09-08-2017 08:34 AM - edited 03-08-2019 11:58 AM
Hi
I have nexus 7010 with 6.2(14) version. It is configured with BGP and VRFs like below
!
router bgp 1
vrf VRF_A
neighbor 2.2.2.1 remote-as 2
neighbor 2.2.2.2 remote-as 2
neighbor 2.2.2.3 remote-as 2
!
vrf VRF_B
neighbor 3.3.3.1 remote-as 3
neighbor 3.3.3.2 remote-as 3
neighbor 3.3.3.3 remote-as 3
!
When I shutdown BGP for specific neighbor under specific VRF it works well.
R1(config-router)# vrf VRF_A
R1(config-router-vrf)#neighbor 2.2.2.1 remote-as 2
R1(config-router-vrf)#shutdown
But when I issue a "shutdown" command under a particular VRF, it shuts down all the peerings under VRF_A (which is quite expected) but as well under VRF_B as well (which is not expected).
R1(config-router)# vrf VRF_A
R1(config-router-vrf)#shutdown
Has somebody encountered this issue? Or can you test this in the lab and share results. Just wondering if I am doing something wrong or there is some bug in the code. Thanks.
Regards
Bharat
Solved! Go to Solution.
09-08-2017 09:21 AM
Hi
Just found this on bug on Cisco website - https://bst.cloudapps.cisco.com/bugsearch/bug/CSCuz01405
This is NXOS BGP enhancement to change the behaviour when 'showdown' command is entered under BGP VRF.
The current design allows shutdown the entire BGP process when the command 'shutdown' was accidently entered under BGP VRF even though 'showdown' is not supported command under BGP VRF.
Some customer intentionally or accidently could type 'shutdown' under BGP VRF in their network operation and implementation.
When entering 'shotdown' under VRF, as this command is not found (not supported) in the current VRF mode, it's automatically finding matching command in router BGP then it executes 'showdown' which results BGP service outage/network down.
The same behaviour is not found from IOS but NXOS.
https://bst.cloudapps.cisco.com/bugsearch/bug/CSCuz01405
Regards
Bharat
09-08-2017 09:21 AM
Hi
Just found this on bug on Cisco website - https://bst.cloudapps.cisco.com/bugsearch/bug/CSCuz01405
This is NXOS BGP enhancement to change the behaviour when 'showdown' command is entered under BGP VRF.
The current design allows shutdown the entire BGP process when the command 'shutdown' was accidently entered under BGP VRF even though 'showdown' is not supported command under BGP VRF.
Some customer intentionally or accidently could type 'shutdown' under BGP VRF in their network operation and implementation.
When entering 'shotdown' under VRF, as this command is not found (not supported) in the current VRF mode, it's automatically finding matching command in router BGP then it executes 'showdown' which results BGP service outage/network down.
The same behaviour is not found from IOS but NXOS.
https://bst.cloudapps.cisco.com/bugsearch/bug/CSCuz01405
Regards
Bharat
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide