[Pdns-users] pdns : still dumping core on me.

[Etienne Ledoux]

Greetings,



snipped from messages:

Jan 13 04:02:28 ns1 pdns[61256]: Recursive query for remote 196.29.240.1 with 
internal id 141 was not answered by backend wi
thin timeout, reusing id
Jan 13 04:02:28 ns1 pdns[61256]: Recursive query for remote 196.41.3.252 with 
internal id 144 was not answered by backend wi
thin timeout, reusing id
Jan 13 04:02:29 ns1 /kernel: pid 61256 (pdns_server), uid 0: exited on signal 
11 (core dumped)
Jan 13 04:02:30 ns1 pdns[15557]: Our pdns instance (61256) exited after signal 
11
Jan 13 04:02:30 ns1 pdns[15557]: Dumped core
Jan 13 04:02:30 ns1 pdns[15557]: Respawning

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

It would do this three times and the fourth time it just says:

Jan 13 04:30:05 ns1 pdns[75021]: Recursive query for remote 196.3.164.40 with 
internal id 85 was not answered by backend wit
hin timeout, reusing id
Jan 13 04:30:05 ns1 pdns[75021]: Recursive query for remote 196.41.3.252 with 
internal id 172 was not answered by backend wi
thin timeout, reusing id
Jan 13 04:30:06 ns1 /kernel: pid 75021 (pdns_server), uid 0: exited on signal 
11 (core dumped)
Jan 13 04:30:07 ns1 /kernel: Limiting icmp unreach response from 236 to 200 
packets per second
Jan 13 04:30:08 ns1 /kernel: Limiting icmp unreach response from 226 to 200 
packets per second
Jan 13 04:30:09 ns1 /kernel: Limiting icmp unreach response from 226 to 200 
packets per second

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

I do get allot of queries that doesn't seem to get answered by my backend 
within the timeout. I assume this is my mysql backend taking to long ?

I did apply the mysql patch. I will probably try to reinstall this from 
scratch today. My other dns server running pdns on redhat 7.0 seems ok. It is 
only the pdns on my fbsd box that seems to be a bit unstable. But my fbsd box 
is also allot more busier then the redhat box.

I think my server is fairly busy. If I graph the queries it seems like i'm 
averaging between 100-200 queries a second depending on the time of day.

So my problems so far seem to be the mysql backend not answering quick enough, 
which might be why pdns is crashing ? Perhaps I didn't apply the patch 
correctly that is why I will try reinstalling again today. In the meantime if 
there is anything else I should/could fix. Any tips would be appreciated. 
Maybe my mysql server is setup to handle the load ? How bout some heavy load 
tips. I've done all the settings for cache-ttl etc. which is recommended for 
heavy load but perhaps I missed something ?

Thanks

e.