05-22-2016 03:05 AM - edited 03-05-2019 04:04 AM
Hello all
what this looks like
R3(config-router-af)#neighbor 10.10.10.10 soft-reconfiguration inbound
% Invalid command for VPNv4 neighbor. Use route refresh instead
thanks all for ur reply
Solved! Go to Solution.
05-22-2016 01:33 PM
Hello
When you apply that command the router stores an unmodified copy of all its peers routes in its adi-rib-out table so its can readily check this cache against any inbound route policy changes you make to the router, which I guest isn't very often?
However to store this cache all the time and not make any route or filtering changes, the router still has to eat up unnecessary cpu/memory resource to accommodate these unprocessed routes in readiness.
The RR is a bgp capability and it applied by default ( I don't know when it came available but I am guessing its been around for some time)
Anyway if this feature is supported by both peers it will negate you having to enable Soft-reconfiguration and thus save valuable cpu/memory resources if you are receiving large amounts of prefixes as it will advertised its RR capable and thus on route change it will request a re-advertisement of the adi-rib-out table.
Thus any network I come across which has Soft-reconfiguration Inbound enabled I check the peering capability and if RR is applicable to both sides I disable the above command.( I have found usually 9/10 its a ghost command copied/edited and the pasted into the new rtr from another router and really isn't required)
Anyway you check if your bgp peering's supports Route refresh:
sh ip bgp neighbors | s Neighbor capabilities
Any capabilities "received" are supported and currently negotiated by the neighbour, The route refresh should be supported and if this is the case the soft-reconfiguration inbound command I would say can be removed or not applied at all.
Now if you have conflicting capabilities, Be it that either router does not support, Then you can be negate that bgp capability check by the enabling the following hidden command:
neighbor x.x.x.x dont-capability-negotiate
Lastly I don't think Soft-reconfiguration is allowed under vpnv4 peering's (your post suggests this) as I guest it would favour RR instead.( but I could be wrong)
So you could try disabling the capability check and then apply soft-reconfiguration inbound to see if M-bgp allows it.
res
Paul
05-22-2016 01:33 PM
Hello
When you apply that command the router stores an unmodified copy of all its peers routes in its adi-rib-out table so its can readily check this cache against any inbound route policy changes you make to the router, which I guest isn't very often?
However to store this cache all the time and not make any route or filtering changes, the router still has to eat up unnecessary cpu/memory resource to accommodate these unprocessed routes in readiness.
The RR is a bgp capability and it applied by default ( I don't know when it came available but I am guessing its been around for some time)
Anyway if this feature is supported by both peers it will negate you having to enable Soft-reconfiguration and thus save valuable cpu/memory resources if you are receiving large amounts of prefixes as it will advertised its RR capable and thus on route change it will request a re-advertisement of the adi-rib-out table.
Thus any network I come across which has Soft-reconfiguration Inbound enabled I check the peering capability and if RR is applicable to both sides I disable the above command.( I have found usually 9/10 its a ghost command copied/edited and the pasted into the new rtr from another router and really isn't required)
Anyway you check if your bgp peering's supports Route refresh:
sh ip bgp neighbors | s Neighbor capabilities
Any capabilities "received" are supported and currently negotiated by the neighbour, The route refresh should be supported and if this is the case the soft-reconfiguration inbound command I would say can be removed or not applied at all.
Now if you have conflicting capabilities, Be it that either router does not support, Then you can be negate that bgp capability check by the enabling the following hidden command:
neighbor x.x.x.x dont-capability-negotiate
Lastly I don't think Soft-reconfiguration is allowed under vpnv4 peering's (your post suggests this) as I guest it would favour RR instead.( but I could be wrong)
So you could try disabling the capability check and then apply soft-reconfiguration inbound to see if M-bgp allows it.
res
Paul
05-23-2016 05:34 AM
thanks Paul , for ur good explanation
09-29-2023 11:19 PM
Hi!
You can use show ip bgp vpnv6 unicast all neighbors XXX routes to see what vpnv6 routes are received, and show ip bgp vpnv4 all neighbors XXX routes to see what vpnv4 routes are received, these commands could achieve what you want.
Hope that can still help you after 7 years.
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