OpenVZ Forum: Support

Home » General » Support » mem leak

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

mem leak [message #29746]

Thu, 24 April 2008 00:55

kapper
Messages: 6
Registered: June 2006
Location: Vienna, Austria

Junior Member

hi
I suspect three openvz-servers administered by us do suffer similiar problems like http://forum.openvz.org/index.php?t=tree&goto=26664

first - all RHEL4 based openvz-servers do not have this problem here.

only RHEL5 based openvz-systems are affected.

roughly 45 days after reboot - those machines - running not a very high load - have eaten up 8 GB swap and 4 GB RAM.

stopping VEs does only free several 100 Megs. stopping openvz alltogether also doesn't help - it seems the kernel is eating up all of this.

only a reboot fixes this.

latest verified kernel with this problem:
2.6.18-53.1.6.el5.028stab053.6

releases of kernels before are unknown but I remember having to reboot those boxes at least for two more times, so probably the problem was introduced during Summer 2007.

we're now running 2.6.18-53.1.13.el5.028stab053.10 - but this only for a few hours - so any suggestions are very welcome.

kindest regards
harald kapper

Report message to a moderator

Re: mem leak [message #29749 is a reply to message #29746]

Thu, 24 April 2008 06:40

vaverin
Messages: 708
Registered: September 2005

Senior Member

Hi Harald,

i would note that I do not see any kernel-related problems here:
linux kernel by design frees the memory only when it really required -- due performance optimisation reasons.

By this way kernel without any significat load will use all RAM, for example for disk cache. however when somebody will request new memory -- kernel will drop oldest cache entries (it is quite fast operations) and give this memory to application.

Thereofre from kernel point of view the situation when kernel and running userspace programs eats toghether all memory -- is normal.

However from userspace point of view huge swap memory usage can point to memory leak in some aplication:
Huge swap usage points to huge userspace memory consumption (kernel do not uses swap for kernel structures, swap contains only rarely used programm's data). It may be normal, or may be caused by some memory leaks in applications.

thank you,
Vasily Averin

Report message to a moderator

Re: mem leak [message #29760 is a reply to message #29749]

Thu, 24 April 2008 07:32

kapper
Messages: 6
Registered: June 2006
Location: Vienna, Austria

Junior Member

Dear Vaverin,
I would absolutely agree - but....

thins on the hardware-node look bad after roughly 45 days uptime:

free -m
total used free shared buffers cached
Mem: 3946 3846 100 0 34 585
-/+ buffers/cache:415 515
Swap: 8191 8191 0

which similar 2.6.9 kernel based host-nodes do not share as a common resource-problem.

this view doesn't change when killing all VEs - swap is still at 8 GB USED - which is for an idle host - definitely nothing normal.

the same box after the reboot yesterday and several hours uptime:

free -m
total used free shared buffers cached
Mem: 3946 3782 163 0 201 2430
-/+ buffers/cache:1150 2796
Swap:8191 0 8191

also three different hardware-nodes do experience the same problem running the EL5 kernel, EL4 kernel based boxes do not suffer this very problem.

finally there are some error-messages that might enlighten anyone on this problem:

WARNING: Kernel Errors Present
[<ffffffff8005ee39>] error_exit+0x0/0x84 ...: 14 Time(s)

and on the other box

WARNING: Kernel Errors Present
[<ffffffff8005ee39>] error_exit+0x0/0x84 ...: 30 Time(s)

and the serverlog shows the outofmemory-killing going on as we invoke yum to upgrade the kernel...

Apr 23 23:49:22 hostname kernel: yum invoked oom-killer: gfp_mask=0x201d2, order=0, oomkilladj=0
Apr 23 23:49:22 hostname kernel:
Apr 23 23:49:22 hostname kernel: Call Trace:
Apr 23 23:49:22 hostname kernel: [<ffffffff800c6672>] out_of_memory+0x9f/0x25e
Apr 23 23:49:22 hostname kernel: [<ffffffff8000eda3>] __alloc_pages+0x242/0x332
Apr 23 23:49:22 hostname kernel: [<ffffffff800124c4>] __do_page_cache_readahead+0x95/0x1d9
Apr 23 23:49:22 hostname kernel: [<ffffffff8002999f>] sync_page+0x0/0x42
Apr 23 23:49:22 hostname kernel: [<ffffffff8005b8d4>] getnstimeofday+0x10/0x28
Apr 23 23:49:22 hostname kernel: [<ffffffff8009d3c7>] ktime_get_ts+0x1a/0x4d
Apr 23 23:49:22 hostname kernel: [<ffffffff800bf752>] delayacct_end+0x5d/0x86
Apr 23 23:49:22 hostname kernel: [<ffffffff80032e9c>] blockable_page_cache_readahead+0x53/0xb2
Apr 23 23:49:22 hostname kernel: [<ffffffff8002fdef>] make_ahead_window+0x82/0x9e
Apr 23 23:49:22 hostname kernel: [<ffffffff80013763>] page_cache_readahead+0x17f/0x1af
Apr 23 23:49:22 hostname kernel: [<ffffffff8000b992>] do_generic_mapping_read+0x126/0x3f8
Apr 23 23:49:22 hostname kernel: [<ffffffff8000c917>] file_read_actor+0x0/0x13c
Apr 23 23:49:22 hostname kernel: [<ffffffff8000bdb0>] __generic_file_aio_read+0x14c/0x190
Apr 23 23:49:22 hostname kernel: [<ffffffff80016800>] generic_file_aio_read+0x34/0x39
Apr 23 23:49:22 hostname kernel: [<ffffffff8000c633>] do_sync_read+0xc7/0x104
Apr 23 23:49:22 hostname kernel: [<ffffffff80066051>] do_page_fault+0x4e9/0x7ef
Apr 23 23:49:22 hostname kernel: [<ffffffff8009b432>] autoremove_wake_function+0x0/0x2e
Apr 23 23:49:22 hostname kernel: [<ffffffff80061597>] __sched_text_start+0x177/0xee2
Apr 23 23:49:22 hostname kernel: [<ffffffff8000af5c>] vfs_read+0xaa/0x150
Apr 23 23:49:22 hostname kernel: [<ffffffff8000b125>] fget_light+0x18/0x7c
Apr 23 23:49:22 hostname kernel: [<ffffffff80012c4a>] sys_pread64+0x54/0xc4
Apr 23 23:49:22 hostname kernel: [<ffffffff8005e166>] system_call+0x7e/0x83
Apr 23 23:49:22 hostname kernel:
Apr 23 23:49:22 hostname kernel: Mem-info:
Apr 23 23:49:22 hostname kernel: Node 0 DMA per-cpu:
Apr 23 23:49:22 hostname kernel: cpu 0 hot: high 0, batch 1 used:0
Apr 23 23:49:22 hostname kernel: cpu 0 cold: high 0, batch 1 used:0
Apr 23 23:49:22 hostname kernel: cpu 1 hot: high 0, batch 1 used:0
Apr 23 23:49:22 hostname kernel: cpu 1 cold: high 0, batch 1 used:0
Apr 23 23:49:22 hostname kernel: Node 0 DMA32 per-cpu:
Apr 23 23:49:22 hostname kernel: cpu 0 hot: high 186, batch 31 used:125
Apr 23 23:49:22 hostname kernel: cpu 0 cold: high 62, batch 15 used:50
Apr 23 23:49:22 hostname kernel: cpu 1 hot: high 186, batch 31 used:15
Apr 23 23:49:22 hostname kernel: cpu 1 cold: high 62, batch 15 used:14
Apr 23 23:49:22 hostname kernel: Node 0 Normal per-cpu:
Apr 23 23:49:22 hostname kernel: cpu 0 hot: high 186, batch 31 used:57
Apr 23 23:49:22 hostname kernel: cpu 0 cold: high 62, batch 15 used:57
Apr 23 23:49:22 hostname kernel: cpu 1 hot: high 186, batch 31 used:36
Apr 23 23:49:22 hostname kernel: cpu 1 cold: high 62, batch 15 used:46
Apr 23 23:49:22 hostname kernel: Node 0 HighMem per-cpu: empty
Apr 23 23:49:22 hostname kernel: Free pages: 21556kB (0kB HighMem)
Apr 23 23:49:22 hostname kernel: Active:482971 inactive:471530 dirty:500 writeback:0 unstable:0 free:5389 slab:26825 mapped-file:9603 mapped-anon:844193 pagetables:10871
Apr 23 23:49:22 hostname kernel: Node 0 DMA free:10980kB min:20kB low:24kB high:28kB active:0kB inactive:0kB present:10532kB pages_scanned:0 all_unreclaimable? yes
Apr 23 23:49:22 hostname kernel: lowmem_reserve[]: 0 3502 4006 4006
Apr 23 23:49:22 hostname kernel: Node 0 DMA32 free:9196kB min:7072kB low:8840kB high:10608kB active:1727524kB inactive:1716528kB present:3586732kB pages_scanned:3724 all_unreclaimable? no
Apr 23 23:49:22 hostname kernel: lowmem_reserve[]: 0 0 504 504
Apr 23 23:49:22 hostname kernel: Node 0 Normal free:1380kB min:1016kB low:1268kB high:1524kB active:204360kB inactive:169592kB present:516096kB pages_scanned:4 all_unreclaimable? no
Apr 23 23:49:22 hostname kernel: lowmem_reserve[]: 0 0 0 0
Apr 23 23:49:22 hostname kernel: Node 0 HighMem free:0kB min:128kB low:128kB high:128kB active:0kB inactive:0kB present:0kB pages_scanned:0 all_unreclaimable? no
Apr 23 23:49:22 hostname kernel: lowmem_reserve[]: 0 0 0 0
Apr 23 23:49:22 hostname kernel: Node 0 DMA: 3*4kB 5*8kB 3*16kB 4*32kB 4*64kB 2*128kB 2*256kB 1*512kB 1*1024kB 0*2048kB 2*4096kB = 10980kB
Apr 23 23:49:22 hostname kernel: Node 0 DMA32: 31*4kB 2*8kB 18*16kB 32*32kB 7*64kB 1*128kB 0*256kB 0*512kB 1*1024kB 1*2048kB 1*4096kB = 9196kB
Apr 23 23:49:22 hostname kernel: Node 0 Normal: 105*4kB 8*8kB 10*16kB 1*32kB 3*64kB 0*128kB 0*256kB 1*512kB 0*1024kB 0*2048kB 0*4096kB = 1380kB
Apr 23 23:49:22 hostname kernel: Node 0 HighMem: empty
Apr 23 23:49:22 hostname kernel: Swap cache: add 2338321, delete 2336791, find 32746461/32778627, race 0+92+75
Apr 23 23:49:22 hostname kernel: Free swap = 0kB
Apr 23 23:49:22 hostname kernel: Total swap = 8388600kB
Apr 23 23:49:22 hostname kernel: Free swap: 0kB
Apr 23 23:49:22 hostname kernel: 1179648 pages of RAM
Apr 23 23:49:22 hostname kernel: 169401 reserved pages
Apr 23 23:49:22 hostname kernel: 73976 pages shared
Apr 23 23:49:22 hostname kernel: 1531 pages swap cached
Apr 23 23:49:22 hostname kernel: Top 10 caches:
Apr 23 23:49:22 hostname kernel: filp : size 1617920 objsize 256
Apr 23 23:49:22 hostname kernel: page_beancounter : size 61296640 objsize 64
Apr 23 23:49:22 hostname kernel: buffer_head : size 1445888 objsize 96
Apr 23 23:49:22 hostname kernel: radix_tree_node : size 3624960 objsize 536
Apr 23 23:49:22 hostname kernel: size-2048 : size 1688960 objsize 2048
Apr 23 23:49:22 hostname kernel: size-128 : size 1941504 objsize 128
Apr 23 23:49:22 hostname kernel: vm_area_struct : size 5017600 objsize 176
Apr 23 23:49:22 hostname kernel: ext3_inode_cache : size 8697216 objsize 808
Apr 23 23:49:22 hostname kernel: task_struct : size 1425408 objsize 2192
Apr 23 23:49:22 hostname kernel: dentry_cache : size 5943296 objsize 248
Apr 23 23:49:22 hostname kernel: Out of memory: Killed process 12866 (apache2).
Apr 23 23:49:22 hostname kernel: Mem-info:
Apr 23 23:49:22 hostname kernel: Node 0 DMA per-cpu:
Apr 23 23:49:22 hostname kernel: cpu 0 hot: high 0, batch 1 used:0
Apr 23 23:49:22 hostname kernel: cpu 0 cold: high 0, batch 1 used:0
Apr 23 23:49:22 hostname kernel: cpu 1 hot: high 0, batch 1 used:0
Apr 23 23:49:22 hostname kernel: cpu 1 cold: high 0, batch 1 used:0
Apr 23 23:49:22 hostname kernel: Node 0 DMA32 per-cpu:
Apr 23 23:49:22 hostname kernel: cpu 0 hot: high 186, batch 31 used:125
Apr 23 23:49:22 hostname kernel: cpu 0 cold: high 62, batch 15 used:49
Apr 23 23:49:22 hostname kernel: cpu 1 hot: high 186, batch 31 used:15
Apr 23 23:49:22 hostname kernel: cpu 1 cold: high 62, batch 15 used:14
Apr 23 23:49:22 hostname kernel: Node 0 Normal per-cpu:
Apr 23 23:49:22 hostname kernel: cpu 0 hot: high 186, batch 31 used:22
Apr 23 23:49:22 hostname kernel: cpu 0 cold: high 62, batch 15 used:61
Apr 23 23:49:22 hostname kernel: cpu 1 hot: high 186, batch 31 used:36
Apr 23 23:49:22 hostname kernel: cpu 1 cold: high 62, batch 15 used:46
Apr 23 23:49:22 hostname kernel: Node 0 HighMem per-cpu: empty
Apr 23 23:49:22 hostname kernel: Free pages: 24136kB (0kB HighMem)
Apr 23 23:49:22 hostname kernel: Active:482668 inactive:471214 dirty:128 writeback:0 unstable:0 free:6034 slab:26851 mapped-file:9603 mapped-anon:844193 pagetables:10871
Apr 23 23:49:22 hostname kernel: Node 0 DMA free:10980kB min:20kB low:24kB high:28kB active:0kB inactive:0kB present:10532kB pages_scanned:0 all_unreclaimable? yes
Apr 23 23:49:22 hostname kernel: lowmem_reserve[]: 0 3502 4006 4006
Apr 23 23:49:22 hostname kernel: Node 0 DMA32 free:11536kB min:7072kB low:8840kB high:10608kB active:1726548kB inactive:1715316kB present:3586732kB pages_scanned:364 all_unreclaimable? no
Apr 23 23:49:22 hostname kernel: lowmem_reserve[]: 0 0 504 504
Apr 23 23:49:22 hostname kernel: Node 0 Normal free:1620kB min:1016kB low:1268kB high:1524kB active:204124kB inactive:169540kB present:516096kB pages_scanned:6 all_unreclaimable? ...

[ Show the rest of the message ]

Report message to a moderator

Re: mem leak [message #29763 is a reply to message #29760]

Thu, 24 April 2008 07:49

vaverin
Messages: 708
Registered: September 2005

Senior Member

Dear Harald,

kernel errors messages are very important. Any kernel error can lead to unpredictible results. It may be crashes or hangs, but memory leakage is also possible.

To found the real cause of the trouble we need to look at the first
error message (because the following errors can be caused by previous).

Could you please check your logs about the oops messages?
http://wiki.openvz.org/When_you_have_an_oops

Also I can do it for You if you give me an access to your node (of course via Private Messaging)

Thank you,
Vasily Averin

Report message to a moderator

Re: mem leak [message #29767 is a reply to message #29763]

Thu, 24 April 2008 08:22

kapper
Messages: 6
Registered: June 2006
Location: Vienna, Austria

Junior Member

hi
the kernel was:
Linux 2.6.18-53.1.13.el5.028stab053.6 compiled by openvz for RHEL5.

now running
Linux hwnode 2.6.18-53.1.13.el5.028stab053.10 #1 SMP Tue Apr 1 14:45:45 MSD 2008 x86_64 x86_64 x86_64 GNU/Linux

----------------

the first error happened on April 21st 10.55 localtime.

should I post the full log or simply private-message it so someone?

the log starts (very first error):

Apr 21 10:55:02 hwnode kernel: apache invoked oom-killer: gfp_mask=0x200d2, order=0, oomkilladj=0
Apr 21 10:55:02 hwnode kernel:
Apr 21 10:55:02 hwnode kernel: Call Trace:
Apr 21 10:55:05 hwnode kernel: [<ffffffff800c6672>] out_of_memory+0x9f/0x25e
Apr 21 10:55:05 hwnode kernel: [<ffffffff8000eda3>] __alloc_pages+0x242/0x332
Apr 21 10:55:05 hwnode kernel: [<ffffffff80010900>] do_wp_page+0x315/0x6b5
Apr 21 10:55:05 hwnode kernel: [<ffffffff80008e8f>] __handle_mm_fault+0xf6c/0x1041
Apr 21 10:55:05 hwnode kernel: [<ffffffff80098ed5>] attach_pid+0x8c/0xb9
Apr 21 10:55:06 hwnode kernel: [<ffffffff8003cbe6>] remove_wait_queue+0x1c/0x2c
Apr 21 10:55:08 hwnode kernel: [<ffffffff80029214>] do_wait+0xad8/0xb74
Apr 21 10:55:08 hwnode kernel: [<ffffffff8006601e>] do_page_fault+0x4b6/0x7ef
Apr 21 10:55:09 hwnode kernel: [<ffffffff800f72ba>] compat_core_sys_select+0x1bf/0x1d0
Apr 21 10:55:10 hwnode kernel: [<ffffffff8005ee39>] error_exit+0x0/0x84
Apr 21 10:55:10 hwnode kernel:
Apr 21 10:55:10 hwnode kernel: Mem-info:
Apr 21 10:55:10 hwnode kernel: Node 0 DMA per-cpu:
Apr 21 10:55:10 hwnode kernel: cpu 0 hot: high 0, batch 1 used:0
Apr 21 10:55:10 hwnode kernel: cpu 0 cold: high 0, batch 1 used:0
Apr 21 10:55:10 hwnode kernel: cpu 1 hot: high 0, batch 1 used:0
Apr 21 10:55:10 hwnode kernel: cpu 1 cold: high 0, batch 1 used:0
Apr 21 10:55:10 hwnode kernel: Node 0 DMA32 per-cpu:
Apr 21 10:55:10 hwnode kernel: cpu 0 hot: high 186, batch 31 used:39
Apr 21 10:55:10 hwnode kernel: cpu 0 cold: high 62, batch 15 used:58
Apr 21 10:55:10 hwnode kernel: cpu 1 hot: high 186, batch 31 used:33
Apr 21 10:55:10 hwnode kernel: cpu 1 cold: high 62, batch 15 used:59
Apr 21 10:55:10 hwnode kernel: Node 0 Normal per-cpu:
Apr 21 10:55:10 hwnode kernel: cpu 0 hot: high 186, batch 31 used:18
Apr 21 10:55:10 hwnode kernel: cpu 0 cold: high 62, batch 15 used:50
Apr 21 10:55:10 hwnode kernel: cpu 1 hot: high 186, batch 31 used:28
Apr 21 10:55:10 hwnode kernel: cpu 1 cold: high 62, batch 15 used:56
Apr 21 10:55:10 hwnode kernel: Node 0 HighMem per-cpu: empty
Apr 21 10:55:10 hwnode kernel: Free pages: 20384kB (0kB HighMem)
Apr 21 10:55:10 hwnode kernel: Active:796844 inactive:123293 dirty:260 writeback:12112 unstable:0 free:5096 slab:60543 mapped-file:21985 mapped-anon:734226 pagetables:11267
Apr 21 10:55:10 hwnode kernel: Node 0 DMA free:10980kB min:20kB low:24kB high:28kB active:0kB inactive:0kB present:10532kB pages_scanned:0 all_unreclaimable? yes
Apr 21 10:55:10 hwnode kernel: lowmem_reserve[]: 0 3502 4006 4006
Apr 21 10:55:10 hwnode kernel: Node 0 DMA32 free:8452kB min:7072kB low:8840kB high:10608kB active:2861528kB inactive:454940kB present:3586732kB pages_scanned:8220 all_unreclaimable? no
Apr 21 10:55:10 hwnode kernel: lowmem_reserve[]: 0 0 504 504
Apr 21 10:55:10 hwnode kernel: Node 0 Normal free:952kB min:1016kB low:1268kB high:1524kB active:325976kB inactive:38232kB present:516096kB pages_scanned:96 all_unreclaimable? no
Apr 21 10:55:10 hwnode kernel: lowmem_reserve[]: 0 0 0 0
Apr 21 10:55:10 hwnode kernel: Node 0 HighMem free:0kB min:128kB low:128kB high:128kB active:0kB inactive:0kB present:0kB pages_scanned:0 all_unreclaimable? no
Apr 21 10:55:10 hwnode kernel: lowmem_reserve[]: 0 0 0 0
Apr 21 10:55:10 hwnode kernel: Node 0 DMA: 3*4kB 5*8kB 3*16kB 4*32kB 4*64kB 2*128kB 2*256kB 1*512kB 1*1024kB 0*2048kB 2*4096kB = 10980kB
Apr 21 10:55:10 hwnode kernel: Node 0 DMA32: 313*4kB 0*8kB 0*16kB 1*32kB 0*64kB 0*128kB 0*256kB 0*512kB 1*1024kB 1*2048kB 1*4096kB = 8452kB
Apr 21 10:55:10 hwnode kernel: Node 0 Normal: 96*4kB 1*8kB 1*16kB 1*32kB 0*64kB 0*128kB 0*256kB 1*512kB 0*1024kB 0*2048kB 0*4096kB = 952kB
Apr 21 10:55:10 hwnode kernel: Node 0 HighMem: empty
Apr 21 10:55:10 hwnode kernel: Swap cache: add 2307383, delete 2290105, find 31844745/31873186, race 0+89+67
Apr 21 10:55:10 hwnode kernel: Free swap = 0kB
Apr 21 10:55:10 hwnode kernel: Total swap = 8388600kB
Apr 21 10:55:10 hwnode kernel: Free swap: 0kB
Apr 21 10:55:10 hwnode kernel: 1179648 pages of RAM
Apr 21 10:55:10 hwnode kernel: 169401 reserved pages
Apr 21 10:55:10 hwnode kernel: 184758 pages shared
Apr 21 10:55:10 hwnode kernel: 17228 pages swap cached

thank you in advance
hk

[Updated on: Thu, 24 April 2008 08:23]

Report message to a moderator

Re: mem leak [message #29768 is a reply to message #29767]

Thu, 24 April 2008 08:40

kir
Messages: 1645
Registered: August 2005
Location: Moscow, Russia

Senior Member

Can you show us what 'top' shows on a "loaded" box?

Run top, then press 'M' (capital M, i.e. usually Shift+M), then copy-paste what you see on the screen to here.

Kir Kolyshkin

http://static.openvz.org/userbars/openvz-developer.png

Report message to a moderator

Re: mem leak [message #29770 is a reply to message #29768]

Thu, 24 April 2008 08:45

kapper
Messages: 6
Registered: June 2006
Location: Vienna, Austria

Junior Member

hi
absolutely - though I probably have to wait for 40 days now after rebooting all three boxes in order to get what you want.

now it shows:

top - 10:44:09 up 10:01,  1 user,  load average: 0.06, 0.09, 0.18
Tasks: 318 total,   1 running, 317 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.3%us,  0.3%sy,  0.0%ni, 99.2%id,  0.0%wa,  0.0%hi,  0.2%si,  0.0%st
Mem:   4040988k total,  3945596k used,    95392k free,   213668k buffers
Swap:  8388600k total,        0k used,  8388600k free,  2513336k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND                                                                       
20537 root      15   0  339m 112m 3760 S    0  2.9   0:13.96 yum-updatesd                                                                  
10730 root      15   0  121m 105m 1912 S    0  2.7   4:17.38 pdns_recursor                                                                 
 8521 100       15   0  134m  50m 4504 S    0  1.3   8:49.31 mysqld                                                                        
12793 103       15   0  160m  42m 5640 S    0  1.1   0:55.73 mysqld                                                                        
20653 root      18   0  164m  27m 1684 S    0  0.7   0:10.86 python                                                                        
21154 33        15   0 41660  24m 3560 S    0  0.6   0:13.76 apache                                                                        
20866 33        15   0 41552  24m 3564 S    0  0.6   0:10.36 apache                                                                        
20860 33        15   0 41432  23m 3252 S    0  0.6   0:09.10 apache                                                                        
21312 33        15   0 41424  23m 3248 S    0  0.6   0:13.96 apache                                                                        
20699 33        15   0 40996  23m 3588 S    0  0.6   0:16.37 apache                                                                        
16445 101       15   0  161m  23m 5428 S    0  0.6   0:01.33 mysqld                                                                        
21460 33        15   0 40992  23m 3248 S    0  0.6   0:08.47 apache                                                                        
21341 33        15   0 40156  22m 3556 S    0  0.6   0:07.32 apache                                                                        
22947 33        15   0 40316  22m 3220 S    0  0.6   0:00.94 apache                                                                        
22672 33        15   0 40140  22m 3216 S    0  0.6   0:03.60 apache                                                                        
14785 101       15   0  158m  22m 5492 S    0  0.6   0:00.93 mysqld                                                                        
16420 101       15   0  158m  22m 5192 S    0  0.6   0:01.37 mysqld                                                                        
20506 101       15   0  158m  21m 5360 S    0  0.6   0:00.76 mysqld                                                                        
10995 101       15   0  158m  21m 5392 S    0  0.6   0:01.63 mysqld                                                                        
17028 33        15   0  115m  16m 4176 S    0  0.4   0:00.17 apache2                                                                       
16504 33        15   0  115m  16m 4036 S    0  0.4   0:03.14 apache2                                                                       
21437 33        15   0  115m  16m 4020 S    0  0.4   0:02.39 apache2                                                                       
16505 33        15   0  115m  16m 3996 S    0  0.4   0:02.69 apache2                                                                       
21433 33        15   0  115m  16m 3996 S    0  0.4   0:02.73 apache2                                                                       
16502 33        15   0  115m  16m 3960 S    0  0.4   0:02.76 apache2                                                                       
16499 33        15   0  115m  16m 3952 S    0  0.4   0:02.56 apache2                                                                       
16503 33        18   0  115m  16m 3952 S    0  0.4   0:02.03 apache2                                                                       
18992 root      18   0  122m  10m 5764 S    0  0.3   0:00.06 apache2                                                                       
19133 33        18   0  122m 9.8m 4744 S    0  0.2   0:10.95 apache2                                                                       
19130 33        15   0  122m 9976 4728 S    0  0.2   0:11.36 apache2                                                                       
21345 33        16   0  123m 9972 4712 S    0  0.2   0:02.85 apache2                                                                       
21336 33        15   0  123m 9968 4704 S    0  0.2   0:07.21 apache2                                                                       
19132 33        15   0  123m 9960 4712 S    0  0.2   0:11.40 apache2                                                                       
19131 33        15   0  122m 9956 4720 S    0  0.2   0:15.42 apache2                                                                       
19129 33        15   0  122m 9952 4708 S    0  0.2   0:06.72 apache2                                                                       
21346 33        15   0  122m 9928 4664 S    0  0.2   0:13.24 apache2                                                                       
21062 33        15   0  122m 9916 4660 S    0  0.2   0:17.51 apache2                                                                       
21337 33        15   0  123m 9868 4620 S    0  0.2   0:03.11 apache2                                                                       
19178 33        15   0 97348 9192 4544 S    0  0.2   0:00.40 apache2                                                                       
19179 33        18   0 97580 9156 4536 S    0  0.2   0:00.30 apache2                                                                       
19176 33        15   0 97604 9124 4512 S    0  0.2   0:00.41 apache2                                                                       
21141 33        15   0 97600 8792 4148 S    0  0.2   0:00.29 apache2                                                                       
19177 33        15   0 97320 8788 4176 S    0  0.2   0:00.33 apache2                                                                       
19175 33        18   0 97568 8760 4156 S    0  0.2   0:00.35 apache2                                                                       
19080 root      18   0 97032 8692 4508 S    0  0.2   0:00.06 apache2                                                                       
20933 33        15   0 82368 8692 4248 S    0  0.2   0:00.28 apache2                                                                       
21434 33        15   0 97564 8684 4116 S    0  0.2   0:00.20 apache2                                                                       
15006 root      18   0  103m 8656 5276 S    0  0.2   0:00.06 apache2                                                                       
20929 33        15   0 82360 8536 4128 S    0  0.2   0:00.51 apache2                                                                       
20931 33        15   0 82364 8528 4140 S    0  0.2   0:00.69 apache2                                                                       
18676 33        15   0 81952 8412 4044 S    0  0.2   0:00.42 apache2                                                                       
18677 33        15   0 81876 8228 3940 S    0  0.2   0:00.32 apache2

[Updated on: Thu, 24 April 2008 08:51] by Moderator

Report message to a moderator

Re: mem leak [message #29772 is a reply to message #29767]

Thu, 24 April 2008 08:50

vaverin
Messages: 708
Registered: September 2005

Senior Member

Hi Harald,

oom-killer messages are not an errors,
oom-killer kills the processes to free the memory in case when kernel cannot free it by any another ways

messages like "WARNING: Kernel Errors Present" -- is more interesting for me.

Could you please give me an access to your node? If not -- I would like to look at the /var/log/dmesg file (to understand configuration of your node) , /var/log/messages* files (to find all kernel error messages) and to dmesg (to look at the current errors). Please sent it to me via PM or via email to vvs@parallels.com

thank you,
Vasily Averin

Report message to a moderator

Re: mem leak [message #29773 is a reply to message #29770]

Thu, 24 April 2008 08:53

kir
Messages: 1645
Registered: August 2005
Location: Moscow, Russia

Senior Member

Yup, so far the situation looks pretty decent -- although yum-updatesd eats a bit too much memory, it's still OK.

Please re-check that in a week or two, and post your results here.

Kir Kolyshkin

Report message to a moderator

Re: mem leak [message #29777 is a reply to message #29772]

Thu, 24 April 2008 10:49

vaverin
Messages: 708
Registered: September 2005

Senior Member

Hi Harald,

thank you for the logs. I've found the follwing messages:

1) Machine check exception message:
Mar 30 12:38:04 k4 kernel: Machine check events logged
it points to some hardware troubles. It may not be related to our current issue, but anyway -- I would recommend You to found its cause.
Some motherboards saves MCE events to the motherboard BIOS. Sometimes userspace daemon writes it to the /var/log/mce.log file.

2) disk-related issue:
Apr 10 13:50:05 k4 kernel:
ata2.00: exception Emask 0x0 SAct 0x70000 SErr 0x0 action 0x2 frozen
...
ata2: hard resetting port
ata2: port is slow to respond, please be patient (Status 0x80)
ata2: COMRESET failed (errno=-16)
ata2: hard resetting port
ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
ata2.00: configured for UDMA/133
ata2: EH complete
SCSI device sdb: 488397168 512-byte hdwr sectors (250059 MB)
sdb: Write Protect is off

scsi subsystem has detected some issue, and started Error Handler to resolve it. Ideally it should not lead to any troubles, however if this operation was not handled properly then it can be the cause of the following troubles on your node.

3) unfrtunaely I've not found any "WARNING: Kernel Errors Present" messages in the logs.

4) I've investigate OOM killer messages and they looks straange for me, I'll consult with my colleguaes and tell You about result a bit later.

Thank you,
Vasily Averin

Report message to a moderator

Re: mem leak [message #29778 is a reply to message #29777]

Thu, 24 April 2008 10:50

vaverin
Messages: 708
Registered: September 2005

Senior Member

btw, if you have similair issue on the some other nodes -- please send me its logs too, it would be useful.

thank you,
Vasily Averin

Report message to a moderator

Previous Topic:	Physical (RH7.3) to Container (CentOS 5)?
Next Topic:	One NIC, Multiple Gateways

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Tue May 21 07:47:02 GMT 2024

Total time taken to generate the page: 0.01544 seconds