We have a "/usr/sbin/squid -k rotate" on logrotate.d and everyday it
seems to give this "assetion failed" right after it..
And this seems to make squid completely unstable because this kid6 is a
disker... (SMP - 5 workers, 3 diskers...)
2018/11/20 00:25:11 kid6| storeDirWriteCleanLogs: Starting...
2018/11/20 00:25:11 kid6| Finished. Wrote 0 entries.
2018/11/20 00:25:11 kid6| Took 0.00 seconds ( 0.00 entries/sec).
2018/11/20 00:25:11 kid6| logfileRotate: daemon:/var/log/squid/access.log
2018/11/20 00:25:11 kid6| logfileRotate: daemon:/var/log/squid/access.log
"!isOpen(new_socket)"*
Please report this to our bugzilla. This looks very similar to
<https://bugs.squid-cache.org/show_bug.cgi?id=4796> but is definitely a
different bug.
If you can replicate the problem on a proxy not actually handling
traffic please supply a cache.log produced with debug_options ALL,9.
Otherwise if ALL,8 is too large please supply a cache log made with
ALL,6 - or at minimum "ALL,2 5,5 50,5 51,5 92,5". We need to know what
FD number is being allocated with the socket and everything Squid has
previously been doing with that FD number.
Amos