From: Niall O Broin (niall at domain linux.ie)
Date: Wed 09 May 2001 - 12:00:48 IST
I've just upgraded the distribution on my main machine to SuSE 7.0 (I know,
I know - I've more to be doing than living on the bleeding edge) and squid
is dying. Whenever I tried to access a URL by name in Netscape, I got a
"Document contains no data." alert and there was an entry like
2001/05/09 09:52:40| assertion failed: dns_internal.c:212: "nns > 0"
in cache.log, immediately followed by
2001/05/09 09:52:43| Starting Squid Cache version 2.3.STABLE3 . . .
which I presume means that Squid has restarted.
My resolv.conf had no nameserver entries (they get inserted on the fly by
wvdial) so I put in a couple. This stops squid from crashing, but instead I
see a lot of entries in cache.log about squid being unable to reach the
nameservers like
2001/05/09 10:20:11| comm_udp_sendto: FD 2, 194.145.128.2, port 53: (101)
Network is unreachable
2001/05/09 10:20:11| idnsSendQuery: FD 2: sendto: (101) Network is unreachable
These are unsurprising as I was offline at the the time :-)
The strange thing is that squid should not have tried to do a DNS lookup for
the requested address - it's on my local LAN, nsswitch.conf is configured
correctly (I can ping the relevant address by name) so why is squid trying
to do a DNS lookup and how do I stop it ?
Regards,
Niall
This archive was generated by hypermail 2.1.6 : Thu 06 Feb 2003 - 13:10:14 GMT