out of memory

Hi,
OMV server crashes from time to time because of out of memory.

in the /var/log/syslog:

Dec 12 11:43:48 omv kernel: [51832.264261] Node 0 Normal: 5787*4kB 10494*8kB 8603*16kB 4*32kB 1*64kB 1*128kB 1*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 245324kB
Dec 12 11:43:48 omv kernel: [51832.264271] 724 total pagecache pages
Dec 12 11:43:48 omv kernel: [51832.264272] 31 pages in swap cache
Dec 12 11:43:48 omv kernel: [51832.264274] Swap cache stats: add 12147, delete 12116, find 307/833
Dec 12 11:43:48 omv kernel: [51832.264276] Free swap = 7777524kB
Dec 12 11:43:48 omv kernel: [51832.264278] Total swap = 7811064kB
Dec 12 11:43:48 omv kernel: [51832.294030] 2064384 pages RAM
Dec 12 11:43:48 omv kernel: [51832.294032] 88871 pages reserved
Dec 12 11:43:48 omv kernel: [51832.294034] 5770 pages shared
Dec 12 11:43:48 omv kernel: [51832.294035] 1901793 pages non-shared
Dec 12 11:43:48 omv kernel: [51832.294038] Out of memory: kill process 2134 (apache2) score 17912 or a child
Dec 12 11:43:48 omv kernel: [51832.294125] Killed process 2134 (apache2)
Dec 12 11:43:48 omv monit[2018]: 'localhost' mem usage of 96.3% matches resource limit [mem usage>90.0%]
Dec 12 11:43:48 omv monit[2018]: 'localhost' loadavg(5min) of 4.9 matches resource limit [loadavg(5min)>4.0]

top sorted by %mem:

top - 11:29:28 up 2:52, 1 user, load average: 2.22, 2.30, 2.00
Tasks: 203 total, 1 running, 202 sleeping, 0 stopped, 0 zombie
Cpu(s): 0.0%us, 0.7%sy, 0.0%ni, 78.7%id, 20.3%wa, 0.0%hi, 0.2%si, 0.0%st
Mem: 7902052k total, 7848568k used, 53484k free, 18248k buffers
Swap: 7811064k total, 0k used, 7811064k free, 7549552k cached

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
2245 root 20 0 88676 13m 2108 S 0 0.2 0:00.28 omv-engined
2083 root 20 0 137m 11m 6972 S 0 0.2 0:00.12 apache2
3695 www-data 20 0 140m 6132 1140 S 0 0.1 0:00.00 apache2
3696 www-data 20 0 140m 6132 1140 S 0 0.1 0:00.00 apache2
3697 www-data 20 0 140m 6132 1140 S 0 0.1 0:00.00 apache2
3698 www-data 20 0 140m 6132 1140 S 0 0.1 0:00.00 apache2
3699 www-data 20 0 140m 6132 1140 S 0 0.1 0:00.00 apache2
3694 www-data 20 0 128m 5368 936 S 0 0.1 0:00.03 apache2
2051 root 20 0 99.4m 4812 3696 S 0 0.1 0:00.02 smbd
2287 root 20 0 76316 4516 3428 S 0 0.1 0:00.03 winbindd
2745 root 20 0 99.8m 4340 3204 S 0 0.1 0:00.03 smbd
2278 root 20 0 76248 4092 3064 S 0 0.1 0:00.13 winbindd
2708 root 20 0 21220 4092 1652 S 0 0.1 0:00.21 bash
2303 root 20 0 82804 4016 2948 S 0 0.1 0:00.01 winbindd
2702 root 20 0 82604 3836 2976 S 0 0.0 0:00.29 sshd

I did also a memtester: everything ok

root@omv:~# dpkg -l | grep openmediavault
ii openmediavault 0.5.17 Open network attached storage solution
ii openmediavault-iscsitarget 0.5.5 OpenMediaVault iSCSI Enterprise Target plugin
ii openmediavault-keyring 0.2 GnuPG archive keys of the OpenMediaVault archive
ii openmediavault-lvm2 0.5.2 OpenMediaVault Logical Volume Manager (LVM2) plugin
ii openmediavault-nut 0.5.4 OpenMediaVault Network UPS Tools (NUT) plugin

driver issue? backports kernel? OMV extras?

Any idea?

thanks in advance

bruno

TERRA SERVER 4530 G1 1100724
CPU Xeon SP E3-1225v2 / LGA 1155 quad core
1 adaptec RAID 6405 SAS/STAT-600 512MB PCIe x 8 Sgl RAID
RAM 8GB DDR3
4 x HD 3.5" SATA2-RAID 1TB HGST HUA722010CLA330/7200

What kind of pb / bug is it?

What is the best way to solve the issue?

thks in advance,

Bruno

I did not installed OMV myself in fact

I guess OMV is installed on the raid array

How can I quickly check on which support has been installed OMV?

I did not run memtest from live-cd or anything else

just:

memtester 7500 1 # test 7500MB in one iteration

from Linux shell

I got no error but it is true that memtester does not test all sectors

difficult for me to stop OMV which is already used by open nebula

Do you think I must repeat the control with memtest?

No, OMV is a unique machine

OMV is one unique machine

we have Open Nebula with two nodes which use one nfs share on OMV

Do you think that an update could restore a normal situation?

Upgrade from latest OMV 0.5.X to OMV 1.0

here is the OS we are using:

Distributor ID: Debian
Description: Debian GNU/Linux 6.0.8 (squeeze)
Release: 6.0.8
Codename: squeeze

Can we continue and upgrade without hesitation?