Configuring Traffic Mirroring

This module describes the configuration of the traffic mirroring feature. Traffic mirroring is sometimes called port mirroring, or switched port analyzer (SPAN). You can then pass this traffic to a destination port on the same router.

Feature Release History

Release

Modification

Release 6.1.3

ERSPAN Traffic to a Destination Tunnel in a Default VRF was introduced.
Release 7.5.3

ERSPAN Traffic to a Destination Tunnel in a Non-Default VRF was introduced.

Introduction to Traffic Mirroring

Traffic mirroring, also referred to as Port mirroring or Switched Port Analyzer (SPAN), is a Cisco proprietary feature that enables you to monitor network traffic passing in or out of a set of ports on a router. You can then mirror this traffic to a remote destination or a destination port on the same router.

Traffic mirroring copies traffic from one or more source ports and sends the copied traffic to one or more destinations for analysis by a network analyzer or other monitoring devices. Traffic mirroring does not affect the flow of traffic on the source interfaces or sub-interfaces. It allows the mirrored traffic to be sent to a destination interface or sub-interface.

For example, you can attach a traffic or network analyzer to the router and capture the ethernet traffic that is sent by host A to host B.

Figure 1. Traffic Mirroring Operation

Traffic Mirroring Terminology

  • Ingress Traffic — Traffic that comes into the router.

  • Egress Traffic — Traffic that goes out of the router.

  • Source port—A port that is monitored with the use of traffic mirroring. It is also called a monitored port.

  • Destination port—A port that monitors source ports, usually where a network analyzer is connected. It is also called a monitoring port.

  • Monitor session—A designation for a collection of SPAN configurations consisting of a single destination and, potentially, one or many source ports.

Characteristics of Source Port

A source port, also called a monitored port, is a routed port that you monitor for network traffic analysis. In a single traffic mirroring session, you can monitor source port traffic. The Cisco NCS540 Series routers support a maximum of up to 800 source ports.

A source port has these characteristics:

  • It can be any data port type, such as Bundle Interface, 100 Gigabit Ethernet physical port, or 10 Gigabit Ethernet physical port.

  • Each source port can be monitored in only one traffic mirroring session.

  • When a port is used as a source port, the same port cannot be used as a destination port.

  • Each source port can be configured with a direction (ingress, egress, or both) to monitor local traffic mirroring. Remote traffic mirroring is supported both in the ingress and egress directions. For bundles, the monitored direction applies to all physical ports in the group.

Characteristics of Monitor Session

A monitor session is a collection of traffic mirroring configurations consisting of a single destination and, potentially, many source interfaces. For any given monitor session, the traffic from the source interfaces (called source ports) is sent to the monitoring port or destination port. If there are more than one source port in a monitoring session, the traffic from the several mirrored traffic streams is combined at the destination port. The result is that the traffic that comes out of the destination port is a combination of the traffic from one or more source ports.

Monitor sessions have these characteristics:

  • A single monitor session can have only one destination port.

  • A single destination port can belong to only one monitor session.

  • A monitor session can have a maximum of 800 source ports. This maximum limit is applicable only when the maximum number of source ports from all monitoring sessions does not exceed 800.

Characteristics of Destination Port

Each session must have a destination port or file that receives a copy of the traffic from the source ports.

A destination port has these characteristics:

  • A destination port cannot be a source port.

  • A destination port must reside on the same router as the source port for local traffic mirroring. For remote mirroring, the destination is always a GRE tunnel.

  • For remote mirroring, the destination is a GRE tunnel.

  • A destination port for local mirroring can be any Ethernet physical port, EFP, GRE tunnel interface, or bundle interface. It can be a Layer 2 or Layer 3 transport interface.

  • A destination port on router cannot be a VLAN subinterface.

  • At any time, a destination port can participate in only one traffic mirroring session. 
A destination port in one traffic mirroring session cannot be a destination port for a second traffic mirroring session. In other words, no two monitor sessions can have the same destination port.

Supported Scale

Restrictions

Generic Restrictions

The following are the generic restrictions related to traffic mirroring:

  • Partial mirroring and sampled mirroring are not supported.

  • From Release 7.6.1, sub-interface configured as source interface is supported on SPAN.

  • The destination bundle interfaces flap when:

    • both the mirror source and destination are bundle interfaces in the Link Aggregation Control Protocol (LACP) mode.

    • mirror packets next-hop is a router or a switch instead of a traffic analyzer.

    This behavior is observed due to a mismatch of LACP packets on the next-hop bundle interface due to the mirroring of LACP packets on the source bundle interface.

  • Bridge group virtual interfaces (BVIs) are not supported as source ports or destination ports.

  • Bundle members cannot be used as destination ports.

  • Fragmentation of mirror copies is not handled by SPAN when SPAN destination MTU is less than the packet size. Existing behaviour if the MTU of destination interface is less than the packet size is as below:

    Platforms

    Rx SPAN

    Tx SPAN

    NCS 540

    Mirror copies are not fragmented. Receives whole packets as mirror copies.

    Mirror copies are fragmented.

    You can configure the SPAN destination with an MTU which is greater than the packet size.

  • SPAN only supports port-level source interfaces.

  • SPAN counters are not supported.

VLAN Sub-interface as Source Restrictions

ACL-based SPAN Restrictions

The following restrictions apply to SPAN-ACL:

Table 1. SPAN-ACL Support

Platforms

Rx Direction

Tx Direction

NCS 540

Supported at the port level, that is, in the ingress direction for IPv4 or IPv6 ACLs.

Not supported.

  • MPLS traffic cannot be captured with SPAN-ACL.

    • ACL for any MPLS traffic is not supported.

  • Traffic mirroring counters are not supported.

  • ACL-based traffic mirroring is not supported with Layer 2 (ethernet-services) ACLs.

  • Main interface as span source interface and ACL with the capture keyword on same main interface's sub-interface are not supported.

  • If a SPAN session with the acl keyword is applied on an interface with no ACL rule attached to that interface, SPAN happens without any filtering.

  • Configure one or more ACLs on the source interface to avoid default mirroring of traffic. If a Bundle interface is a source interface, configure the ACL on the bundle interface (not bundle members). Also, ensure that the ACL configured is a UDK (with capture field) and of the same protocol type and direction as the SPAN configuration. For example, if you configure SPAN with ACL for IPv4 or IPv6, configure an ingress IPv4 UDK (with capture) or IPv6 UDK (with capture) on that network processing unit respectively.

  • Configure one or more ACLs on the source interface or any interface on the same network processing unit as the source interface, to avoid default mirroring of traffic. If a Bundle interface is a source interface, configure the ACL on any interface on the same network processing unit as all active bundle-members. Bundle members can be on multiple NPUs. Also, ensure that the ACLs configured are of the same protocol type and direction as the SPAN configuration. For example, if you configure SPAN with ACL for IPv4 or IPv6, configure an ingress IPv4 or IPv6 ACL on that network processing unit respectively.

ERSPAN Restrictions

This section provides the restrictions that apply to ERSPAN and multiple ERSPAN sessions.

The following restrictions apply to ERSPAN:

  • The value of ERSPAN session-ID is always zero. IOS XR command for configuring ERSPAN is not available.

  • ERSPAN next-hop must have ARP resolved. Any other traffic or protocol will trigger ARP.

  • ERSPAN packets with outgoing interface having MPLS encapsulation are not supported.

    • Additional routers may encapsulate in MPLS.

  • ERSPAN sessions can be created only on physical interfaces. The sessions cannot be created on sub-interfaces.

  • ERSPAN tunnel statistics is not supported.

  • ERSPAN decapsulation is not supported.

  • ERSPAN does not work if the GRE next hop is reachable over sub-interface. For ERSPAN to work, the next hop must be reachable over the main interface.

  • ERSPAN decapsulation is not supported. Tunnel destination should be network analyzer.

SPAN Types, Supported Features, and Configurations

Local SPAN

This is the most basic form of traffic mirroring. The network analyzer or sniffer is attached directly to the destination interface. In other words, all monitored ports are located on the same router as the destination port.

SPAN on Subinterfaces

VLAN Subinterface as Ingress or Egress Source for Traffic Mirroring

Table 2. Feature History Table

Feature Name

Release Information

Feature Description

VLAN Subinterface as Ingress or Egress Source for Traffic Mirroring

 Release 7.6.1

You can now configure the VLAN subinterface as an egress or ingress source for traffic mirroring. This feature enables the monitoring of traffic mirrored on either egress or ingress or both directions.

You could configure mirror functionality only at the main interface level in earlier releases.

VLAN subinterface provides the flexibility to monitor ingress or egress, or both ingress/egress traffic from all the active subinterfaces of the source VLAN. The active subinterfaces in the source VLAN are considered as source subinterfaces. When subinterfaces are added or removed from the source VLAN, the corresponding traffic is added or removed from the monitoring sources.

VLAN Interface as Ingress Source for Traffic Mirroring

Configuration Example

Router# configure
Router(config)# monitor-session mon1 ethernet
Router(config-mon)# destination interface tunnel-ip 3
Router(config-mon)# exit
Router(config)# interface HundredGigE 0/1/0/1.10
Router(config-subif)# 
Router(config-if-mon)# commit
Running Configuration

Router# show run monitor-session mon1
monitor-session mon1 ethernet
 destination interface tunnel-ip3
!


Router# show run interface HundredGigE 0/1/0/1.10
 interface HundredGigE0/1/0/1.10
  encapsulation dot1q 10
   ipv4 address 101.1.2.1 255.255.255.252
    monitor-session mon1 ethernet 
    !
   !
  !
Verification
Router# show monitor-session status
Monitor-session mon1
Destination interface tunnel-ip3
================================================================================
Source Interface Dir Status
--------------------- ---- ----------------------------------------------------
HundredGigE 0/1/0/1.10  Both Operational

VLAN Interface as Egress Source for Traffic Mirroring

Configuration Example

Router(config)# interface HundredGigE 0/1/0/1.10
Router(config-subif)# 
Router(config-if-mon)# commit
Running Configuration

Router# show run monitor-session mon1
monitor-session mon1 ethernet
 destination interface tunnel-ip3
!


Router# show run interface HundredGigE 0/1/0/1.10
interface HundredGigE0/1/0/1.10
 encapsulation dot1q 20
  ipv4 address 102.1.2.1 255.255.255.252
   monitor-session mon1 ethernet 
   !
  !
 !
Verification
Router# show monitor-session status
Monitor-session mon1
Destination interface tunnel-ip3
================================================================================
Source Interface Dir Status
--------------------- ---- ----------------------------------------------------
HundredGigE 0/1/0/1.10  Both Operational

Monitoring Traffic Mirroring on a Layer 2 Interface

This section describes the configuration for monitoring traffic on a Layer 2 interface.

Configuration

To monitor traffic mirroring on a Layer 2 interface, configure the monitor under l2transport sub-config of the interface: 

RP/0/RP0/CPU0:router(config)# interface TenGigE0/0/0/42
RP/0/RP0/CPU0:router(config-if)# l2transport
RP/0/RP0/CPU0:router(config-if-l2)# monitor-session EASTON ethernet port-level

Verification

RP/0/RP0/CPU0:router# show monitor-session status 
Thu Aug 29 21:42:22.829 UTC
Monitor-session EASTON
Destination interface TenGigE0/0/0/20
=============================================
Source Interface      Dir   Status
--------------------- ----  -----------------
Te0/0/0/42 (port)     Both  Operational

ACL-based SPAN

Traffic is mirrored based on the configuration of the interface ACL.

You can mirror traffic based on the definition of an interface access control list. When you mirror Layer 3 traffic, the ACL is configured using the ipv4 access-list or the ipv6 access-list command with the capture option. The permit and deny commands determine if the packets in the traffic are permitted or denied. The capture option designates the packet is to be mirrored to the destination port, and it is supported only on permit type of Access Control Entries (ACEs).


Note

  • Prior to Release 6.5.1, ACL-based traffic mirroring required the use of UDK (User-Defined TCAM Key) with the enable-capture option so that the capture option can be configured in the ACL.

  • ACL must be defined before attaching the ACL name to SPAN source interface.

Configuring Security ACLs for Traffic Mirroring

This section describes the configuration for creating security ACLs for traffic mirroring.

In ACL-based traffic mirroring, traffic is mirrored based on the configuration of the interface ACL. You can mirror traffic based on the definition of an interface access control list. When you're mirroring Layer 3 or Layer 2 traffic, the ACL is configured using the ipv4 access-list or the ipv6 access-list command with the capture option. The permit and deny commands determine the behavior of the regular traffic.

Configure an IPv4 ACL for Traffic Mirroring

Use the following steps to configure ACLs for traffic mirroring.

/* Create an IPv4 ACL (TM-ACL) for traffic mirroring */ 
Router(config)# ipv4 access-list TM-ACL 
Router(config-ipv4-acl)# 10 permit udp 10.1.1.0 0.0.0.255 eq 10 any capture
Router(config-ipv4-acl)# 20 permit udp 10.1.1.0 0.0.0.255 eq 20 any 
Router(config-ipv4-acl)# exit
Router(config)# commit 

/* Validate the configuration */
Router(config)# show run
Thu May 17 11:17:49.968 IST
Building configuration...
!! IOS XR Configuration 0.0.0
!! Last configuration change at Thu May 17 11:17:47 2018 by user
…
ipv4 access-list TM-ACL
 10 permit udp 10.1.1.0 0.0.0.255 eq 10 any capture
 20 permit udp 10.1.1.0 0.0.0.255 eq 20 any
!
…

You have successfully configured an IPv4 ACL for traffic mirroring.

Configuring UDF-Based Security ACL for Traffic Mirroring

Before you begin

This section describes the configuration steps for UDF-based security ACLs for traffic mirroring.

Procedure

Step 1

configure

Example:


RP/0/RP0/CPU0:router# configure

Enters global configuration mode.

Step 2

udf udf-name header {inner | outer} {l2 | l3 | l4} offset offset-in-bytes length length-in-bytes

Example:


RP/0/RP0/CPU0:router(config)# udf udf3 header outer l4 offset 0 length 1
(config-mon)#

Example:


RP/0/RP0/CPU0:router(config)# udf udf3 header inner l4 offset 10 length 2 
(config-mon)#

Example:


RP/0/RP0/CPU0:router(config)# udf udf3 header outer l4 offset 50 length 1
(config-mon)#

Configures individual UDF definitions. You can specify the name of the UDF, the networking header from which offset, and the length of data to be extracted.

The inner or outer keywords indicate the start of the offset from the unencapsulated Layer 3 or Layer 4 headers, or if there is an encapsulated packet, they indicate the start of offset from the inner L3/L4.

Note

 

The maximum offset allowed, from the start of any header, is 63 bytes

The length keyword specifies, in bytes, the length from the offset. The range is from 1 to 4.

Step 3

ipv4 access-list acl-name

Example:


RP/0/RP0/CPU0:router(config))# ipv4 access-list acl1

Creates ACL and enters IP ACL configuration mode. The length of the acl-name argument can be up to 64 characters.

Step 4

permit regular-ace-match-criteria udf udf-name1 value1 ... udf-name8 value8

Example:


RP/0/RP0/CPU0:router(config-ipv4-acl)# 10 permit ipv4 any any udf udf1 0x1234 0xffff udf3 0x56 0xff capture
RP/0/RP0/CPU0:router(config-ipv4-acl)# 30 permit ipv4 any any dscp af11 udf udf5 0x22 0x22 capture

Configures ACL with UDF match.

Step 5

exit

Example:


RP/0/RP0/CPU0:router(config-ipv4-acl)# exit

Exits IP ACL configuration mode and returns to global configuration mode.

Step 6

interfacetype number

Example:


RP/0/RP0/CPU0:router(config)# interface HundredGigE 0/0/1/0

Configures interface and enters interface configuration mode.

Step 7

ipv4 access-group acl-name ingress

Example:


RP/0/RP0/CPU0:router(config-if)# ipv4 access-group acl1 ingress

Applies access list to an interface.

Step 8

commit

Example:


RP/0/RP0/CPU0:router(config-if)# commit

Applies access list to an interface.

Verifying UDF-based Security ACL

Use the show monitor-session status detail command to verify the configuration of UDF on security ACL. 

RP/0/RP0/CPU0:leaf1# show monitor-session 1 status detail 

Fri May 12 19:40:39.429 UTC
Monitor-session 1
  Destination interface tunnel-ip3
  Source Interfaces
  -----------------
  TenGigE0/0/0/15
    Direction:  Rx-only
    Port level: True
    ACL match:  Enabled
    Portion:    Full packet
    Interval:   Mirror all packets
    Status:     Not operational (destination not active)

Attaching the Configurable Source Interface

Procedure

Step 1

configure

Example:


RP/0/RP0/CPU0:router# configure

Enters global configuration mode.

Step 2

interface type number

Example:


RP/0/RP0/CPU0:router(config)# interface HundredGigE 0/0/1/0

Enters interface configuration mode for the specified source interface. The interface number is entered in rack /slot /module /port notation. For more information about the syntax for the router, use the question mark (?) online help function.

Step 3

ipv4 access-group acl-name {ingress | egress}

Example:


RP/0/RP0/CPU0:router(config-if)# ipv4 access-group acl1 ingress

Controls access to an interface.

Step 4

monitor-session session-name ethernet direction rx-onlyport-level acl

Example:


RP/0/RP0/CPU0:router(config-if)# monitor-session mon1 ethernet direction rx-only port-level acl
RP/0/RP0/CPU0:router(config-if-mon)#

Attaches a monitor session to the source interface and enters monitor session configuration mode.

Note

 

rx-only specifies that only ingress traffic is replicated.

Step 5

acl

Example:


RP/0/RP0/CPU0:router(config-if-mon)# acl

Specifies that the traffic mirrored is according to the defined ACL.

Note

 

If an ACL is configured by name, then this step overrides any ACL that may be configured on the interface.

Step 6

exit

Example:


RP/0/RP0/CPU0:router(config-if-mon)# exit
RP/0/RP0/CPU0:router(config-if)#

Exits monitor session configuration mode and returns to interface configuration mode.

Step 7

end or commit

Example:


RP/0/RP0/CPU0:router(config-if)# end

or 


RP/0/RP0/CPU0:router(config-if)# commit

Saves configuration changes.

  • When you issue the end command, the system prompts you to commit changes: 

    
Uncommitted changes found, commit them before exiting (yes/no/cancel)?
[cancel]:

    - Entering yes saves configuration changes to the running configuration file, exits the configuration session, and returns the router to EXEC mode.

    - Entering no exits the configuration session and returns the router to EXEC mode without committing the configuration changes.

    - Entering cancel leaves the router in the current configuration session without exiting or committing the configuration changes.

  • Use the commit command to save the configuration changes to the running configuration file and remain within the configuration session.

Step 8

show monitor-session [session-name] status [detail] [error]

Example:


RP/0/RP0/CPU0:router# show monitor-session status

Displays information about the monitor session.

ERSPAN

Encapsulated Remote Switched Port Analyzer (ERSPAN) transports mirrored traffic over an IP network. The traffic is encapsulated at the source router and is transferred across the network. The packet is decapsulated at the destination router and then sent to the destination interface.

Encapsulated Remote SPAN (ERSPAN) enables generic routing encapsulation (GRE) for all captured traffic and allows it to be extended across Layer 3 domains.

ERSPAN involves mirroring traffic through a GRE tunnel to a remote site. For more information on configuring the GRE tunnel that is used as the destination for the monitor sessions, see the chapter Configuring GRE Tunnels.


Note

A copy of every packet includes the Layer 2 header if the ethernet keyword is configured. As this renders the mirrored packets unroutable, the end point of the GRE tunnel must be the network analyzer.

Introduction to ERSPAN Egress Rate Limit

With ERSPAN egress rate limit feature, you can monitor traffic flow through any IP network. This includes third-party switches and routers.

ERSAPN operates in the following modes:

  • ERSPAN Source Session – box where the traffic originates (is SPANned).

  • ERSPAN Termination Session or Destination Session – box where the traffic is analyzed.

This feature provides rate limiting of the mirroring traffic or the egress traffic. With rate limiting, you can limit the amount of egress traffic to a specific rate, which prevents the network and remote ERSPAN destination traffic overloading. Be informed, if the egress rate-limit exceeds then the system may cap or drop the monitored traffic.

You can configure the QoS parameters on the traffic monitor session.

  • Traffic Class (0 through 7)

    • Traffic class 0 has the lowest priority and 7 the highest.

    • The default traffic class is the same as that of the original traffic class.

  • The Discard Class (0 through 2):

    • The default is 0.

    • The discard class configuration is used in WRED.

Benefits

With ERSPAN Egress rate limit feature, you can limit the egress traffic or the mirrored and use the mirrored traffic for data analysis.

Topology

Figure 2. Topology for ERSPAN Egress Rate Limit

The encapsulated packet for ERSPAN is in ARPA/IP format with GRE encapsulation. The system sends the GRE tunneled packet to the destination box identified by an IP address. At the destination box, SPAN-ASIC decodes this packet and sends out the packets through a port. ERSPAN egress rate limit feature is applied on the router egress interface to rate limit the monitored traffic.

The intermediate switches carrying ERSPAN traffic from source session to termination session can belong to any L3 network.

Configure ERSPAN Egress Rate Limit

Use the following steps to configure ERSPAN egress rate limit:

monitor-session ERSPAN ethernet
destination interface tunnel-ip1
!
RP/0/RP0/CPU0:pyke-008#sh run int tunnel-ip 1
interface tunnel-ip1
ipv4 address 4.4.4.1 255.255.255.0
tunnel mode gre ipv4
tunnel source 20.1.1.1
tunnel destination 20.1.1.2
!

RP/0/RP0/CPU0:pyke-008#sh run int hundredGigE 0/0/0/16

interface HundredGigE0/0/0/16
ipv4 address 215.1.1.1 255.255.255.0
ipv6 address 3001::2/64
monitor-session ERSPAN ethernet direction rx-only port-level
  acl
!
ipv4 access-group ACL6 ingress
Running Configuration
!! Policy-map to be used with the ERSPAN Destination (egress interface)
!! Traffic class is set to 5. For packets in this class, apply shaping
!! as well as WRED.
class-map match-any TC5
 match traffic-class 5 
 end-class-map
! 
policy-map shape-foo
 class TC5
  random-detect discard-class 0 10000 bytes 40000 bytes 
  random-detect discard-class 1 40000 bytes 80000 bytes 
  random-detect discard-class 2 80000 bytes 200000 bytes 
  shape average percent 15 
 !  
 class class-default
 ! 
 end-policy-map
! 
!!GRE Tunnel Interface
interface Loopback49
 ipv4 address 49.49.49.49 255.255.255.255
!
interface tunnel-ip100
 ipv4 address 130.100.1.1 255.255.255.0
 tunnel mode gre ipv4
 tunnel source 49.49.49.49
 tunnel destination 10.8.1.2
!
!!ERSPAN Monitor Session with GRE tunnel as the Destination Interface, and with QoS configuration
monitor-session FOO ethernet
 destination interface tunnel-ip100
 traffic-class 5
 discard-class 1
!
!!ERSPAN Source Interface
interface TenGigE0/6/0/4/0
 description connected to TGEN 9/5
 ipv4 address 10.4.90.1 255.255.255.0
 monitor-session FOO ethernet port-level
 !
!
!!ERSPAN Destination ip-tunnel00’s underlying interface, with egress policy-map shape-foo attached
interface TenGigE0/6/0/9/0
 service-policy output shape-foo
 ipv4 address 10.8.1.1 255.255.255.0
Verification
RP/0/RP0/CPU0:ios#show monitor-session FOO status detail   
Wed May  2 15:14:05.762 UTC
Monitor-session FOO
  Destination interface tunnel-ip100
  Source Interfaces
  -----------------
  TenGigE0/6/0/4/0
    Direction:  Both
    Port level: True
    ACL match:  Disabled
    Portion:    Full packet
    Interval:   Mirror all packets
    Status:     Operational
RP/0/RP0/CPU0:ios#
show monitor-session <sess-id> status internal

RP/0/RP0/CPU0:ios#show monitor-session FOO status internal 
Wed May  2 15:13:06.063 UTC
Information from SPAN Manager and MA on all nodes:
Monitor-session FOO (ID 0x00000001) (Ethernet)
SPAN Mgr: Destination interface tunnel-ip100 (0x0800001c)
          Last error: Success
          Tunnel data:
            Mode: GREoIPv4
            Source IP: 49.49.49.49
            Dest IP: 10.8.1.2
            VRF: 
            ToS: 0 (copied)
            TTL: 255
            DFbit: Not set
0/6/CPU0: Destination interface tunnel-ip100 (0x0800001c)
          Tunnel data:
            Mode: GREoIPv4
            Source IP: 49.49.49.49
            Dest IP: 10.8.1.2
            VRF: 
            ToS: 0 (copied)
            TTL: 255
            DFbit: Not set

Information from SPAN EA on all nodes:
Monitor-session 0x00000001 (Ethernet)
0/6/CPU0: Name 'FOO', destination interface tunnel-ip100 (0x0800001c)
Platform, 0/6/CPU0:

  Dest Port: 0xe7d

  ERSPAN Encap:
    Tunnel ID: 0x4001380b
    ERSPAN Tunnel ID: 0x4001380c
    IP-NH Grp key: 0x3140000cc5
    IP-NH hdl: 0x308a5fa5e0
    IP-NH IFH: 0x30002a0
    IP-NH IPAddr: 10.4.91.2

  NPU   MirrorRx     MirrorTx    
  00    0x00000003   0x00000004
  01    0x00000003   0x00000004
  02    0x00000003   0x00000004
  03    0x00000003   0x00000004
  04    0x00000003   0x00000004
  05    0x00000003   0x00000004
RP/0/RP0/CPU0:ios#

ERSPAN Traffic to a Destination Tunnel in a Default VRF

Table 3. Feature History Table

Feature Name

Release Information

Description

ERSPAN Traffic to a Destination Tunnel in a Default VRF

Release 6.1.3

Encapsulated Remote Switched Port Analyzer (ERSPAN) now transports mirrored traffic through GRE tunnels that belongs to the default VRF thus ensuring a network design with a single Layer 3 device.

This feature enables the tunnels to be grouped under the default VRF domain towards which you can segregate the traffic.

Running Configuration

The following example shows a tunnel interface configured with endpoints in a default VRF (vrf: green ): 

Router#show run int tunnel-ip 2
Thu Feb  3 06:18:28.075 UTC
interface tunnel-ip2
 ipv4 address 102.1.1.100 255.255.255.0
 tunnel tos 32
 tunnel mode gre ipv4
 tunnel source 120.1.1.100
 tunnel vrf green
 tunnel destination 120.1.1.1
Router#show monitor-session status
Thu Feb  3 06:18:11.061 UTC
Monitor-session ERSPAN-2
Destination interface tunnel-ip2
===============================================================
Source Interface      Dir   Status
--------------------- ----  ----------------------------------------------------
Te0/0/0/5 (port)      Rx    Operational

Verification

The following CLI output shows how to verify the default VRF configuration:

Router#show monitor-session ERSPAN-2 status internal
Thu Feb  3 06:19:50.014 UTC

Information from SPAN Manager and MA on all nodes:
Monitor-session ERSPAN-2 (ID 0x00000003) (Ethernet)
SPAN Mgr: Destination interface tunnel-ip2 (0x20008024)
          Last error: Success
          Tunnel data:
            Mode: GREoIPv4
            Source IP: 120.1.1.100
            Dest IP: 120.1.1.1
            VRF: green
            VRF TBL ID: 0
            ToS: 32
            TTL: 255
            DFbit: Not set

ERSPAN Traffic to a Destination Tunnel in a Non-Default VRF

Table 4. Feature History Table

Feature Name

Release Information

Description

ERSPAN Traffic to a Destination Tunnel in a Non-Default VRF

Release 7.5.3

The tunnels are grouped under the VRFs and you can segregate the traffic towards a specific VRF domain.

Encapsulated Remote Switched Port Analyzer (ERSPAN) now transports mirrored traffic through GRE tunnels with multiple VRFs, helping you design your network with multiple Layer 3 partitions.

In earlier releases, ERSPAN transported mirrored traffic through GRE tunnels that belonged to only default VRF.

Here, the tunnel interface, where the traffic mirroring is destined, is now in a VRF.

The traffic coming out of the interfaces of a router do not have any grouping. By configuring a specific VRF, you can now identify the incoming traffic group.

Configuration

Use the following command to configure a specific VRF:

RP/0/RP0/CPU0:router# configure 
RP/0/RP0/CPU0:router(config)# interface tunnel-ip 2
RP/0/RP0/CPU0:router(config)# tunnel vrf red

For more information on enabling the tunnel mode in GRE, see Configuring GRE Tunnels.

Configuration example

The following example shows a tunnel interface configured with endpoints in a non-default VRF (vrf: red ): 

Router#show run int tunnel-ip 2
Thu Feb  3 06:18:28.075 UTC
interface tunnel-ip2
 ipv4 address 102.1.1.100 255.255.255.0
 tunnel tos 32
 tunnel mode gre ipv4
 tunnel source 120.1.1.100
 tunnel vrf red
 tunnel destination 120.1.1.1
Router#show monitor-session status
Thu Feb  3 06:18:11.061 UTC
Monitor-session ERSPAN-2
Destination interface tunnel-ip2
===============================================================
Source Interface      Dir   Status
--------------------- ----  ----------------------------------------------------
Te0/0/0/5 (port)      Rx    Operational

Verification

The following CLI output shows how to verify, if the configured tunnel VRF is programmed in the session:

Router#show monitor-session ERSPAN-2 status internal
Thu Feb  3 06:19:50.014 UTC

Information from SPAN Manager and MA on all nodes:
Monitor-session ERSPAN-2 (ID 0x00000003) (Ethernet)
SPAN Mgr: Destination interface tunnel-ip2 (0x20008024)
          Last error: Success
          Tunnel data:
            Mode: GREoIPv4
            Source IP: 120.1.1.100
            Dest IP: 120.1.1.1
            VRF: red
            VRF TBL ID: 0
            ToS: 32
            TTL: 255
            DFbit: Not set

Troubleshoot Traffic Mirroring

When you encounter any issue with traffic mirroring, begin troubleshooting by checking the output of the show monitor-session status command. This command displays the recorded state of all sessions and source interfaces: 


# show monitor-session status
Monitor-session 5
rx destination interface tunnel-ip5
tx destination is not specified
================================================================================
Source Interface  Dir  Status
--------------------- ---- ----------------------------------------------------
Te0/0/0/23 (port) Rx   Operational

In the preceding example, the line marked as <Session status> can indicate one of these configuration errors:

Session Status

Explanation

Session is not configured globally

The session does not exist in global configuration. Review the show run command output and ensure that a session with a correct name has been configured.

Destination interface <intf> (<down-state>)

The destination interface is not in Up state in the Interface Manager. You can verify the state using the show interfaces command. Check the configuration to determine what might be keeping the interface from coming up (for example, a sub-interface needs to have an appropriate encapsulation configured).

The <Source interface status> can report these messages:

Source Interface Status

Explanation

Operational

Everything appears to be working correctly in traffic mirroring PI. Please follow up with the platform teams in the first instance, if mirroring is not operating as expected.

Not operational (Session is not configured globally)

The session does not exist in global configuration. Check the show run command output to ensure that a session with the right name has been configured.

Not operational (destination not known)

The session exists, but it either does not have a destination interface specified, or the destination interface named for the session does not exist. For example, if the destination is a sub-interface that has not been created.

Not operational (source same as destination)

The session exists, but the destination and source are the same interface, so traffic mirroring does not work.

Not operational (destination not active)

The destination interface is not in the Up state. See the corresponding Session status error messages for suggested resolution.

Not operational (source state <down-state>)

The source interface is not in the Up state. You can verify the state using the show interfaces command. Check the configuration to see what might be keeping the interface from coming up (for example, a sub-interface needs to have an appropriate encapsulation configured).

Error: see detailed output for explanation

Traffic mirroring has encountered an error. Run the show monitor-session status detail command to display more information.

The show monitor-session status detail command displays full details of the configuration parameters and any errors encountered. For example:

RP/0/RP0/CPU0:router show monitor-session status detail 


Monitor-session sess1
 Destination interface is not configured
 Source Interfaces
 -----------------
 TenGigE0/0/0/1
  Direction: Both
  ACL match: Disabled
  Portion:  Full packet
  Status:  Not operational (destination interface not known)
 TenGigE0/0/0/2
  Direction: Both
  ACL match: Disabled
  Portion:  First 100 bytes
  Status: Not operational (destination interface not known). Error: 'Viking SPAN PD' detected the 'warning' condition 'PRM connection 
          creation failure'.
Monitor-session foo
 Destination next-hop TenGigE 0/0/0/0
 Source Interfaces
 -----------------
 TenGigE 0/0/0/1.100:
  Direction: Both
  Status:  Operating
 TenGigE 0/0/0/2.200:
  Direction: Tx
  Status:  Error: <blah>

Monitor session bar
 No destination configured
 Source Interfaces
 -----------------
 TenGigE 0/0/0/3.100:
  Direction: Rx
  Status:  Not operational(no destination)

Here are additional trace and debug commands: 


RP/0/RP0/CPU0:router# show monitor-session trace ?

 platform  Enable platform trace
 process   Filter debug by process(cisco-support)

RP/0/RP0/CPU0:router# show monitor-session trace platform ?

 errors  Display error traces(cisco-support)
 events  Display event traces(cisco-support)

RP/0/RP0/CPU0:router#show monitor-session trace platform events location all ?

usrtdir  Specify directory to collect unsorted traces(cisco-support)
|        Output Modifiers
<cr>

RP/0/RP0/CPU0:router#show monitor-session trace platform errors location all ?

usrtdir  Specify directory to collect unsorted traces(cisco-support)
|        Output Modifiers
<cr>

#

 

RP/0/RP0/CPU0:router# debug monitor-session process all

RP/0/RP0/CPU0:router# debug monitor-session process ea

RP/0/RP0/CPU0:router# debug monitor-session process ma

RP/0/RP0/CPU0:router# show monitor-session process mgr

 detail  Display detailed output
 errors  Display only attachments which have errors
 internal Display internal monitor-session information
 |     Output Modifiers

RP/0/RP0/CPU0:router# show monitor-session status

RP/0/RP0/CPU0:router# show monitor-session status errors

RP/0/RP0/CPU0:router# show monitor-session status internal

RP/0/RP0/CPU0:router# show tech-support span ?

 file       Specify a valid file name (e.g. disk0:tmp.log)
 list-CLIs  list the commands that would be run (don't execute)(cisco-support)
 location   Specify a location(cisco-support)
 rack       Specify a rack(cisco-support)
 time-out   per show command timeout configuration(cisco-support)
 <cr>