. For inbound lists, permit means continue to process the packet after receiving it on an inbound interface; deny means to discard the packet.

• Outbound access lists process packets before they leave the router. Incoming packets are routed to the outbound interface and then processed through the outbound access list. For outbound lists, permit means send it to the output buffer; deny means discard the packet.

• An access list cannot be removed if that access list is being applied by an access group in use. To remove an access list, remove the access group that is referencing the access list and then remove the access list.

• Before removing an interface, which is configured with an ACL that denies certain traffic, you must remove the ACL and commit your configuration. If this is not done, then some packets are leaked through the interface as soon as the no interface <interface-name> command is configured and committed.

• An access list must exist before you can use the ipv4 access group command.

• ACL-based Forwarding (ABF) is supported in common ACLs.

• Filtering of MPLS packets with the explicit-null or de-aggregation label is supported on the ingress direction.

• If the Ternary Content-Addressable Memory (TCAM) utilization is high and large ACLs are modified, then an error may occur. During such instances, remove the ACL from the interface and reconfigure the ACL. Later, reapply the ACL to the interface.

• You can configure an ACL name with a maximum of 64 characters.

• You can configure an ACL name to comprise of only letters and numbers.

ACL Filtering by Wildcard Mask and Implicit Wildcard Mask

Address filtering uses wildcard masking to indicate whether the software checks or ignores corresponding IP address bits when comparing the address bits in an access-list entry to a packet being submitted to the access list. By carefully setting wildcard masks, an administrator can select a single or several IP addresses for permit or deny tests.

Wildcard masking for IP address bits uses the number 1 and the number 0 to specify how the software treats the corresponding IP address bits. A wildcard mask is sometimes referred to as an inverted mask, because a 1 and 0 mean the opposite of what they mean in a subnet (network) mask.

• A wildcard mask bit 0 means check the corresponding bit value.

• A wildcard mask bit 1 means ignore that corresponding bit value.

You do not have to supply a wildcard mask with a source or destination address in an access list statement. If you use the host keyword, the software assumes a wildcard mask of 0.0.0.0.

Unlike subnet masks, which require contiguous bits indicating network and subnet to be ones, wildcard masks allow noncontiguous bits in the mask.

You can also use CIDR format (/x) in place of wildcard bits. For example, the IPv4 address 1.2.3.4 0.255.255.255 corresponds to 1.2.3.4/8 and for IPv6 address 2001:db8:abcd:0012:0000:0000:0000:0000 corresponds to 2001:db8:abcd:0012::0/64.

Including Comments in Access Lists

You can include comments (remarks) about entries in any named IP access list using the remark access list configuration command. The remarks make the access list easier for the network administrator to understand and scan. Each remark line is limited to 255 characters.

The remark can go before or after a permit or deny statement. You should be consistent about where you put the remark so it is clear which remark describes which permit or deny statement. For example, it would be confusing to have some remarks before the associated permit or deny statements and some remarks after the associated statements. Remarks can be sequenced.

Remember to apply the access list to an interface or terminal line after the access list is created.

Limitations

• For Cisco NCS 5700 Series platforms and NC57 line cards in native mode, if you configure an IPv4 ACL with an ACE that permits UDP packets on the L2TP reserved port, the UDP packets on the port are implicitly denied despite meeting the permit conditions.

Notes and Restrictions for Configuring IPv4 Ingress ACLs

IPv4 ingress ACLs are characterized by the following behavior.

• Ingress IPv4 ACLs are supported on all interfaces except management interfaces.

• ACL-based Forwarding (ABF) is supported only in the ingress direction.

• The following line card limits apply for traditional ingress ACLs:

  • A maximum of 127 different traditional ingress ACLs per LC in shared ACL mode.

  • A maximum of 255 different hybrid ingress ACLs per LC in shared ACL mode.

  • Unique ACL mode allows you to configure more than 128 different traditional ingress ACLs per LC.

  • From Cisco IOS XR Release 7.6.1 onwards, the router supports ACL logging with input interface (using the log-input keyword). Prior to Cisco IOS XR Release 7.6.1, the router does not support ACL logging with input interface on the router.

  From Cisco IOS XR Release 7.6.1 onwards, the line card limits have increased for traditional and hybrid ingress ACLs in shared ACL mode on the Cisco NCS 5700 Series Routers and routers with the Cisco NC57 line cards installed and operating in either native or compatibility mode, as listed below:

  • A maximum of 512 different traditional ingress ACLs per LC.

  • A maximum of 1000 different hybrid ingress ACLs per LC.

• The SE variant of the NCS 5700 line cards cannot match the IP-in-IP tunnel decapsulation packet outer header with Ingress ACL including the Hybrid ACL in external TCAM.

Notes and Restrictions for Configuring IPv4 Egress ACLs

IPv4 egress ACLs are characterized by the following behavior.

• Egress IPv4 ACLs are supported on main physical interfaces and bundle interfaces.

  Note	If you configure an egress ACL on a main interface that has sub-interfaces, the ACL action is also applied to the sub-interface traffic. This egress ACL behavior holds true even if the sub-interfaces are configured after the ACL is applied to the main interface.

• The total number of egress ACLs allowed per NPU is 255.

• ACL is not supported on Management interface on egress direction.

• Apart from the throughput limitation, router-generated traffic is not be affected by egress IPv4 ACLs.

• The number of attached ACEs allowed per line card is 4096.

• When hierarchical QoS is enabled, egress ACLs are supported on the sub-interfaces. However, an egress ACL configured on the main interfaces is not to be applied to the sub-interface traffic.

• Filtering for egress IPv4 multicast traffic is not supported if H-QoS is configured on the router.

• Starting with Release 7.7.1, shared ACL mode supports egress ACLs and the same egress ACL applied to multiple physical interfaces utilizes the same NPU resource. But, if you enable the H-QoS profile, then the egress ACLs operate as unique ACLs and utilize separate NPU resources. Egress ACL supports shared ACL mode only in physical interfaces and not in subinterfaces, bundle interfaces, and BVI interfaces. The Egress ACL in shared ACL mode is applicable only for Cisco routers that have the Cisco NC57 line cards installed and operate in the native mode and Cisco NCS 5700 fixed port routers.

Configuring an Ingress IPv4 ACL on a Gigabit Ethernet Interface

Use the following configuration to configure an ingress IPv4 ACL on a GigE interface.

/* Configure a GigE interface with an IPv4 address */
Router(config)#  
Router(config-if)# ipv4 address 10.1.1.1 255.255.255.0
Router(config-if)# no shut
Router(config-if)# commit
Thu Jan 25 10:07:54.700 IST
Router(config-if)# exit

/* Verify if the interface is up */
Router(config)# do show ipv4 interface brief
Thu Jan 25 10:08:49.087 IST

Interface                      IP-Address      Status          Protocol Vrf-Name         10.1.1.1        Up              Up       default 

/* Configure an IPv4 ingress ACL */
Router(config)# ipv4 access-list V4-ACL-INGRESS
Router(config-ipv4-acl)# 10 permit tcp 10.2.1.1 0.0.0.255 any
Router(config-ipv4-acl)# 20 deny udp any any
Router(config-ipv4-acl)# 30 permit ipv4 10.2.0.0 0.255.255.255 any
Router(config-ipv4-acl)# commit
Thu Jan 25 10:16:11.473 IST

/* Verify the ingress ACL creation */
Router(config)# do show access-lists ipv4 
Thu Jan 25 10:25:19.896 IST
...
ipv4 access-list V4-ACL-INGRESS
 10 permit tcp 10.2.1.0 0.0.0.255 any
 20 deny udp any any
 30 permit ipv4 10.0.0.0 0.255.255.255 any


/* Apply the ingress ACL to the GigE interface */
Router(config)# interface 
Router(config-if)# ipv4 access-group V4-ACL-INGRESS ingress
Router(config-if)# commit
Thu Jan 25 10:28:19.671 IST
Router(config-if)# exit

/* Verify if the ingress ACL has been successfully applied to the interface */
Router(config)# do show ipv4 interface 
Thu Jan 25 10:29:44.944 IST
 is Up, ipv4 protocol is Up
  Vrf is default (vrfid 0x60000000)
  Internet address is 10.1.1.1/24
  MTU is 1514 (1500 is available to IP)
  Helper address is not set
  Directed broadcast forwarding is disabled
  Outgoing access list is not set
  Inbound  common access list is not set, access list is V4-ACL-INGRESS
  Proxy ARP is disabled
  ICMP redirects are never sent
  ICMP unreachables are always sent
  ICMP mask replies are never sent
  Table Id is 0xe0000000

You have successfully configured an IPv4 ingress ACL on a Gigabit Ethernet interface.

Configuring an Egress IPv4 ACL on a Gigabit Ethernet Interface

Use the following configuration to configure an egress IPv4 ACL on a GigE interface.

/* Configure a GigE interface with an IPv4 address */
Router(config)# 
Router(config-if)# ipv4 address 20.1.1.1 255.255.255.0
Router(config-if)# no shut 
Router(config-if)# commit
Thu Jan 25 10:08:38.767 IST
Router(config-if)# exit

/* Verify if the interface is up */
Router(config)# do show ipv4 interface brief
Thu Jan 25 10:08:49.087 IST

Interface                      IP-Address      Status          Protocol Vrf-Name
         10.1.1.1        Up              Up       default 
         20.1.1.1        Up              Up       default 

/* Configure an IPv4 egress ACL */
Router(config)# ipv4 access-list V4-ACL-EGRESS
Router(config-ipv4-acl)# 10 permit ipv4 10.2.0.0 0.255.255.255 20.2.0.0 0.255.255.255
Router(config-ipv4-acl)# 20 deny ipv4 any any
Router(config-ipv4-acl)# commit
Thu Jan 25 10:25:04.655 IST

/* Verify the egress ACL creation */
Router(config)# do show access-lists ipv4 
Thu Jan 25 10:25:19.896 IST
ipv4 access-list V4-ACL-EGRESS
 10 permit ipv4 10.0.0.0 0.255.255.255 20.0.0.0 0.255.255.255
 20 deny ipv4 any any
...

/* Apply the egress ACL to the GigE interface */
Router(config)# 
Router(config-if)# ipv4 access-group V4-ACL-EGRESS egress
Router(config-if)# commit
Thu Jan 25 10:28:45.937 IST
Router(config-if)# exit

/* Verify if the egress ACL has been successfully applied to the interface */
Router(config)# do show ipv4 interface 
Thu Jan 25 10:29:44.944 IST
 is Up, ipv4 protocol is Up
  Vrf is default (vrfid 0x60000000)
  Internet address is 20.1.1.1/24
  MTU is 1514 (1500 is available to IP)
  Helper address is not set
  Directed broadcast forwarding is disabled
  Outgoing access list is V4-ACL-EGRESS
  Inbound  common access list is not set, access list is not set
  Proxy ARP is disabled
  ICMP redirects are never sent
  ICMP unreachables are always sent
  ICMP mask replies are never sent
  Table Id is 0xe0000000
...

You have successfully configured an IPv4 egress ACL on a Gigabit Ethernet interface.

Notes and Restrictions for Configuring IPv6 Ingress ACLs

IPv6 ingress ACLs are characterized by the following behavior.

• Ingress IPv6 ACLs are supported on all interfaces.

• ACL-based Forwarding (ABF) is supported only in the ingress direction.

• The following line card limits apply for traditional ingress ACLs:

  From Cisco IOS XR Release 7.6.1 onwards, the line card limits have increased for traditional and hybrid ingress ACLs in shared ACL mode on the Cisco NCS 5700 Series Routers and routers with the Cisco NC57 line cards installed and operating in either native or compatibility mode, as listed below:

  • A maximum of 512 different traditional ingress ACLs per LC.

  • A maximum of 1000 different hybrid ingress ACLs per LC.

• The number of attached ACEs allowed per line card is 2048 in the ingress direction.

• From Cisco IOS XR Release 7.6.1 onwards, the router supports ACL logging with input interface (using the log-input keyword). Prior to Cisco IOS XR Release 7.6.1, the router does not support ACL logging with input interface on the router.

• Packet Length (using the pkt-length keyword) is not supported.

  Note	Packet Length using the pkt-length keyword) is supported on the NC57-24DD and NC57-18DD-SE line cards for IPv6 ingress ACLs.

• Without an external TCAM, traditional IPv6 DSCP and IPv6 Precedence are not supported in the ingress direction for NC57 line cards.

• The SE variant of the NCS 5700 line cards cannot match the IP-in-IP tunnel decapsulation packet outer header with Ingress ACL icluding the Hybrid ACL in external TCAM.

Notes and Restrictions for Configuring IPv6 Egress ACLs

IPv6 egress ACLs are characterized by the following behavior:

• Configuring packet length is not supported on egress ACLs.

• TCP flags are not supported on egress ACLs.

• For NCS 5700 line cards, you can configure a destination address of upto 96 bits prefix for an ACE.

• For NC57-24DD and NC57-18DD-SE line cards, you can configure packet length, DSCP, fragments, and IPv6 extension headers, and TCP flags on egress ACLs.

• Egress ACLs are not supported on L2 interfaces. For NC57-24DD and NC57-18DD-SE line cards, ACLs are not supported on L2 interfaces.

• Egress IPv6 ACLs are supported on BVI interfaces for NC57 line cards, in native mode. Reload the router to commit the configuration.

• Configuring qos-group is not supported on egress ACLs.

• A throughput of 50% or less is supported on egress ACLs.

  Note	For the NC57-24DD and NC57-18DD-SE line cards, 100% throughput is supported on IPv6 egress ACLs.

• Apart from the throughput limitation, router-generated traffic is not be affected by egress IPv6 ACLs.

• The total number of egress ACLs allowed per NPU is 255.

• The total number of attached ACEs allowed per line card is 2048 in the egress direction.

• Configuring dynamic TCAM key is not supported on egress ACLs.

• Upto 160GB of total IPv6 egress ACL is supported per NPU because the Egress IPv6 ACLs take the recycle path.

  Note	For the NC57-24DD and NC57-18DD-SE line cards, 100% throughput for IPv6 egress ACL is supported per NPU.

• When hierarchical QoS is enabled, egress ACLs are supported on the sub-interfaces. However, an egress ACL configured on the main interfaces is not to be applied to the sub-interface traffic.

• Filtering for egress IPv6 multicast traffic is not supported if H-QoS is configured on the router.

• Starting with Release 7.7.1, shared ACL mode supports egress ACLs and the same egress ACL applied to multiple physical interfaces utilizes the same NPU resource. But, if you enable the H-QoS profile, then the egress ACLs operate as unique ACLs and utilize separate NPU resources. Egress ACL supports shared ACL mode only in physical interfaces and not in subinterfaces, bundle interfaces, and BVI interfaces. The Egress ACL in shared ACL mode is applicable only for Cisco routers that have the Cisco NC57 line cards installed and operate in the native mode and Cisco NCS 5700 fixed port routers.

• In IPv6 Egress ACLs, TCP flag filtering does not function for IPv6 packets with a fragmentation header. As a result, IPv6 packets with both a fragmentation header and a TCP header (ACK+SYN flags) are not appropriately filtered by the ACL rules.

Configuring an Ingress IPv6 ACL on a Gigabit Ethernet Interface

Use the following configuration to configure an ingress IPv6 ACL on a GigE interface.

/* Configure a GigE interface with an IPv6 address */
Router(config)#  
Router(config-if)# ipv6 address 1001::1/64
Router(config-if)# no shut
Router(config-if)# commit
Thu Jan 25 10:07:54.700 IST
Router(config-if)# exit

/* Verify if the interface is up */
Router(config)# do show ipv6 interface brief
Thu Jan 25 12:38:35.742 IST
 [Up/Up]
    fe80::bd:b9ff:fea9:5606                       
    1001::1                                       
…

/* Configure an IPv6 ingress ACL */
Router(config)# ipv6 access-list V6-INGRESS-ACL
Router(config-ipv6-acl)# 10 permit ipv6 any any
Router(config-ipv6-acl)# 20 deny udp any any
Router(config-ipv6-acl)# commit
Thu Jan 25 11:31:24.488 IST
Router(config-ipv6-acl)# exit 


/* Verify the ingress ACL creation */
Router(config)# do show access-lists ipv6
Thu Jan 25 11:34:56.911 IST
ipv6 access-list V6-INGRESS-ACL
 10 permit ipv6 any any
 20 deny udp any any


/* Apply the ingress ACL to the GigE interface */
Router(config)# 
Router(config-if)# ipv6 access-group V6-INGRESS-ACL ingress
Router(config-if)# commit
Thu Jan 25 11:32:55.194 IST
Router(config-if)# exit

/* Verify if the ingress ACL has been successfully applied to the interface */
Router(config)# do show ipv6 interface 
Thu Jan 25 11:34:08.028 IST
 is Up, ipv6 protocol is Up, Vrfid is default (0x60000000)
  IPv6 is enabled, link-local address is fe80::bd:b9ff:fea9:5606 
  Global unicast address(es):
    1001::1, subnet is 1001::/64 
  Joined group address(es): ff02::1:ff00:1 ff02::1:ffa9:5606 ff02::2
      ff02::1
  MTU is 1514 (1500 is available to IPv6)
  ICMP redirects are disabled
  ICMP unreachables are enabled
  ND DAD is enabled, number of DAD attempts 1
  ND reachable time is 0 milliseconds
  ND cache entry limit is 1000000000
  ND advertised retransmit interval is 0 milliseconds
  Hosts use stateless autoconfig for addresses.
  Outgoing access list is not set
  Inbound  common access list is not set, access list is V6-INGRESS-ACL
  Table Id is 0xe0800000
  Complete protocol adjacency: 0
  Complete glean adjacency: 0
  Incomplete protocol adjacency: 0
  Incomplete glean adjacency: 0
  Dropped protocol request: 0
  Dropped glean request: 0
…

You have successfully configured an IPv6 ingress ACL on a Gigabit Ethernet interface.

Configuring an Egress IPv6 ACL on a Gigabit Ethernet Interface

Use the following configuration to configure an egress IPv6 ACL on a GigE interface.

/* Configure a GigE interface with an IPv6 address */
Router(config)# 
Router(config-if)# ipv6 address 2001::1/64
Router(config-if)# no shut
Router(config-if)# commit
Thu Jan 25 11:41:25.778 IST
Router(config-if)# exit

/* Verify if the interface is up */
Router(config)# do show ipv6 interface brief
Thu Jan 25 12:38:35.742 IST
 [Up/Up]
    fe80::bd:b9ff:fea9:5606                       
    1001::1                                       
 [Up/Up]
    fe80::23:e9ff:fea8:a44e                       
    2001::1    

/* Configure an IPv6 egress ACL */
Router(config)# ipv6 access-list V6-EGRESS-ACL 
Router(config-ipv6-acl)# 10 permit ipv6 any any
Router(config-ipv6-acl)# 20 deny udp any any
Router(config-ipv6-acl)# commit
Thu Jan 25 11:44:03.969 IST
Router(config-ipv6-acl)# exit

/* Verify the egress ACL creation */
Router(config)# do show access-lists ipv6
Thu Jan 25 11:45:53.823 IST
ipv6 access-list V6-EGRESS-ACL
 10 permit ipv6 any any
 20 deny udp any any
…


/* Verify if the egress ACL has been successfully applied to the interface */
Router(config)# do show ipv6 interface    
Thu Jan 25 11:46:43.234 IST
…
 is Up, ipv6 protocol is Up, Vrfid is default (0x60000000)
  IPv6 is enabled, link-local address is fe80::23:e9ff:fea8:a44e 
  Global unicast address(es):
    2001::1, subnet is 2001::/64 
  Joined group address(es): ff02::1:ff00:1 ff02::1:ffa8:a44e ff02::2
      ff02::1
  MTU is 1514 (1500 is available to IPv6)
  ICMP redirects are disabled
  ICMP unreachables are enabled
  ND DAD is enabled, number of DAD attempts 1
  ND reachable time is 0 milliseconds
  ND cache entry limit is 1000000000
  ND advertised retransmit interval is 0 milliseconds
   Hosts use stateless autoconfig for addresses.
  Outgoing access list is V6-EGRESS-ACL
  Inbound  common access list is not set, access list is not set
  Table Id is 0xe0800000
  Complete protocol adjacency: 0
  Complete glean adjacency: 0
  Incomplete protocol adjacency: 0
  Incomplete glean adjacency: 0
  Dropped protocol request: 0
  Dropped glean request: 0
...

You have successfully configured an IPv6 egress ACL on a Gigabit Ethernet interface.

Configuring Ingress and Egress IPv6 ACLs on Bundle Interfaces

Use the following configuration to configure ingress and egress IPv6 ACLs on a bundle interface.

/* Configure a bundle interface with an IPv6 address */
Router(config)# interface Bundle-Ether 1
Router(config-if)# ipv6 address 3001::1/64
Router(config-if)# no shut
Router(config-if)# commit
Thu Jan 25 13:53:47.435 IST
Router(config-if)# exit 

/* Configure an IPv6 egress ACL */
Router(config)# ipv6 access-list V6-EGRESS-ACL-bundle interface
Router(config-ipv6-acl)# 100 permit tcp any any eq www
Router(config-ipv6-acl)# 110 permit tcp any any eq https
Router(config-ipv6-acl)# 120 permit tcp any any eq ssh
Router(config-ipv6-acl)# 130 permit udp any any eq snmp
Router(config-ipv6-acl)# commit 
Thu Jan 25 13:57:14.960 IST
Router(config-ipv6-acl)# exit

/* Configure an IPv6 ingress ACL to deny ingress traffic on the bundle interface */
Router(config)# ipv6 access-list V6-DENY-INGRESS-ACL
Router(config-ipv6-acl)# 10 deny ipv6 any any
Router(config-ipv6-acl)# commit
Thu Jan 25 13:59:23.198 IST
Router(config-ipv6-acl)# exit

/* Verify the egress and ingress ACL creation */
Router(config)# do show access-lists ipv6
Thu Jan 25 14:00:24.055 IST
ipv6 access-list V6-DENY-INGRESS-ACL
 10 deny ipv6 any any
ipv6 access-list V6-EGRESS-ACL-bundle
 100 permit tcp any any eq www
 110 permit tcp any any eq https
 120 permit tcp any any eq ssh
 130 permit udp any any eq snmp
...

/* Apply the egress and ingress ACLs to the bundle interface */
Router(config)# interface Bundle-Ether 1
Router(config-if)# ipv6 access-group V6-EGRESS-ACL-bundle egress
Router(config-if)# ipv6 access-group V6-DENY-INGRESS-ACL ingress
Router(config-if)# commit
Thu Jan 25 14:04:19.536 IST
Router(config-if)# exit

/* Verify if the ACLs have been successfully applied to the interface */
Router(config)# do show ipv6 interface    
Thu Jan 25 11:46:43.234 IST
…
Thu Jan 25 14:04:51.322 IST
Bundle-Ether1 is Down, ipv6 protocol is Down, Vrfid is default (0x60000000)
  IPv6 is enabled, link-local address is fe80::1:10ff:fe87:8d04 [TENTATIVE]
  Global unicast address(es):
    3001::1, subnet is 3001::/64 [TENTATIVE]
  Joined group address(es): ff02::2 ff02::1
  MTU is 1514 (1500 is available to IPv6)
  ICMP redirects are disabled
  ICMP unreachables are enabled
  ND DAD is enabled, number of DAD attempts 1
  ND reachable time is 0 milliseconds
  ND cache entry limit is 1000000000
  ND advertised retransmit interval is 0 milliseconds
  ND router advertisements are sent every 160 to 240 seconds
  ND router advertisements live for 1800 seconds
  Hosts use stateless autoconfig for addresses.
  Outgoing access list is V6-EGRESS-ACL-BI
  Inbound  common access list is not set, access list is V6-DENY-INGRESS-ACL
  Table Id is 0xe0800000
  Complete protocol adjacency: 0
  Complete glean adjacency: 0
  Incomplete protocol adjacency: 0
  Incomplete glean adjacency: 0
  Dropped protocol request: 0
  Dropped glean request: 0

You have successfully configured ingress and egress IPv6 ACLs on a bundle interface.

TCP Flags

The Transmission Control Protocol (TCP) is one of the most widely used protocol for data transmission in networks.The TCP header contains several one-bit boolean fields known as flags used to influence the flow of data across a TCP connection. TCP packets use TCP flags during a packet transfer to indicate connection state or provide additional information about the packet transfer. The various TCP flags are SYN, ACK, FIN, RST, URG, PSH, and EST.

The following TCP flags can be present in a packet:

• SYN: Both the sender and receiver devices use the synchronisation (SYN) flag in only the first packet that is sent. This flag initiates a TCP connection by sending a synchronization request from the sender device to the receiver device.

• ACK: The receiver devices use the acknowledgment (ACK) flag in the packet that is sent to acknowledge the successful receipt of a packet.

• FIN: The sender device uses the finished (FIN) flag in the last packet to indicate that there is no more data to be sent.

• RST: The receiver device uses the reset (RST) flag in the packet sent to the sender device when the receiver device receives a packet that is not expected.

• URG: The sender device uses the urgent (URG) flag in the packets to notify the receiver device to process the urgent packets before processing all other packets.

• PSH: The receiver device uses the push (PSH) flag that is similar to the URG flag and tells the receiver to process these packets as soon as they are received without waiting for any other packets to be received.

• EST: When a remote host receives TCP packets with a SYN flag set and if it does not support such a service, the remote host replies with an EST flag set in the packet. EST flag signifies both ACK and RST flags set in the packet.

Restrictions and Configuration Guidelines

The following are the restrictions and gudelines for using TCP flags in Hybrid ACLs:

• The NC57 line cards must operate in the native mode. To enable native mode, use the hw-module profile npu native-mode-enable command.

Configuring ACLs Based on TCP Flags

You can use the match-any keyword in ACLs to permit or deny packets based on whether any of the configured TCP flags is set. Use the match-all keyword in IPv4 or IPv6 ACLs to permit or deny packets based on whether all the configured TCP flags are set.

Based on the TCP flags that you configure in your hybrid ACLs, the hybrid ACLs filter the packets.

The following example shows you how to create a hybrid IPv4 ACL and apply the TCP flags. After you save your configuration, attach the ACL to an interface and apply compress level 3. In this example, we have configured the following TCP flags:

• The match-any keyword and the ack, psh, and urg TCP flags set.

• The match-any keyword, and the syn TCP flag set and ack TCP flag not set.

• The match-all keyword, and the urg and fin TCP flags set.

• The match-all keyword, and the syn TCP flag set and ack TCP flag not set.

/* Enter the global configuration mode and create an ACL with name ACL-TCP.*/
Router# configure
Router(config)# ipv4 access-list ACL-TCP

/* Configure an ACL that permits packets with TCP flag that is either ACK, PSH, or URG */
Router(config-ipv4-acl)# 10 permit tcp any any match-any + ack + psh + urg
Router(config-ipv4-acl)# 12 permit tcp any any match-any + syn - ack
Router(config-ipv4-acl)# 14 permit tcp any any match-all + urg + fin
Router(config-ipv4-acl)# 16 permit tcp any any match-all + syn - ack
Router(config-ipv4-acl)# commit

/*  Attach the ACL to the interface and apply compress level 3 */
Router(config)#int hundredGigE 0/2/0/0 
Router(config-if)#ipv4 access-group ACL-TCP egress compress level 3

Note	Repeat the same steps to set TCP flags in an IPv6 hybrid ACL.

Running Configuration

Router#show run interface HundredGigE0/5/0/4
	interface HundredGigE0/5/0/4
	Ipv4 address 10.1.1.1/24
	Ipv6 add 10::1/64
	ipv4 access-group ACL-TCP egress compress level 3


Router#show run ipv4 access-list ACL-TCP
	ipv4 access-list ACL-TCP
	 
 	10 permit tcp net-group ACL-TCP_port_neq_4000 net-group ACL-TCP_port_neq_4000 match-all + ack + psh + urg

 	12 permit tcp net-group ACL-TCP_port_neq_4000 net-group network_V4_object_group port-group compress_port_neq_4000 match-any + syn – ack

 	14 permit tcp ACL-TCP_port_neq_4000 ACL-TCP_port_neq_4000 match-all + urg + fin

	 16 permit tcp ACL-TCP_port_neq_4000 net-group network_V4_object_group port-group compress_port_neq_4000 match-all + syn - ack

Verification

The number of packets/bytes matching the ACE with TCP flags can be verified using the below command:

Router#show access-lists ipv4 ACL-TCP hardware egress location 0/5/CPU0

ipv4 access-list ACL-TCP

 10 permit tcp net-group ACL-TCP_port_neq_4000 net-group ACL-TCP_port_neq_4000 match-all +ack +psh +urg (1915 matches) (294910 bytes)
 12 permit tcp net-group ACL-TCP_port_neq_4000 net-group ACL-TCP_neq_4000 match-any + syn – ack (1916 matches) (295064 bytes)

 14 permit tcp net-group ACL-TCP_port_neq_4000 net-group network_V4_object_group port-group compress_port_neq_4000 match-all + urg + fin
(1915 matches) (294910 bytes)
 
16 permit tcp net-group ACL-TCP_port_neq_4000 net-group ACL-TCP_port_neq_4000 match-all + syn - ack (1915 matches) (294910 bytes)
 
 

Feature Highlights

• ABF is only supported on ingress ACL.

• ABF supports nexthop modifications. You can modify a nexthop, remove a nexthop, or make changes between existing nexthops.

  Note	While defining an ACE rule, you must specify the VRF for all nexthops unless the nexthop is in the default VRF. This will ensure that the packets take the right path towards the nexthop.

• VRF-aware ABF is supported for IPv4 and IPv6 with up to three next hops.

• IPv4 ABF nexthops routed over GRE interfaces are supported.

• As ABF is ACL-based, packets that do not match an existing rule (ACE) in the common ACL, will check for a match in the interface ACL. If the packets do not find a match in both the ACLs are subject to the default ACL rule (drop all). If the ACL is being used for ABF-redirect only (not for security), then include an explicit ACE rule at the end of the ACL (lowest user priority) to match and "permit" all traffic. This ensures that all traffic that does not match an ABF rule is permitted and forwarded as normal.

• ABF is supported on permit rules only.

• VRF-select (where only the VRF is configured for the nexthop) is supported in ABF for IPv4 and IPv6 addresses, where up to three VRF next hops are allowed.

• ABF default route is not supported.

• Packets punted in the ingress direction from the NPU to the linecard CPU are not subjected to ABF treatment due to lack of ABF support in the slow path. These packets will be forwarded normally based on destination-address lookup by the software dataplane. Some examples of these types of packets are (but are not limited to) packets with IPv4 options, IPv6 extension headers, and packets destined for glean (unresolved/incomplete) adjacencies.

• Packets destined to the local IP interface ("for-us" packets) are subjected to redirect if they match the rule containing the ABF action. This can be avoided by either designing the rule to be specific enough to avoid matching the “for-us” packets or placing an explicit permit ACE rule (with higher priority) into the ACL before the matching ABF rule.

Configuration Example

To configure ACL-based forwarding, use the following configuration example:

/* Enter IPv4 access list configuration mode and configure an ACL: */
Router# configure
Router(config)# ipv4 access-list abf-acl

/* Set the conditions for the ACL and configure ABF: */
/* The next hop for this entry is specified. */
Router(config-ipv4-acl)# 10 permit ipv4 192.168.18.0 0.255.255.255 any nexthop1 ipv4 192.168.20.2
Router(config-ipv4-acl)# 15 permit ipv4 192.168.21.0 0.0.0.255 any
Router(config-ipv4-acl)# 20 permit ipv4 192.168.22.0 0.0.255.255 any nexthop1 ipv4 192.168.23.2
/* More than two nexthops */
Router(config-ipv4-acl)# 25 permit tcp any range 2000 3000 any range 4000 5000 nexthop1 ipv4 192.168.23.1 nexthop2 ipv4 192.168.24.1 nexthop3 ipv4 192.168.25.1

/* VRF support on ABF */
Router(config-ipv4-acl)# 30 permit tcp any eq www host 192.168.12.2 precedence immediate nexthop1 vrf vrf1_ipv4 ipv4 192.168.13.2 nexthop2 vrf vrf1_ipv4 ipv4 192.168.14.2

Router(config-ipv4-acl)# 35 permit ipv4 any any

Router(config-ipv4-acl)# commit

/* (Optional) Display ACL information: */
Router# show access-lists ipv4 abf-acl

To configure ABF with VRF-select with three next hops, use the following configuration example:

/* Enter IPv4 access list configuration mode and configure an ACL. */
Router# configure
Router(config)# ipv4 access-list abf-vrf-select

/* Set the conditions for the ACL and configure ABF with VRF-select with three next hops. */
Router(config-ipv4-acl)# 10 permit ipv4 60.1.1.5 0.0.0.255 any nexthop1 vrf VRF1 nexthop2 vrf VRF2 nexthop3 vrf VRF3
Router(config-ipv4-acl)# commit

Running Configuration

ipv4 access-list abf-acl
10 permit ipv4 192.168.18.0 0.255.255.255 any nexthop1 192.168.20.2
15 permit ipv4 192.168.21.0 0.0.0.255 any
20 permit ipv4 192.168.22.0 0.0.255.255 any nexthop1 192.168.23.2
25 permit tcp any range 2000 3000 any range 4000 5000 nexthop1 ipv4 192.168.23.1 nexthop2 ipv4 192.168.24.1 nexthop3 ipv4 192.168.25.1
30 permit tcp any eq www host 192.168.12.2 precedence immediate nexthop1 vrf vrf1_ipv4 ipv4 192.168.13.2 nexthop2 vrf vrf1_ipv4 ipv4 192.168.14.2
35 permit ipv4 any any
commit
!

ipv4 access-list TEST
 10 permit ipv4 60.1.1.5 0.0.0.255 any nexthop1 vrf VRF1 nexthop2 vrf VRF2 nexthop3 vrf VRF3
!

Verification

Use the following command to verify the IP nexthop state in ABF to ensure that the expected nexthop is up:

router# show access-lists ipv4 abf nexthops client pfilter_ea location 0/3/CPU0
Tue May 17 22:25:05.940 UTC
 
 ACL name : abf-acl
  ACE seq.           NH-1                   NH-2                   NH-3
 ---------  ---------------------  ---------------------  ---------------------
        20    Global 192.168.23.2    Not present        Not present
    status                     UP    Not present            Not present
     exist                     No           Not present            Not present
    pd ctx                Present          Not present            Not present
                Track not present      Track not present                     --
        25    Global 192.168.23.1    Global 192.168.24.1        Global 192.168.25.1
    status                     UP                     UP            UP
     exist                     Yes                    Yes            Yes
    pd ctx                Present                Present            Present
                Track not present      Track not present            Track not present

Use the following command to verify if ABF is currently attached to any interfaces at any linecard:

show access-lists usage pfilter location all

Configuration

The following section describes the procedure for configuring IPv4 ingress and egress ACLs on BVIs.

To configure IPv4 ingress and egress ACLs on a BVI, use the following procedure with sample configuration.

1. Configure the HW module.

   Router(config)# hw-module profile acl egress layer3 interface-based
   Router(config)# commit
   Router(config)# exit
   

2. Enter the Global Configuration mode, and configure an IPv4 ingress ACL.

   Router(config)# ipv4 access-list v4-acl-ingress
   Router(config-ipv4-acl)# 10 permit tcp any 10.1.1.0/24 dscp cs6
   Router(config-ipv4-acl)# 20 deny udp any any eq ssh
   Router(config-ipv4-acl)# 30 permit ipv4 any any
   Router(config-ipv4-acl)# commit
   Router(config-ipv4-acl)# exit
   
   

3. Configure an IPv4 egress ACL.

   Router(config)# ipv4 access-list v4-acl-egress
   Router(config-ipv4-acl)# 10 deny ipv4 any any fragments log
   Router(config-ipv4-acl)# 20 deny tcp any any ack
   Router(config-ipv4-acl)# 30 permit ipv4 any any
   Router(config-ipv4-acl)# commit
   Router(config-ipv4-acl)# exit
   
   

4. Configure the Gigabit Ethernet interface that must be mapped to the BVI, and enable it for Layer 2 transport.

   Router(config)# interface GigabitEthernet 0/0/0/0
   Router(config-if)# l2transport 
   Router(config-if-l2)# commit
   
   

5. Attach the ingress and egress ACLs to the BVI.

   Router(config)# interface BVI1
   Router(config-if)# ipv4 access-group v4-acl-ingress ingress
   Router(config-if)# ipv4 access-group v4-acl-egress egress
   Router(config-if)# commit
   Router(config-if)# exit
   
   

6. Configure the bridge domain with the Gigabit Ethernet interface and BVI.

   Router(config)# l2vpn
   Router(config-l2vpn)# bridge group BG1
   Router(config-l2vpn-bg)# bridge-domain B1
   Router(config-l2vpn-bg-bd)# interface GigabitEthernet 0/0/0/0  
   Router(config-l2vpn-bg-bd-ac)# routed interface BVI1                     
   Router(config-l2vpn-bg-bd)# commit
   Router(config-l2vpn-bg-bd)# exit
   Router(config-l2vpn-bg)# exit
   Router(config-l2vpn)# exit
   
   

7. Confirm that your configuration has been successfully committed.

   Router(config)# show run
   ...
   !
   ipv4 access-list v4-acl-egress
    10 deny ipv4 any any fragments log
    20 deny tcp any any ack
    30 permit ipv4 any any
   !
   ipv4 access-list v4-acl-ingress
    10 permit tcp any 10.1.1.0/24 dscp cs6
    20 deny udp any any eq ssh
    30 permit ipv4 any any
   !
   interface GigabitEthernet0/0/0/0
    l2transport
    !
   !
   interface BVI1
   ipv4 address 209.165.200.224/27
    ipv4 access-group v4-acl-ingress ingress
    ipv4 access-group v4-acl-egress egress
    
   !
   l2vpn
    bridge group BG1
     bridge-domain B1
      interface GigabitEthernet0/0/0/0
      !
      routed interface BVI1
     !
    !
   !
   end
   
   

You have successfully configured and enabled IPv4 ingress and egress ACL on a BVI.

Verification

This section explains how to verify the IPv4 ACL configuration.

In Executive Privilege mode, confirm that the ACLs are in operation.

Router# show access-lists interface bvi1
Tue May 9 10:01:25.732 EDT
Input ACL (common): GigabitEthernet 0/0/0/0 (interface): v4-acl-ingress
Output ACL: v4-acl-egress

Router# show access-lists summary
Tue May 9 10:02:01.167 EDT
ACL Summary:
Total ACLs configured: 2
Total ACEs configured: 6


Router# show access-lists ipv4 v4-acl-egress hardware egress location 0/0/CPU0  
ipv4 access-list v4-acl-egress
10 deny ipv4 any any fragments log (15214 matches)
20 deny tcp any any ack (15214 matches)
30 permit ipv4 any any (15214 matches)

The output clearly shows the configured ACLs, the total number of ACEs (three per ACL), and also the ACE matches in hardware.

Configuration

You can use the following configuration to configure the fragments keyword for an IPv4 access list:

/* Configure an Access List */
Router# configure
Router(config)# ipv4 access-list IPv4_Fragments

/* Configure the fragments keyword for the IPv4 access list */
Router(config-ipv4-acl)# 10 permit ipv4 any any fragments
Router(config-ipv4-acl)# commit

You can use the following configuration to configure the fragments keyword for an IPv6 access list:

/* Configure an Access List */
Router# configure
Router(config)# ipv6 access-list IPv6_Fragments

/* Configure the fragments keyword for the IPv6 access list */
Router(config-ipv6-acl)# 10 permit ipv6 any any fragments
Router(config-ipv6-acl)# commit

Associated Commands

• deny (IPv4)

• deny (IPv6)

• permit (IPv4)

• permit (IPv6)

Associated Topics

• Configuring an IPv4 ACL to Match on Fragment Type

• Matching by Fragment Offset in ACLs

Restrictions

Packet length filtering feature in ACL is subjected to these restrictions:

• Packet length filtering is supported only on ingress direction, for both traditional (non-compression) and hybrid (compression) ACLs.

• Only quantized (value divisible by 16) packet length filtering is supported for traditional ACLs on IPv4.

• Packet length filtering is not supported in the default TCAM key, but instead requires a User-Defined TCAM Key (UDK) that can be specified using the hw-module profile tcam format command as described in the configuration section.

• Packet length filtering is supported in the default TCAM key for NC57-18DD-SE line cards, without using the User-Defined-Key or compression mode.

Associated Commands

• deny (IPv4)

• deny (IPv6)

• packet-length

• permit (IPv4)

• permit (IPv6)

Limitations for using TTL matching and rewriting for IPv4 ACLs

Using TTL matching and rewriting for IPv4 ACLs is known to have the following limitations.

• TTL matching is supported only for ingress ACLs.

• ACL logging is not supported for ingress ACLs after a User-Defined TCAM Key (UDK) is configured with the enable-set-ttl option.

• If a TTL rewrite is applied to the outer IPv4 header of an IP-in-IP header, then when the outer IPv4 header is decapsulated, (by GRE decapsulation) the TTL rewrite is also applied to the inner IPv4 header.

• TTL matching is not supported in the default TCAM key, but instead requires a User-Defined TCAM Key (UDK) using the hw-module profile tcam format command as described in the configuration section.

Configuration

Use the following steps to configure TTL matching and rewriting for IPv4 ACLs.

/* Enable TTL matching and rewriting in the global configuration mode by using the hw-module command */
Router(config)# hw-module profile tcam format access-list ipv4 dst-addr dst-port proto port-range enable-set-ttl ttl-match

/* Configure an IPv4 ACL with the TTL parameters */
Router(config)# ipv4 access-list acl-v4
Router(config-ipv4-acl)# 10 deny tcp any any ttl eq 100
Router(config-ipv4-acl)# 20 permit tcp any any ttl range 1 50 set ttl 200
Router(config-ipv4-acl)# 30 permit tcp any any ttl neq 100 set ttl 255
Router(config-ipv4-acl)# commit
Thu Nov  2 12:22:58.948 IST

/* Attach the IPv4 ACL to the GigE interface */
Router(config)# interface GigabitEthernet 0/0/0/0
Router(config-if)# ipv4 address 15.1.1.1 255.255.255.0
Router(config-if)# ipv4 access-group acl-v4 ingress
Router(config-if)# commit

Running Configuration

Validate your configuration by using the show run command.

Router(config)# show run
Thu Nov  2 14:01:53.376 IST
Building configuration...
!! IOS XR Configuration 0.0.0
!! Last configuration change at Thu Nov  2 12:22:59 2017 by annseque
!
hw-module profile tcam format access-list ipv4 dst-addr dst-port proto port-range enable-set-ttl ttl-match
!
ipv4 access-list acl-v4
 10 deny tcp any any ttl eq 100
 20 permit tcp any any ttl range 1 50 set ttl 200
 30 permit tcp any any ttl neq 100 set ttl 255
!
interface GigabitEthernet0/0/0/0
 ipv4 address 15.1.1.1 255.255.255.0
 ipv4 access-group acl-v4 ingress
!

You have successfully configured TTL matching and rewriting for IPv4 ACLs.

Configuration Using UDK on Internal TCAM

Use the following configuration to enable unique, interface-based IPv4 ACLs on internal TCAMs.

Note

A reboot of the line cards is required after entering the hw-module profile command to activate the command. TCAM is allocated for each ACL on an interface, and is not shared across ACLs. Hence, for instance, if an ACL utitlizes 10 TCAM entries, and is applied to 100 interfaces, the total number of TCAM entries will be 1000.

/* Enable interface-based, unique IPv4 ACLs on internal TCAMs*/
Router(config)# hw-module profile tcam format access-list ipv4 src-addr dst-addr src-port dst-port proto precedence ttl-match tcp-flags frag-bit enable-set-ttl interface-based

/* Configure an IPv4 ACL with the TTL parameters */
Router(config)# ipv4 access-list acl-v4
Router(config-ipv4-acl)# 10 deny tcp any any ttl eq 100
Router(config-ipv4-acl)# 20 permit tcp any any ttl range 1 50 set ttl 200
Router(config-ipv4-acl)# 30 permit tcp any any ttl neq 100 set ttl 255
Router(config-ipv4-acl)# commit
Thu Nov  2 12:22:58.948 IST

/* Attach the IPv4 ACL to the GigE interface */
Router(config)# interface GigabitEthernet 0/0/0/0
Router(config-if)# ipv4 address 15.1.1.1 255.255.255.0
Router(config-if)# ipv4 access-group acl-v4 ingress
Router(config-if)# commit

Running Configuration

Validate your configuration by using the show run command.

Router(config)# show run
Thu Nov  2 14:01:53.376 IST
Building configuration...
!! IOS XR Configuration 0.0.0
!! Last configuration change at Thu Nov  2 12:22:59 2017 by annseque
!
hw-module profile tcam format access-list ipv4 src-addr dst-addr src-port dst-port proto precedence ttl-match tcp-flags frag-bit enable-set-ttl interface-based
!
ipv4 access-list acl-v4
 10 deny tcp any any ttl eq 100
 20 permit tcp any any ttl range 1 50 set ttl 200
 30 permit tcp any any ttl neq 100 set ttl 255
!
interface GigabitEthernet0/0/0/0
 ipv4 address 15.1.1.1 255.255.255.0
 ipv4 access-group acl-v4 ingress
!

You have successfully configured unique, interface-based IPv4 ACLs.

Configuration Without Using UDK on External TCAM

You can also configure unique, interface-based IPv4 ACLs on external TCAMs. Through this configuration you can use the internal TCAM space for other configurations. On NC57-18DD-SE line cards, you can configure interface-based unique IPv4 ACLs through the hw-module profile acl ingress layer3 ipv4 interface-based location location command as well.

Use the following configuration to enable unique, interface-based IPv4 ACLs on external TCAMs:

/* Enable interface-based, unique IPv4 ACLs on external TCAMs*/
Router(config)# hw-module profile acl ingress layer3 ipv4 interface-based location 0/1/CPU0

/* Configure an IPv4 ACL with the TTL parameters */
Router(config)# ipv4 access-list acl-v4
Router(config-ipv4-acl)# 10 permit tcp any any eq 1024
Router(config-ipv4-acl)# 20 permit tcp any any eq 2048 set ttl 200
Router(config-ipv4-acl)# 30 permit tcp any any neq 100 set ttl 210
Router(config-ipv4-acl)# commit
Thu Nov  2 12:22:58.948 IST

Limitations for using TTL matching and rewriting for IPv6 ACLs

Using TTL matching and rewriting for IPv6 ACLs is known to have the following limitations.

• TTL matching is supported only for ingress ACLs.

• ACL logging is not supported for ingress ACLs after a User-Defined TCAM Key (UDK) is configured with the enable-set-ttl option.

• If a TTL rewrite is applied to the outer IPv6 header of an IP-in-IP header, then when the outer IPv6 header is decapsulated, (by GRE decapsulation) the TTL rewrite is also applied to the inner IPv6 header.

• TTL matching is not supported in the default TCAM key, but instead requires a User-Defined TCAM Key (UDK) using the hw-module profile tcam format command as described in the Configuration section.

Configuration

Use the following steps to configure TTL matching and rewriting for IPv6 ACLs.

/* Enable TTL matching and rewriting in the global configuration mode by using the hw-module command */
Router(config)# hw-module profile tcam format access-list ipv6 dst-addr dst-port src-port next-hdr enable-set-ttl ttl-match

/* Configure an IPv6 ACL with the TTL parameters */
Router(config)# ipv6 access-list acl-v6
Router(config-ipv6-acl)# 10 deny tcp any any ttl eq 50
Router(config-ipv6-acl)# 20 permit tcp any any ttl lt 50 set ttl 255
Router(config-ipv6-acl)# 30 permit tcp any any ttl gt 50 set ttl 200
Router(config-ipv6-acl)# commit
Thu Nov  2 12:22:58.948 IST

/* Attach the IPv6 ACL to the GigE interface */
Router(config)# interface GigabitEthernet 0/0/0/0
Router(config-if)# ipv6 address 2001:2:1::1/64
Router(config-if)# ipv6 access-group acl-v6 ingress
Router(config-if)# commit

Running Configuration

Validate your configuration by using the show run command.

Router(config)# show run
Thu Nov  2 14:01:53.376 IST
Building configuration...
!! IOS XR Configuration 0.0.0
!! Last configuration change at Thu Nov  2 12:22:59 2017 by annseque
!hw-module profile tcam format access-list ipv6 dst-addr dst-port src-port next-hdr enable-set-ttl ttl-match
!
ipv6 access-list acl-v6
 10 deny tcp any any ttl eq 50
20 permit tcp any any ttl lt 50 set ttl 255
30 permit tcp any any ttl gt 50 set ttl 200
!
interface GigabitEthernet0/0/0/0
 ipv6 address 2001:2:1::1/64
 ipv6 access-group acl-v6 ingress
!

You have successfully configured TTL matching and rewriting for IPv6 ACLs.

Configuration

Use the following configuration to enable unique, interface-based IPv6 ACLs on internal TCAMs.

Note

A reboot of the line cards is required after entering the hw-module profile command to activate the command. TCAM is allocated for each ACL on an interface, and is not shared across ACLs. Hence, for instance, if an ACL utitlizes 10 TCAM entries, and is applied to 100 interfaces, the total number of TCAM entries will be 1000.

/* Enable interface-based, unique IPv6 ACLs on internal TCAMs*//
Router(config)# hw-module profile tcam format access-list ipv6 src-addr src-port dst-addr dst-port next-hdr traffic-class enable-capture enable-set-ttl ttl-match interface-based

/* Configure an IPv6 ACL with the TTL parameters */
Router(config)# ipv6 access-list acl-v6
Router(config-ipv6-acl)# 10 deny tcp any any ttl eq 100
Router(config-ipv6-acl)# 20 permit tcp any any ttl range 1 50 set ttl 200
Router(config-ipv6-acl)# 30 permit tcp any any ttl neq 100 set ttl 255
Router(config-ipv6-acl)# commit
Thu Nov  2 12:22:58.948 IST

/* Attach the IPv6 ACL to the GigE interface */
Router(config)# interface GigabitEthernet 0/0/0/0
Router(config-if)# ipv6 address 2001:2:1::1/64
Router(config-if)# ipv6 access-group acl-v6 ingress
Router(config-if)# commit

Running Configuration

Validate your configuration by using the show run command.

Router(config)# show run
Thu Nov  2 14:01:53.376 IST
Building configuration...
!! IOS XR Configuration 0.0.0
!! Last configuration change at Thu Nov  2 12:22:59 2017 by annseque
!
hw-module profile tcam format access-list ipv6 src-addr src-port dst-addr dst-port next-hdr traffic-class enable-capture enable-set-ttl ttl-match interface-based
!
ipv6 access-list acl-v6
 10 deny tcp any any ttl eq 100
 20 permit tcp any any ttl range 1 50 set ttl 200
 30 permit tcp any any ttl neq 100 set ttl 255
!
interface GigabitEthernet0/0/0/0
 ipv6 address 2001:2:1::1/64
 ipv6 access-group acl-v6 ingress
!

You have successfully configured unique, interface-based IPv6 ACLs.

Configuration Without Using UDK on External TCAM

You can also configure unique, interface-based IPv6 ACLs on external TCAMs. Through this configuration you can use the internal TCAM space for other configurations. On NC57-18DD-SE line cards, you can configure interface-based unique IPv6 ACLs through the hw-module profile acl ingress layer3 ipv6 interface-based location location command as well.

Use the following configuration to enable unique, interface-based IPv6 ACLs on external TCAMs:

/* Enable interface-based, unique IPv6 ACLs on external TCAMs*/
Router(config)# hw-module profile acl ingress layer3 ipv6 interface-based location 0/1/CPU0

/* Configure an IPv6 ACL with the TTL parameters */
Router(config)# ipv6 access-list acl-v6
Router(config-ipv6-acl)# 10 permit tcp any any eq 1024
Router(config-ipv6-acl)# 20 permit tcp any any eq 2048 set ttl 200
Router(config-ipv6-acl)# 30 permit tcp any any neq 100 set ttl 210
Router(config-ipv6-acl)# commit

Thu Nov  2 12:22:58.948 IST

Configuration Example

Router# configure
Router(config)# hw-module profile acl IPv6 ext-header permit

Router(config)# commit

Associated Commands

• acl ipv6 ext-header

Configuration Example

Creates an IPv4 named access list "acl_1". This access list permits ICMP protocol packets with any source and destination IPv4 address and denies TCP protocol packets with any source and destination IPv4 address and destination port greater than 5000.

Router#configure
Router(config)#ipv4 access-list acl_1


Router(config-ipv4-acl)#20 permit icmp any any
Router(config-ipv4-acl)#30 deny tcp any any gt 5000
Router(config-ipv4-acl)#commit

Running Configuration

Router#  show running-config ipv4 access-list acl_1
ipv4 access-list acl_1
 20 permit icmp any any
 30 deny tcp any any gt 5000
!

Verification

Verify that the permit and deny settings are according to the set configuration.

Router# show access-lists acl_1
ipv4 access-list acl_1
 20 permit icmp any any
 30 deny tcp any any gt 5000
Router#

Associated Commands

What to Do Next

After creating an access list, you must apply it to a line or an interface. ACL commit fails while adding and removing unique Access List Entries (ACE). This happens due to the absence of an assigned manager process. The user has to exit the ACL configuration mode and re-enter it before adding the first ACE.

Enable Logging on ACE

This section shows you how to enable the ACE of an ACL to log informational messages when it matches incoming packets, using the optional keyword log . The router supports this feature only for IPv4 or IPv6 ingress ACLs. The logging message includes the access list number, whether the packet was permitted or denied, the source IP address of the packet, and the number of packets from that source permitted or denied in the prior 5-minute interval.

Router#configure
Router(config)#ipv4 access-list test
Router(config-ipv4-acl)#10 permit udp 10.85.1.0 255.255.255.0 log
Router(config-ipv4-acl)#exit
Router(config)# interface FortyGigE0/0/0/22
Router(config-if)# ipv4 access-group test ingress
Router(config-if)# commit

Note

Set log-level to informational or higher with the logging console command, so that the router displays the ACL log-messages on the console. Router#configure Router(config)#logging console informational Router(config)# commit For more information on log-levels, see section Syslog Message Severity Levels in the Implementing System Logging chapter of the System Monitoring Configuration Guide.

The following snippet shows a sample log message:

Router: ipv4_acl_mgr[350]: %ACL-IPV4_ACL-6-IPACCESSLOGP : access-list test (10) permit udp 
10.85.1.2(0) -> 10.0.0.1(0), 1 packet 

Enable Ingress Interface Logging on ACE

This section shows you how to configure the ACE of an ACL with the optional keyword log-input . This option provides the same functionality as the log keyword, as described in the previous section Enable Logging on ACE, except that the log-message also includes the ingress interface on which the router receives the packet. The router supports this feature for both IPv4 and IPv6 ingress ACLs on main interfaces, sub-interfaces and bridged-virtual interfaces (BVI).

Router#configure
Router(config)#ipv4 access-list test
Router(config-ipv4-acl)#10 deny udp 10.1.1.0 255.255.255.0 log-input
Router(config-ipv4-acl)#exit
Router(config)# interface FortyGigE0/0/0/22
Router(config-if)# ipv4 access-group test ingress
Router(config-if)# commit

The following snippet shows a sample log message when the user has enabled this option on an ACE:

Router: ipv4_acl_mgr[132]: %ACL-IPV4_ACL-6-IPACCESSLOGP : access-list test (10) deny udp 
10.1.1.2(0) FortyGigE0/0/0/22-> 10.2.2.2(0), 63782 packets 

BGP Filtering Using Prefix Lists

Prefix lists can be used as an alternative to access lists in many BGP route filtering commands. It is configured under the Global configurations of the BGP protocol. The advantages of using prefix lists are as follows:

• Significant performance improvement in loading and route lookup of large lists.

• Incremental updates are supported.

• More user friendly CLI. The CLI for using access lists to filter BGP updates is difficult to understand and use because it uses the packet filtering format.

• Greater flexibility.

Before using a prefix list in a command, you must set up a prefix list, and you may want to assign sequence numbers to the entries in the prefix list.

How the System Filters Traffic by Prefix List

Filtering by prefix list involves matching the prefixes of routes with those listed in the prefix list. When there is a match, the route is used. More specifically, whether a prefix is permitted or denied is based upon the following rules:

• An empty prefix list permits all prefixes.

• An implicit deny is assumed if a given prefix does not match any entries of a prefix list.

• When multiple entries of a prefix list match a given prefix, the longest, most specific match is chosen.

Sequence numbers are generated automatically unless you disable this automatic generation. If you disable the automatic generation of sequence numbers, you must specify the sequence number for each entry using the sequence-number argument of the permit and deny commands in IPv4 prefix list configuration command. Use the no form of the permit or deny command with the sequence-number argument to remove a prefix-list entry.

The show commands include the sequence numbers in their output.

Configuration Example

Creates a prefix-list "pfx_2" with a remark "Deny all routes with a prefix of 10/8". This prefix-list denies all prefixes matching /24 in 128.0.0.0/8.

Router#configure
Router(config)#ipv4 prefix-list pfx_2


Router(config-ipv4_pfx)#10 remark Deny all routes with a prefix of 10/8
Router(config-ipv4_pfx)#20 deny 128.0.0.0/8 eq 24
/* Repeat the above step as necessary. Use the no sequence-number command to delete an entry. */

Router(config-ipv4_pfx)#commit

Running Configuration

Router#show running-config ipv4 prefix-list pfx_2  
ipv4 prefix-list pfx_2
 10 remark Deny all routes with a prefix of 10/8
 20 deny 128.0.0.0/8 eq 24
!

Verification

Verify that the permit and remark settings are according to the set configuration.

Router# show prefix-list pfx_2
ipv4 prefix-list pfx_2
 10 remark Deny all routes with a prefix of 10/8
 20 deny 128.0.0.0/8 eq 24
RP/0/RP0/CPU0:ios#

Associated Commands

Configuration Example

Assigns sequence numbers to entries in a named prefix list and how to add or delete an entry to or from a prefix list. It is assumed a user wants to revise a prefix list. Resequencing a prefix list is optional.

Note	It is possible to resequence ACLs for prefix-list but not for security ACLs.

Router#config
Router(config)#ipv4 prefix-list cl_1 


Router(config)#10 permit 172.16.0.0 0.0.255.255
/* Repeat the above step as necessary adding statements by sequence number where you planned; use the no sequence-number command to delete an entry */

Router(config)#commit
end
Router#resequence prefix-list ipv4 cl_1 20 15

Running Configuration

/*Before resequencing/*
Router#show running-config ipv4 prefix-list cl_1
ipv4 prefix-list cl_1
 10 permit 172.16.0.0/16
!
/* After resequencing using the resequence prefix-list ipv4 cl_1 20 15 command: */
Router#show running-config ipv4 prefix-list cl_1 
ipv4 prefix-list cl_1
 20 permit 172.16.0.0/16
!

Verification

Verify that the prefix list has been resequenced:

Router#show prefix-list cl_1  
ipv4 prefix-list cl_1
 20 permit 172.16.0.0/16

Associated Commands

Guidelines

• ACL-based Policing feature is available only in the ingress direction.

• Both IPv4 and IPv6 ACLs support ACL-based Policing.

• Layer-2 ACL does not support ACL-based Policing.

• The policing rate limit for this feature is not supported in PPS (Packets Per Second) although you could configure policer rate in PPS in the router. From the release 7.10.1 onwards, PPS is supported.

• ACL-based Policing does not work with ABF, capture, and set options.

• Each ACE can only have a single policing rate. The policing rate could be in bps, kbps, mbps, or gbps.

• ACL-based Policing is supported in hybrid ACLs only.

• ACL-based Policing feature is supported only on the Cisco NCS 5500 series routers that have the Cisco NC57 line cards that are installed and operating in the native mode.

Configuration

The following section details configuring policing rate for ACEs in IPv4 and IPv6 ACLs:

Configuring policing rate for IPv4 ACL

Router(config)# ipv4 access-list Test1
Router(config-ipv4-acl)# 10 permit 192.168.34.0 0.0.0.255 police 333 mbps
Router(config-ipv4-acl)# 20 permit 172.16.0.0 0.0.255.255 police 10 gbps
Router(config-ipv4-acl)# 30 permit 10.0.0.0 0.255.255.255 police 1274 kbps

Configuring policing rate for IPv6 ACL

Router(config)# ipv6 access-list Test2
Router(config-ipv6-acl)# 10 permit fec0:0:0:2::/64 any police 10 gbps
Router(config-ipv6-acl)# 20 permit any any police 1274 kbps

Verification

This section details the traffic patterns results for the policing rate in the ACEs:

Router(config)# show ipv4 access-list Test1 hardware ingress location 0/1/CPU0
10 permit 192.168.34.0 0.0.0.255 (Accepted: 130 packets, Dropped: 0 packets)
20 permit 172.16.0.0 0.0.255.255 (Accepted: 1005 packets, Dropped: 0 packets)
30 permit 10.0.0.0 0.255.255.255 (Accepted: 10303 packets, Dropped: 7 packets)

Router# show ipv6 access-lists Test2 hardware ingress location 0/1/CPU0
10 permit fec0:0:0:2::/64 any (Accepted: 24303 packets, Dropped: 0 packets)
20 permit any any (Accepted: 13 packets, Dropped: 0 packets)