Cisco Support Community
Community Member

PBR on ME3800x

Hello All ,

               I was trying to do this simple PBR on an ME3800 switch "15.2(2)S1" .


R2#sh route-map test

route-map test, permit, sequence 10

  Match clauses:

    ip address (access-lists): 10

  Set clauses:

    vrf VPN_A

  Policy routing matches: 99 packets, 9990 bytes

R2#sh ip access-lists 10

Standard IP access list 10

    10 permit (278 matches)

R2#sh run int te0/2

Building configuration...

Current configuration : 179 bytes


interface TenGigabitEthernet0/2

no switchport

mtu 9216

ip vrf receive VPN_A

ip address

ip policy route-map test

ipv6 address 2001:10:1:2::2/64


R2#debug ip policy dynamic

Dynamic PBR debugging is on




*Sep 16 20:31:40.841: IP: s= (TenGigabitEthernet0/2), d=, len 100, policy match

*Sep 16 20:31:40.841: IP: route map test, item 10, permit

*Sep 16 20:31:40.841: IP: s= (TenGigabitEthernet0/2), d=, len 100, set vrf

*Sep 16 20:31:40.841: IP: set vrf tableid 2

*Sep 16 20:31:40.841: IP: s= (TenGigabitEthernet0/2), d= (Vlan24), len 100, policy routed

*Sep 16 20:31:40.841: IP: TenGigabitEthernet0/2 to Vlan24


As you can packetes with the source of get matched , but the ping are fails . When I checked futher on the next hop router I have found that the packet gets normally routed and not tagged at all !!!.


*Sep 16 20:31:23.505: IP: s= (Vlan24), d=, len 100, input feature, MCI Check(85), rtype 0, forus FALSE, sendself FALSE, mtu 0, fwdchk FALSE

*Sep 16 20:31:23.505: IP: tableid=0, s= (Vlan24), d= (GigabitE

R4#thernet0), routed via FIB

*Sep 16 20:31:23.505: IP: s= (Vlan24), d= (GigabitEthernet0), len 100, dropped due to input interface drop feature


I have checked and found that this feautre is not supported on the ME3800x

Can any body confirm that PBR for VRF does not work correctly ?

Everyone's tags (3)
Community Member

PBR on ME3800x

Hello George,

On both platforms me3600/3800 neither VRF source selection nor PBR with "set vrf" are supported. So you are correct there is a limitation on these platforms.

Thank you

Best regards,
Dmitry Skotnikov

-- Best regards, Dmitry Skotnikov
CreatePlease to create content