Re: [squid-users] errors with diskd

On Wed, 21 Mar 2001, Jan Satko wrote:

> Hi.
> I'm running True64 4.0F and squid 2.4stable1.
> # ./configure --enable-stacktraces --disable-ident-lookups
> --enable-err-language=Slovak --enable-cache-digest --enable-storeio=diskd
> --enable-kill-parent-hack --enable-removal-policies=heap --enable-delay-pools
> --enable-dlmalloc
>
> On linux box it works OK. But on True64 i'got errors(in cache.log)
> on startup:
>
> 2001/03/21 12:30:13| storeDiskdSend UNLINK: (35) Operation would block
> 2001/03/21 12:30:13| storeDiskdSend: msgsnd: (35) Operation would block
> 2001/03/21 12:30:13| storeDiskdSend UNLINK: (35) Operation would block
> 2001/03/21 12:30:13| storeDiskdSend: msgsnd: (35) Operation would block
> 2001/03/21 12:30:13| storeDiskdSend UNLINK: (35) Operation would block
> 2001/03/21 12:30:13| storeDiskdSend: msgsnd: (35) Operation would block
> 2001/03/21 12:30:13| storeDiskdSend UNLINK: (35) Operation would block
> 2001/03/21 12:30:13| storeDiskdSend: msgsnd: (35) Operation would block
> 2001/03/21 12:30:13| storeDiskdSend UNLINK: (35) Operation would block
> 2001/03/21 12:30:13| storeDiskdSend: msgsnd: (35) Operation would block
> 2001/03/21 12:30:13| storeDiskdSend UNLINK: (35) Operation would block
> 2001/03/21 12:30:13| storeDiskdSend: msgsnd: (35) Operation would block
> 2001/03/21 12:30:13| assertion failed: store_io_diskd.c:492:
> "++send_errors < 100"
> 2001/03/21 12:31:04| Starting Squid Cache version 2.4.STABLE1 for
> alpha-dec-osf4.0...
> 2001/03/21 12:31:04| Process ID 14573
> 2001/03/21 12:31:04| With 4096 file descriptors available
> 2001/03/21 12:31:04| Performing DNS Tests...
> 2001/03/21 12:31:04| Successful DNS name lookup tests...
> ...
> ..
> 2001/03/21 12:31:04| Unlinkd pipe opened on FD 7
> 2001/03/21 12:31:04| Swap maxSize 6860800 KB, estimated 527753 objects
> 2001/03/21 12:31:04| Target number of buckets: 26387
> 2001/03/21 12:31:04| Using 32768 Store buckets
> 2001/03/21 12:31:04| Max Mem size: 90112 KB
> 2001/03/21 12:31:04| Max Swap size: 6860800 KB
> 2001/03/21 12:31:04| Rebuilding storage in /squid-cache (DIRTY)
> 2001/03/21 12:31:04| Rebuilding storage in /squid-cache2 (DIRTY)
> 2001/03/21 12:31:04| Rebuilding storage in /var/squid-cache3 (DIRTY)
> 2001/03/21 12:31:04| Using Least Load store dir selection
> 2001/03/21 12:31:04| Set Current Directory to /var/squid-cache3/
> 2001/03/21 12:31:04| storeDiskdSend: msgsnd: (35) Operation would block
> 2001/03/21 12:31:04| storeDiskdSend OPEN: (35) Operation would block
> 2001/03/21 12:31:04| storeDiskdSend: msgsnd: (35) Operation would block
> 2001/03/21 12:31:04| storeDiskdSend OPEN: (35) Operation would block
> 2001/03/21 12:31:04| storeDiskdSend: msgsnd: (35) Operation would block
> 2001/03/21 12:31:04| storeDiskdSend OPEN: (35) Operation would block
> 2001/03/21 12:31:04| storeDiskdSend: msgsnd: (35) Operation would block
> 2001/03/21 12:31:04| storeDiskdSend OPEN: (35) Operation would block
>
> Can someone help me with this ?

Chances are that you don't have message queues in your kernel, or if you
do, the limits are very low. The FAQ has hints for increasing the
message queue and shared memory parameters.
Received on Wed Mar 21 2001 - 09:59:54 MST