Re: Nameserver messages on tor's logfile

[Olaf Selke <olaf.selke@xxxxxxxxxxxx>]

Hans de Hartog schrieb:
> 
> Tor's logfile irregulary receives the following messages:
> 
> May 01 20:38:14.542 [warn] eventdns: All nameservers have failed
> May 01 20:38:15.034 [notice] eventdns: Nameserver 127.0.0.1 is back up
> May 01 20:52:01.832 [warn] eventdns: All nameservers have failed
> May 01 20:52:02.942 [notice] eventdns: Nameserver 127.0.0.1 is back up
> May 02 09:23:49.074 [warn] eventdns: All nameservers have failed
> May 02 09:23:49.503 [notice] eventdns: Nameserver 127.0.0.1 is back up
> May 02 12:52:15.519 [warn] eventdns: All nameservers have failed
> May 02 12:52:25.533 [notice] eventdns: Nameserver 127.0.0.1 is back up
> May 02 12:53:52.209 [warn] eventdns: All nameservers have failed
> May 02 12:53:52.296 [notice] eventdns: Nameserver 127.0.0.1 is back up
> 
> I'm running tor 0.2.0.34 on FreeBSD 7.1-RELEASE and FWIW, I'm also
> running the standard nameserver locally: BIND 9.4.2-P2.
> It started after being up (as exit-relay) for about 8 days and after
> that, I see
> those messages several times per day. My libevent-version is 1.4.10
> Why is this happening? I don't see anything logged by the nameserver so
> it must be libevent who thinks that the nameserver fails.

I've seen this with all Linux versions within the last two years. No
matter if 32 or 64 bit code. I don't believe my nameservers being
unreachable. And using pdnsd as nameserver tor constantly core dumps.

Olaf