[rancid] Re: Frequent unknown changes against Quagga

[john heasley]

Fri, Mar 19, 2010 at 10:23:49AM -0400, Steve Bertrand:
> Hi everyone,
> 
> I've got an annoying issue using RANCID against Quagga. I poll every
> half-hour, and each day, there are two or three updates that randomly
> appear like this:
> 
> diff -u -4 -r1.105 hostname
> @@ -61,8 +61,10 @@
>    ip address 208.70.x.x/30
>    ipv6 nd suppress-ra
>    ipv6 ospf6 cost 1
>    ipv6 ospf6 dead-interval 40
> + exit
> + exit
>    ipv6 ospf6 hello-interval 10
>    ipv6 ospf6 instance-id 0
>    ipv6 ospf6 priority 1
>    ipv6 ospf6 retransmit-interval 5
> 
> ...on the next poll, the 'exit' lines are removed. It is inconsistent
> where in the config these exit entries appear. Nobody has logged into
> the router in question.
> 
> My .clogin rc looks like such:
> 
> add method hostname ssh
> add autoenable hostname 1
> add user hostname rancid
> add password hostname blah
> 
> ...and the router.db
> 
> hostname:zebra:up
> 
> The SSH user is bound directly to vtysh as its shell.
> 
> Is there anyone who might be able to give me some clue as to why these
> config changes happen?

its probably quagga doing something odd.  i run it against quagga regularly
w/o issue, but dont use vtysh.  try to replicate it with clogin -c 'list
of commands from zrancid'.  if you can cause it regularly, do it again
with -d and collect the output.