BGP Best Path Selection

[bizhelp]

Can anyone help me figure out why the best path is shown as #3? I would expect it to be #1.

The reason I even looked at this in the first place was that one of our circuits went down and then came back online an hour later. I believe in the past #1 would have been chosen as the best path.

show ip bgp 72.32.79.195
BGP routing table entry for 72.32.0.0/16, version 23508337
Paths: (4 available, best #3, table Default-IP-Routing-Table)
Multipath: eBGP iBGP
Advertised to update-groups:
3
12232 3346 1299 33070 33070
111.123.191.153 from 111.123.191.153 (12.23.45.82)
Origin IGP, localpref 400, valid, external
DMZ-Link Bw 1250 kbytes
12232 3346 1299 33070 33070, (received-only)
111.123.191.153 from 111.123.191.153 (12.23.45.82)
Origin IGP, localpref 100, valid, external
DMZ-Link Bw 1250 kbytes
11122 3346 1299 33070 33070
22.11.33.225 from 22.11.33.225 (22.11.63.253)
Origin IGP, localpref 400, valid, external, best
DMZ-Link Bw 375 kbytes
11122 3346 1299 33070 33070, (received-only)
22.11.33.225 from 22.11.33.225 (22.11.63.253)
Origin IGP, localpref 100, valid, external
DMZ-Link Bw 375 kbytes

I receive similiar results with "show ip bgp" for other networks too.

 

[burtsbees]

Post a sh run and let's see all the parameters...weight, etc.

Burt

 

[bizhelp]

I'll have to grab the config later.

Shouldn't all the decision information be displayed in what I posted? That is, the results of the parameters applying weight, local pref, etc would be shown above yes no?

 

[burtsbees]

There are more than just local preference and weight...now you're gonna make me dust off my books...it's just that since I can't see anything pointing to why in the sh ip bgp
I would tend to agree with you, having the pref at 400 and the higher bandwidth also...

Burt

 

[unclerico]

What version of IOS are you running? The reason I ask is because if you're running an older version the show ip bgp command may not output the results you're looking for. Also, the results will vary depending on which router you run the commands from.

I know that one of the last criteria for placing bgp routes into the routing table is EBGP Peering. All things being equal it will prefer the oldest route.

I hate all Uppercase... I don't want my groups to seem angry at me all the time! =)
- ColdFlame (vbscript forum)

 

[bizhelp]

12.4. Unclerico - you are saying that in this case it is ignoring dmz-link bw? (I notice there is no presence of "multipath" except at the top..) Would a "clear ip bgp neighborIP in" and/or "clear ip bgp neighborIP out" do anything potentially usefull?

 

[burtsbees]

Here is something from another post...

Quote from the cisco site

In order to be candidates for multipath, paths to the same destination need to have these characteristics equal to the best-path characteristics:

Weight

Local preference

AS-PATH length

Origin

MED

Burt

 

[unclerico]

you are saying that in this case it is ignoring dmz-link bw?

In this case I believe that is what is happening. I think that the dmzlink-bw is being ignored since multipathing isn't truly functioning.

Would a "clear ip bgp neighborIP in" and/or "clear ip bgp neighborIP out" do anything potentially usefull?

Yes it will. If you issue a clear ip bgp neighborIP soft in then the routes will be re-evaluated.

BGP sure is fun isn't it? LOL

I hate all Uppercase... I don't want my groups to seem angry at me all the time! =)
- ColdFlame (vbscript forum)

 

[dialerstring]

Do a show ip bgp | include (type route here)

Then you can get a better look at the attributes. I believe there is 10 and sometime you will go all the way to the bottom of the list. I can't recall all, but the path will take the router id, closest next hop, path that originated locally (You can get the correct order of the net)
These can be adjusted depending if it's iBGP or ibgp.

Burt started the order off correctly but there are like 4 more.

CCNA
Network Admin