Memory leak: slabinfo details

Boris Tobotras (boris@xtalk.msk.su)
Wed, 17 Sep 1997 10:54:08 +0400


Maybe this will help somehow.

Since 2.1.5x (54?) my linux box can't survive updatedb
invokation. I tried to manually run "find /" and monitor memory status.
"Cached", "buffers" and "free" came to <1M, system thrashed. This was in /
proc/slabinfo at the moment:

slabinfo - version: 1.0
kmem_cache 22 42
tcp_open_request 0 0
sock 70 80
filp 290 294
buffer_head 269 840
mm_struct 36 62
vm_area_struct 471 693
files_cache 38 49
uid_cache 2 127
size-131072 0 0
size-65536 0 0
size-32768 0 0
size-16384 1 1
size-8192 12 12
size-4096 3 4
size-2048 12 16
size-1024 51 56
size-512 372 376
size-256 54 56
size-128 61 75
size-64 60039 60060
size-32 59801 59850
slab_cache 64 126

It it of any help? Details: 2.1.55, 32M RAM, AMD K5.

-- 
	Best regards, -- Boris.