Repost:
I'm having a memory leak:
2.0.rc3 on Generic and
2.0.1 on Generic
PIII/500MHz 128Mb
Longshine HFC-S LCS-8051A
Longshine HFC-S LCS-8051(not A?)
/proc/meminfo:
00002: Thu May 27 12:05:04 CEST 2010
00004: MemTotal: 124216 kB
00005: MemFree: 81040 kB
00027: Slab: 24832 kB
00029: SUnreclaim: 22132 kB
....
00453: Wed Jun 9 14:36:56 CEST 2010
00455: MemTotal: 124216 kB
00456: MemFree: 3320 kB
00478: Slab: 113016 kB
00480: SUnreclaim: 111912 kB
/proc/slabinfo:
00002: Thu May 27 12:08:36 CEST 2010
00005: # name <active_objs> <num_objs> <objsize> <objperslab> <pagesperslab> : tunables <limit> <batchcount> <sharedfactor> : slabdata <active_slabs> <num_slabs> <sharedavail>
00063: skbuff_head_cache 29472 29472 160 24 1 : tunables 120 60 0 : slabdata 1228 1228 0
00120: size-512 29664 29664 512 8 1 : tunables 54 27 0 : slabdata 3708 3708 0
.....
01476: Wed Jun 9 14:37:24 CEST 2010
01479: # name <active_objs> <num_objs> <objsize> <objperslab> <pagesperslab> : tunables <limit> <batchcount> <sharedfactor> : slabdata <active_slabs> <num_slabs> <sharedavail>
01537: skbuff_head_cache 162466 162480 160 24 1 : tunables 120 60 0 : slabdata 6770 6770 0
01594: size-512 162664 162664 512 8 1 : tunables 54 27 0 : slabdata 20333 20333 0
Slab is growing continuously; only growing slabs shown. In the end, I was still able to connect via ssh but Asterisk had died down - I had to reboot.
Has anybody an idea how to trace this problem further?
