IPv6 design, deployment, standards, and best practices.
New Member
Sat Mar 21, 2015 1:21 am

Splitting up an ipv6 block between peers

Sat Mar 21, 2015 1:36 am

Hi everyone, this is my first post here. I wanted to join this forum and do a little crowdsourcing for an issue I'm having regarding some transit providers and the ipv6 block that they're advertising to us.

So one of our transit providers is sending us an ipv6 /24 over a 10G link that we have, but unfortunately I would guess that this link becomes oversubscribed by almost 25% on the outbound during peak times.

In an attempt to do some traffic engineering, I subnetted the block 2601:300::/24 (not the real address) into 2601:300::/25 & 2601:380::/25.

Then I created a static route to point the second block to a different transit provider like this ipv6 route 2601:380::/25 2001:::6f:1.

I added that prefix to the list to allow redistribution ipv6 prefix-list ipv6-static-into-bgp permit 2601:380::/25

It seemed to work partially, the interface in which 2001:::6f:1 was the next-hop increased in traffic by roughly 40%, however the other half of the traffic, I'm assuming what would normally have been the 2601:300::/25 portion was gone, presumably routing elsewhere. However, a few hours later I got an email because a phone call came into our NOC from a notoriously large telco complaining that we were highjacking their routes and advertising them out. I now know that if I do this again, I need to ensure that these static-into-bgp static routes don't get advertised to any peers.

What I was wondering if there is a *better* way to do this. I can't see that originating somebody elses routes and hacking them up in your own network, then trying not to leak them is the best way to go about this. Anybody with experience with this care to help out?

Return to IPv6

Who is online

Users browsing this forum: No registered users and 8 guests