[squid-users] Squid 3.2 snapshot ... vanishing processes

From: alex sharaz <alexsharaz_at_me.com>
Date: Fri, 06 Jan 2012 12:03:11 +0000

Running squid 3.2 on a number of machines. Gradually the worker
process die with the FATAL xalloc message shown below. Config at
bottom of email. Then have to clear out cache directory,recreate and
restart squid. Usually wait till there only 1 or 2 workers left and
then do the lot. Would send a config but list end objects to size of
email being >50K
Rgds
Alex

2012/01/06 11:46:55 kid8| Starting Squid Cache version
3.2.0.14-20111228-r11479 for x86_64-unknown-linux-gnu...
2012/01/06 11:46:55 kid8| Process ID 31974
2012/01/06 11:46:55 kid8| Process Roles: worker
2012/01/06 11:46:55 kid8| With 49152 file descriptors available
2012/01/06 11:46:55 kid8| Initializing IP Cache...
2012/01/06 11:46:55 kid8| DNS Socket created at 0.0.0.0, FD 7
2012/01/06 11:46:55 kid8| Adding nameserver 150.237.84.21 from
squid.conf
2012/01/06 11:46:55 kid8| Adding nameserver 150.237.198.21 from
squid.conf
2012/01/06 11:46:55 kid8| helperOpenServers: Starting 5/20 'helper-
mux.pl' processes
2012/01/06 11:46:55 kid8| helperOpenServers: Starting 0/10
'basic_pam_auth' processes
2012/01/06 11:46:55 kid8| helperOpenServers: No 'basic_pam_auth'
processes needed.
2012/01/06 11:46:55 kid8| Logfile: opening log daemon:/logs/access.log
2012/01/06 11:46:55 kid8| Logfile Daemon: opening log /logs/access.log
2012/01/06 11:46:55 kid8| Local cache digest enabled; rebuild/rewrite
every 3600/3600 sec
2012/01/06 11:46:55 kid8| Store logging disabled
2012/01/06 11:46:55 kid8| WARNING: disk-cache maximum object size is
unlimited but mem-cache maximum object size is 32.00 KB
2012/01/06 11:46:55 kid8| Swap maxSize 4060160 + 262144 KB, estimated
332484 objects
2012/01/06 11:46:55 kid8| Target number of buckets: 16624
2012/01/06 11:46:55 kid8| Using 32768 Store buckets
2012/01/06 11:46:55 kid8| Max Mem size: 262144 KB [shared]
2012/01/06 11:46:55 kid8| Max Swap size: 4060160 KB
2012/01/06 11:46:55 kid8| Version 1 of swap file with LFS support
detected...
2012/01/06 11:46:55 kid8| Rebuilding storage in /cache/8 (CLEAN)
2012/01/06 11:46:55 kid8| Using Least Load store dir selection
2012/01/06 11:46:55 kid8| Set Current Directory to /usr/local/squid/
var/cache/squid
2012/01/06 11:46:55 kid8| Loaded Icons.
2012/01/06 11:46:55 kid8| HTCP Disabled.
2012/01/06 11:46:55 kid8| IcmpSquid.cc(255) Open: Pinger socket opened
on FD 24
2012/01/06 11:46:55 kid8| Ready to serve requests.
2012/01/06 11:46:55 kid8| Done reading /cache/8 swaplog (35 entries)
2012/01/06 11:46:55 kid8| Finished rebuilding storage from disk.
2012/01/06 11:46:55 kid8| 35 Entries scanned
2012/01/06 11:46:55 kid8| 0 Invalid entries.
2012/01/06 11:46:55 kid8| 0 With invalid flags.
2012/01/06 11:46:55 kid8| 35 Objects loaded.
2012/01/06 11:46:55 kid8| 0 Objects expired.
2012/01/06 11:46:55 kid8| 0 Objects cancelled.
2012/01/06 11:46:55 kid8| 0 Duplicate URLs purged.
2012/01/06 11:46:55 kid8| 0 Swapfile clashes avoided.
2012/01/06 11:46:55 kid8| Took 0.02 seconds (1772.87 objects/sec).
2012/01/06 11:46:55 kid8| Beginning Validation Procedure
2012/01/06 11:46:55 kid8| Completed Validation Procedure
2012/01/06 11:46:55 kid8| Validated 35 Entries
2012/01/06 11:46:55 kid8| store_swap_size = 13489710508802048.00 KB
FATAL: xcalloc: Unable to allocate 18446744073664220870 blocks of 1
bytes!

Squid Cache (Version 3.2.0.14-20111228-r11479): Terminated abnormally.
Received on Fri Jan 06 2012 - 12:04:41 MST

This archive was generated by hypermail 2.2.0 : Fri Jan 06 2012 - 12:00:02 MST