Cisco Support Community
cancel
Showing results for 
Search instead for 
Did you mean: 
Announcements

Welcome to Cisco Support Community. We would love to have your feedback.

For an introduction to the new site, click here. And see here for current known issues.

New Member

eBGP multihop and dual ISP redundancy

Hi all,

I have to implement a bgp-multihomed, in the beginning my routers were going to replaced the CE, but now the ISP's are saying no, we have to have the CE, which they managed.  I have my own public ASN and /24 public address so I need to control my internal routing.   So, the only option I have to implement bgp-dualhomed is to use eBGP multihop to their PE and the CE would be there just to provided access to their PE - has anyone implemented this before? are there any issues to consider, is there a nice how to link (I haven't found anything good yet), I would really appreciated it.

Thanks in advanced.

Cheers!

1 ACCEPTED SOLUTION

Accepted Solutions
Hall of Fame Super Silver

Re: eBGP multihop and dual ISP redundancy

Hello Mguzman4158,

I tested a similar scenario some years ago.

routers on the path (CE) have to be able to route appropriately BGP messages and user traffic by means of static routes or dynamic routing

it is possible to have a session with both the CE and the PE in the path with eBGP session with CE using BGP community no-advertise for your IP subnet but allowing to advertise the subnet in between.

on the session with the PE node you advertise your network

using static routes is probably easier

of course you need to use  ebgp-multihop option for session to PE node.

with modern IOS you can even use IP SLA and tracking objects to check PE address reachability in this way the route to PE address will disappear when the tracked object fails ( ip reachability of PE address used as BGP neighbor)

in this way you can achieve faster convergence in case of failure you just need to avoid to have a backup static route as your objective is to declare down the eBGP multihop session as soon as you detect the BGP endpoint does not answer anymore.

As the route to BGP nexthop disappear BGP next-hop tracking can declare down the session.

Again this a feature present in recent IOS versions.

In this case recovery means to use the iBGP session between your routers to send traffic to your other router that has an alive multihop eBGP session

You need to coordinate with SP in order to ensure they are doing the same.

see

http://www.cisco.com/en/US/docs/ios/12_3/12_3x/12_3xe/feature/guide/dbackupx.html

for BGP next hop tracking:

http://www.cisco.com/en/US/docs/ios/12_3t/12_3t14/feature/guide/gt_bnht.html

Hope to help

Giuseppe

1 REPLY
Hall of Fame Super Silver

Re: eBGP multihop and dual ISP redundancy

Hello Mguzman4158,

I tested a similar scenario some years ago.

routers on the path (CE) have to be able to route appropriately BGP messages and user traffic by means of static routes or dynamic routing

it is possible to have a session with both the CE and the PE in the path with eBGP session with CE using BGP community no-advertise for your IP subnet but allowing to advertise the subnet in between.

on the session with the PE node you advertise your network

using static routes is probably easier

of course you need to use  ebgp-multihop option for session to PE node.

with modern IOS you can even use IP SLA and tracking objects to check PE address reachability in this way the route to PE address will disappear when the tracked object fails ( ip reachability of PE address used as BGP neighbor)

in this way you can achieve faster convergence in case of failure you just need to avoid to have a backup static route as your objective is to declare down the eBGP multihop session as soon as you detect the BGP endpoint does not answer anymore.

As the route to BGP nexthop disappear BGP next-hop tracking can declare down the session.

Again this a feature present in recent IOS versions.

In this case recovery means to use the iBGP session between your routers to send traffic to your other router that has an alive multihop eBGP session

You need to coordinate with SP in order to ensure they are doing the same.

see

http://www.cisco.com/en/US/docs/ios/12_3/12_3x/12_3xe/feature/guide/dbackupx.html

for BGP next hop tracking:

http://www.cisco.com/en/US/docs/ios/12_3t/12_3t14/feature/guide/gt_bnht.html

Hope to help

Giuseppe

615
Views
0
Helpful
1
Replies