bgp route-map question

[Mike Schooley]

Ok, I have a route-map on my bgp peer for advertising out routes, it references an access list.  I added a route to the access-list, what do I have to do to re read the access-list.  Reset the peer, will a soft reset work, remove and reapply the route-map ?

[Mike Schooley]

actually need to modify this, my access-list is on a redistribute eigrp to bgp command.  I've added the route and its not being redist

[letsgomets]

Is the BGP network command configured to include this new subnet?

router bgp 65000

network 10.1.1.0 mask 255.255.255.0

You could also use an aggregate command

aggregate-address 10.0.0.0 255.0.0.0 summary-only

depends on how you wish to advertise this network.

It is important that your network statement matches the mask bit mark exactly or it will ignore it.

[Mike Schooley]

not the question, the redistribution is currently working, with a redistr eigrp 136 route-map eigrp_to_bgp.  I added another route to the eigrp_to_bgp access list, now its not being read to redist, what needs to be done to make it start using the new access-list

[letsgomets]

Mike,

I understand your redistribution is working. The fact that you mentioned you added another route, which I am sure you mean you added another prefix to the ACL then it is possible BGP is not aware of the prefix. Which is why you are not seeing being advertised via BGP.

BGP will not advertise a network, regardless of redistribution policies if it doesn't have the applicable network statement configured.

If you are using a prefix list to determine which routes to redistribute then you should not need to do anything after modifying the list.

However if you want to try something you can do a soft clear of BGP

clear ip bgp 1.1.1.1 soft

This will not affect your bgp session at all.

If this does not fix your problem I would need to see the configuration to help troubleshoot.

[milan.kulik]

Hi,

I don't agree with

"BGP will not advertise a network, regardless of redistribution policies if it doesn't have the applicable network statement configured."

If the EIGRP prefix is the best one (copied into the RIB table, i.e., the "sh ip route" command shows "D" or "EX" in the beginning of the line assigned to the prefix), EIGRP will redistribute it to BGP and BGP will advertise it to the BGP neighbors.

BR,

Milan

[Mike Schooley]

thats correct milan as i have no bgp network statements, do it strictly by redistribut eigrp and the route-map

router bgp 257
no synchronization
bgp log-neighbor-changes
redistribute eigrp 136 route-map eigrp_to_bgp
neighbor 10.33.254.3 remote-as 1
neighbor 10.33.254.3 soft-reconfiguration inbound
neighbor 10.33.254.4 remote-as 1
neighbor 10.33.254.4 soft-reconfiguration inbound
neighbor 10.63.255.1 remote-as 257
neighbor 10.63.255.1 next-hop-self
neighbor 10.63.255.1 soft-reconfiguration inbound

no auto-summary

[letsgomets]

Thank you Milan for correcting me. I have learned something today.

[milan.kulik]

Hi,

IMHO, you need to remove the EIGRP redistribution to BGP and enable it again.

I.e.,

router bgp xy

no redistribute eigrp 136 route-map eigrp_to_bgp

redistribute eigrp 136 route-map eigrp_to_bgp

This can make a short outage in your network as the EIGRP prefixes might be withdrawn from the BGP for a short time, but I'm afraid there's no "softer" way.

("clear ip bgp * soft" or "clear ip eigrp 136 nei  soft" might be tried, but I'm afraid won't work.)

BR,

Milan

[Mike Schooley]

yeah thats what I was afraid you would say, guess I will try it at midnight tonite, pretty sure that will fix

it was just curious if there was another way, tried doing a clear peer soft, that didnt work, a

hard clearing of peers might work to, but either removing the redistr probably quickest and less obtrusive of the two.

[letsgomets]

Mike,

If you can send your scrubbed configuration it may help.

Adding a prefix to your redist list should be seamless.