Hi,

I have eBGP running in my ASR router. Few weeks ago I used to get only my network when I give command "sh ip bgp nei x.x.x.x advertised-routes". But recently I get global routing table when I give this command. Will this cause any kind of problem or it's fine.

How can I get my router to show only my advertised network in above command.

Thanks in advance.

Are you using IOS-XR on your ASR? If so I am not sure how the command is working as you don't need "ip" in the syntax of the command. Like all route commands on IOS-XR it is all IP-based so you can leave it out. For example, "show bgp neigh x.x.x.x {etc., etc.}"

At any rate the "advertised-routes" option shows the routes the router is advertising to the BGP peer you are specifying. From what you are saying, I am assuming that some how you are now redistributing other routes into BGP. This is a problem if you don't want to advertise those networks out to your peer. In most situations what causes unexpected advertised routes is a change in a route-map/prefix list/access-list for IOS or in the case of IOS-XR is a route-policy. "Show rpl route-policy {policy name} detail" would be a good place to start. At any rate, if you are just wanting to verify that you are advertising a specific route to a BGP peer than I would just simply do a pipe include, i.e. show bgp neigh 1.2.3.4 advertised-routes | i 4.3.2.1

I hope that helps

_________________
Age si quid agis. Age quod agis. - If you do something, do it well. Do well in whatever you do.

http://www.zingo-net.com/
http://twitter.com/zingonet

Sounds like you're sending the wrong shit.

_________________
http://blog.alwaysthenetwork.com

Router came with IOS-XE.

Do you mean wrong configuration of route-map/access-list can cause this issue?
i checked in router and found that few access-lists were added in route-map (match ip address) even though those access-list are not configured in router. How can I stop redistributing global network ?

You've become a transit AS. Well done. I'm surprised your peers are accepting these routes and are happy to send all traffic your way...

Either you need to have a route-map that calls an as-path access list which only stop the advertisement of any prefix going through another AS, or you set the no-export community on all routes received from your peers. Or you could do both

_________________
www.mellowd.co.uk/ccie/

issue resolved. As suggested by Zingonet, rechecked all access-lists and route-maps and corrected them. now i can see only my network in "sh ip bgp nei x.x.x.x advertised-routes'

thanx for help.

I doubt they were, doesn't stop him from sending though.

_________________
http://blog.alwaysthenetwork.com

I've had this both ways. I have one upstream that drops my connection upon hitting a prefix limit, and another that takes them but does nothing with them.


Sent from my ADR6400L using Tapatalk 2

I'd expect any non-shitty service provider to only allow ^CUSTOMER_ASN$ inbound, otherwise, they have no business running BGP either. Of course, there would still be overhead of processing all those updates and storing them needlessly as "received-only" in memory.

Secondly, OP, I have route-maps/prefix-lists/communities/local-prefs defined for all of my eBGP peers (~50). It's worth the extra maintenance -- I would never send everything I had or even anything that I locally originated in BGP without a very specific matching policy. Lazy networks are often the ones that break most.

That's gangster eloquence.

EDIT - I think I'd love working for Vito, at least until I was fired two weeks later for incompetence.

I get the feeling he's the kinda guy that would make you feel small, but you couldn't stop cracking up at what he says.

_________________
Regards,

Steven King
San Diego Cisco User Group - http://www.sdcug.com
"The only time something is impossible is when you think it is." - Kevin Corbin, CCIE #11577

I've worked with BGP for too long to know that you don't trust anyone

_________________
www.mellowd.co.uk/ccie/

This is very true. I can think of several times I could have been burned by not filtering.