[f-nsp] IPv6 subnets flapping on Bigiron 4000 jetcore

[Jeroen Wunnink]

We have two Bigiron 4000 Jetcores running B2P08.0.01s firmware and we're 
experiencing very unstable IPv6 behaviour.
These Bigirons have several ve's with customer subnets on there in a 
dual stack config (/24 for IPv4 and /64 for IPv6) and OSPF as IGP 
between devices. (Also two XMR's, but they work just fine)

On at least one of of these Bigirons, one or more /64 IPv6 subnets 
frequently becomes unavailable, the /24 IPv4 subnet on the ve works 
fine, but IPv6 enabled servers are suddenly unable to ping the ve's IPv6 
address (which is the gateway) and the Bigiron is unable to ping the 
host's address over IPv6.

It's happening on the whole subnet, we have three servers with IPv6 
connectivity monitoring this and they all report the ve's IP becoming 
unreachable. Removing the ve's IPv6 address and adding it back solves it 
for a while, but after a day or so it starts acting up again.  Yet it's 
not all subnets at once, one time it's a customer subnet, at other times 
it's the /126 subnet between XMR and Bigiron

I've been going over the release notes for the Bigiron (since the latest 
release is 08.0.01w), but I can't find any IPv6 related fixes that match 
this.

Anyone else seeing odd IPv6 behaviour on Bigirons ? (yes we have future 
MLX upgrade plans, but for now these units will have to suffice :-)

-- 

Met vriendelijke groet,

Jeroen Wunnink,
EasyHosting B.V. Systeembeheerder
systeembeheer at easyhosting.nl

telefoon:+31 (035) 6285455              Postbus 48
fax: +31 (035) 6838242                  3755 ZG Eemnes

http://www.easyhosting.nl
http://www.easycolocate.nl