Solved: Re: ASR Router with BGP Neighbours and Failover

mahesh18 · ‎05-08-2019

We have ASR Routers with 3 BGP neighbour's which are 3 different ISP's.

We are learning routes from all of them

From ISP 1 we are learning full Internet Routing table.

IF we lose physical link to ISP 1 will our Router learn all the routes from ISP 2 and 3?

We do not have any link monitoring configured on physical links connecting to the 3 ISP's

Giuseppe Larosa · ‎05-09-2019

Hello Mahesh,

just look at the output of show run | begin router bgp

if you see commands that sets local preference per neighbor or any route-map applied in direction in per neighbor some form of manipulation is configured otherwise simply BGP uses the default order of criteria to choice the best path.

Hope to help

Giuseppe

View solution in original post

Francesco Molino · ‎05-09-2019

Based on your previous output, there is no local preference or whatsoever. The route chosen is based on as-path.

Thanks
Francesco
PS: Please don't forget to rate and select as validated answer if this answered your question

View solution in original post

Giuseppe Larosa · ‎05-10-2019

Hello Mahesh,

the route-maps are not applied to any neighbor so no type of route manipulation is occurring in this moment in your router.

You have a route-map that could be used to increase the local-preference to 150 for all received routes if you add a command like

neighbor <IP-address-ebGP-neighbor> route-map set-local-pre in

under router bgp or router bgp address-family ipv4-unicast

The use of a command like the one proposed above should be considered only for the eBGP neighbors that are sending you partial BGP routes like ISP2 and ISP3.

As I have written in another thread I would give local preference 200 to ISP3 that gives you a small number of routes, then local preference 150 to ISP2 that sends you more routes. You can leave ISP1 and the iBGP neighbor with default settings of local-preference 100.

However, you need to consider the current link usage of links towards ISP3, ISP2 and ISP1 and their effective speed that can be different. Likely ISP1 has a link with higher bandwidth.

An higher local-preference can make a prefix with a longer AS path preferred because local preference is considered before AS path length in BGP best path selection.

At the moment as noted by Francesco no form of route manipulation is occurring in your router.

Hope to help

Giuseppe

View solution in original post

Francesco Molino · ‎05-08-2019

Hi

You said you have 3 bgp peering with 3 ISPs.
From 1 you're learning the full internet table. What about with the 2 others?

The routing table will install prefixes learned from the 2 others when isp1 will be down but depending on what you're learning you may have only a default route.

How is your bgp configuration? Do you have done mechanisms to prioritize prefixes over isp1 ?
Can you share the configuration and output of sh ip bgp summary (if your ISPs are in a front vrf, adapt the config to check the specific vrf)

Thanks
Francesco
PS: Please don't forget to rate and select as validated answer if this answered your question

mahesh18 · ‎05-08-2019

here is output

show ip bgp summary
BGP router identifier 72.29.230.181, local AS number 16569
BGP table version is 25827281, main routing table version 25827281
747509 network entries using 107641296 bytes of memory
1491087 path entries using 119286960 bytes of memory
232055/116395 BGP path/bestpath attribute entries using 35272360 bytes of memory
202832 BGP AS-PATH entries using 8603774 bytes of memory
1038 BGP community entries using 47350 bytes of memory
95 BGP extended community entries using 2556 bytes of memory
0 BGP route-map cache entries using 0 bytes of memory
103 BGP filter-list cache entries using 1648 bytes of memory
BGP using 270855944 total bytes of memory
722 received paths for inbound soft reconfiguration
BGP activity 1905580/1158047 prefixes, 5019660/3528573 paths, scan interval 60 s ecs

Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down State /PfxRcd
72.29.230.182 4 25983 7054589 67060 25827252 0 0 3w0d 745 300
192.41.x.x 4 16569 3656685 6134122 25827281 0 0 3w0d 723 291
199.116.233.104 4 15296 248994 67057 25827252 0 0 3w0d 17 821
Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down State/PfxRcd
199.116.234.90 4 15296 90783 67053 25827252 0 0 3w0d 3951

show ip route
Codes: L - local, C - connected, S - static, R - RIP, M - mobile, B - BGP
D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
E1 - OSPF external type 1, E2 - OSPF external type 2
i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2
ia - IS-IS inter area, * - candidate default, U - per-user static route
o - ODR, P - periodic downloaded static route, H - NHRP, l - LISP
a - application route
+ - replicated route, % - next hop override

Gateway of last resort is 72.29.230.182 to network 0.0.0.0

Francesco Molino · ‎05-08-2019

Based on your show ip bgp summary output, there are 4 peerings:

1 —> 72.29.230.182 4 25983 7054589 67060 25827252 0 0 3w0d 745 300
2 —> 192.41.x.x 4 16569 3656685 6134122 25827281 0 0 3w0d 723 291
3 —> 199.116.233.104 4 15296 248994 67057 25827252 0 0 3w0d 17 821
4 —> 199.116.234.90 4 15296 90783 67053 25827252 0 0 3w0d 3951

for peering 1 and 2, you’re receiving the full table. However with peer 3 and 4 you’re receiving some prefixes but not the full internet bgp table.

Your show ip route is empty which is weird because you should get something in.

Can you share the output of show run | s r b please?

Also rerun the show ip route to validate you have something in.

Thanks
Francesco
PS: Please don't forget to rate and select as validated answer if this answered your question

mahesh18 · ‎05-08-2019

1# show run | s r b
1#

we have routes in output of show ip route

they are all BGP and full routing table

mahesh18 · ‎05-08-2019

Also if we lose link to ISP 1 72.29.230.182

we can learn routes from second ISP 192.x.x.x?

Francesco Molino · ‎05-08-2019

I don’t know if you have copy/paste issues but don’t see your router config.

Anyways, based on prefixes you’re learning, yes routes from 192.x.x.x ISPs should be installed in your RIB.

But if you didn’t managed which ISP must be prioritized, maybe you already have some prefixes from ISP2 already installed into your RIB. ISP 192.x.x.x has the same AS as your own router which means you’re running iBGP whereas ISP1 is in ebgp.

Thanks
Francesco
PS: Please don't forget to rate and select as validated answer if this answered your question

mahesh18 · ‎05-08-2019

here is output

show ip route
Codes: L - local, C - connected, S - static, R - RIP, M - mobile, B - BGP
D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
E1 - OSPF external type 1, E2 - OSPF external type 2
i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2
ia - IS-IS inter area, * - candidate default, U - per-user static route
o - ODR, P - periodic downloaded static route, H - NHRP, l - LISP
a - application route
+ - replicated route, % - next hop override

Gateway of last resort is 72.29.230.182 to network 0.0.0.0

B* 0.0.0.0/0 [20/0] via 72.29.230.182, 3w0d
1.0.0.0/8 is variably subnetted, 2660 subnets, 14 masks
B 1.0.0.0/24 [200/0] via 192., 2w1d
B 1.0.4.0/22 [200/0] via 192.x.x.x, 3w0d

B 1.0.128.0/19 [200/1710] via 192.x.x.x, 3w0d
B 1.0.128.0/24 [200/0] via 192.x.x.x2w4d
B 1.0.129.0/24 [200/430] via 192.xxx, 3w0d
B 1.0.131.0/24 [200/0] via 192.x, 2w2d

Seems it is preferring IBGP route from 192.x.x.x?

any reason why it is doing that?

Francesco Molino · ‎05-08-2019

can you share output of:
sh ip bgp 1.0.131.0/24

Thanks
Francesco
PS: Please don't forget to rate and select as validated answer if this answered your question

mahesh18 · ‎05-08-2019

sh ip bgp 1.0.131.0/24
BGP routing table entry for 1.0.131.0/24, version 6538875
Paths: (2 available, best #2, table default)
Not advertised to any peer
Refresh Epoch 1
25983 6327 3491 38040 23969
72.29.230.182 from 72.29.230.182 (72.29.224.129)
Origin incomplete, localpref 100, valid, external
Community: 6327:2001 6327:30999
rx pathid: 0, tx pathid: 0
Refresh Epoch 1
6327 3491 38040 23969
192.x.x.x from x.x.x.x (64.141.118.154)
Origin incomplete, metric 0, localpref 100, valid, internal, best
rx pathid: 0, tx pathid: 0x0

Francesco Molino · ‎05-08-2019

We can see the prefixe learned from 72.29.230.182 has a longer as-path which is why it prefers the ibgp route.

Thanks
Francesco
PS: Please don't forget to rate and select as validated answer if this answered your question

mahesh18 · ‎05-08-2019

can you please tell me which config points to longer as-path?

BGP prefers path with shorest AS path as per my understanding

Francesco Molino · ‎05-09-2019

Yes BGP prefers a shortest AS-PATH that's why you see routes into your RIB pointing to 192.x.x.x ISP.

sh ip bgp 1.0.131.0/24
BGP routing table entry for 1.0.131.0/24, version 6538875
Paths: (2 available, best #2, table default)
Not advertised to any peer
Refresh Epoch 1
25983 6327 3491 38040 23969
72.29.230.182 from 72.29.230.182 (72.29.224.129)
Origin incomplete, localpref 100, valid, external
Community: 6327:2001 6327:30999
rx pathid: 0, tx pathid: 0
Refresh Epoch 1
6327 3491 38040 23969
192.x.x.x from x.x.x.x (64.141.118.154)
Origin incomplete, metric 0, localpref 100, valid, internal, best
rx pathid: 0, tx pathid: 0x0

Based on your output you can see that learned prefix from 192.x.x.x has a shortest as-path.

Thanks
Francesco
PS: Please don't forget to rate and select as validated answer if this answered your question

paul driver · ‎05-09-2019

Hello

Just like to add I doubt very much you are receiving full bgp internet table from all peers after bgp has fully converged.

@mahesh18 wrote:
show ip bgp summary

BGP router identifier 72.29.230.181, local AS number 16569
BGP table version is 25827281, main routing table version 25827281
747509 network entries using 107641296 bytes of memory
1491087 path entries using 119286960 bytes of memory
232055/116395 BGP path/bestpath attribute entries using 35272360 bytes of memory
202832 BGP AS-PATH entries using 8603774 bytes of memory
1038 BGP community entries using 47350 bytes of memory
95 BGP extended community entries using 2556 bytes of memory
0 BGP route-map cache entries using 0 bytes of memory
103 BGP filter-list cache entries using 1648 bytes of memory
BGP using 270855944 total bytes of memory
722 received paths for inbound soft reconfiguration
BGP activity 1905580/1158047 prefixes, 5019660/3528573 paths, scan interval 60 s ecs

Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down State /PfxRcd
72.29.230.182 4 25983 7054589 67060 25827252 0 0 3w0d 745300
192.41.x.x 4 16569 3656685 6134122 25827281 0 0 3w0d 723291
199.116.233.104 4 15296 248994 67057 25827252 0 0 3w0d 17821
199.116.234.90 4 15296 90783 67053 25827252 0 0 3w0d 3951

Please rate and mark as an accepted solution if you have found any of the information provided useful.
This then could assist others on these forums to find a valuable answer and broadens the community’s global network.

Kind Regards
Paul

mahesh18 · ‎05-10-2019

Just saw your note now yes we are receiving full routing table from the ISP.