OpenVZ Forum: Support » Problem dcachesize

Home » General » Support » Problem dcachesize

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

Problem dcachesize [message #29071]

Fri, 04 April 2008 11:02

afrare
Messages: 11
Registered: November 2007

Junior Member

Dear Friends,

I incremented parameter dcachesize, but I continue receive failcnt. And when execute services and other commande I have received message "Cannot allocate memory".

I use centos 4.6 but, the container is centos 5.

Below, I show conf file.

ONBOOT="no"

# UBC parameters (in form of barrier:limit)
# Primary parameters
AVNUMPROC="40:40"
NUMPROC="200:200"
NUMTCPSOCK="280:280"
NUMOTHERSOCK="280:280"
VMGUARPAGES="65536:2147483647"
# Secondary parameters
KMEMSIZE="22020096:23488096"
TCPSNDBUF="3194880:5242880"
TCPRCVBUF="3194880:5242880"
OTHERSOCKBUF="1320960:3368960"
DGRAMRCVBUF="132096:132096"
OOMGUARPAGES="6144:2147483647"
# Auxiliary parameters
LOCKEDPAGES="32:32"
SHMPAGES="8192:8192"
PRIVVMPAGES="264144:524288"
NUMFILE="4096:4096"
NUMFLOCK="100:110"
NUMPTY="16:16"
NUMSIGINFO="256:256"
DCACHESIZE="1048576:1097728"

PHYSPAGES="0:2147483647"
NUMIPTENT="128:128"

# Disk quota parameters (in form of softlimit:hardlimit)
DISKSPACE="4194304:4613736"
DISKINODES="800000:880000"
QUOTATIME="0"
DISK QUOTA=yes
# CPU fair sheduler parameter
CPUUNITS="1000"
OFFLINE_MANAGEMENT="yes"
IP_ADDRESS="192.168.100.18"
VE_ROOT="/vz/root/$VEID"
VE_PRIVATE="/vz/private/$VEID"
OSTEMPLATE="centos-5-i386-default"
HOSTNAME="xx.xxx.xxx"
NAMESERVER="123.123.123.123"

FEATURES="sysfs:on "
====================================

Thanks for help.

Adriano

Report message to a moderator

Re: Problem dcachesize [message #29073 is a reply to message #29071]

Fri, 04 April 2008 11:20

maratrus
Messages: 1495
Registered: August 2007
Location: Moscow

Senior Member

Hi,

Could you please also show us
"cat /proc/user_beancounters" output and also "df -h" and "df -i" output from inside your VE.
And please exact command you used and exact error you received if it is possible.

Report message to a moderator

Re: Problem dcachesize [message #29075 is a reply to message #29073]

Fri, 04 April 2008 11:35

afrare
Messages: 11
Registered: November 2007

Junior Member

Hi,

Below, I show output commands

[root@srv002 afrare]# vzctl exec 108 df -h
Filesystem Size Used Avail Use% Mounted on
/dev/simfs 4.0G 1.6G 1.2G 58% /
tmpfs 314M 8.0K 314M 1% /dev
[root@srv002 afrare]# vzctl exec 108 df -i
Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/simfs 800000 72393 727607 10% /
tmpfs 80177 899 79278 2% /dev

Please, see log.

====================================================
[root@srv002 afrare]# vzctl exec 108 tail -f /var/log/secure
Apr 4 08:18:01 srv008 crond[26090]: PAM unable to dlopen(/lib/security/pam_unix.so)
Apr 4 08:18:01 srv008 crond[26090]: PAM [error: libcrack.so.2: cannot open shared object file: Cannot allocate memory]
Apr 4 08:18:01 srv008 crond[26090]: PAM adding faulty module: /lib/security/pam_unix.so
Apr 4 08:18:01 srv008 crond[26090]: PAM unable to dlopen(/lib/security/pam_access.so)
Apr 4 08:18:01 srv008 crond[26090]: PAM [error: libnsl.so.1: cannot open shared object file: Cannot allocate memory]
Apr 4 08:18:01 srv008 crond[26090]: PAM adding faulty module: /lib/security/pam_access.so
Apr 4 08:22:18 srv008 sshd[18394]: Received signal 15; terminating.
Apr 4 08:22:18 srv008 sshd[20330]: pam_unix(sshd:session): session closed for user afrare
Apr 4 08:22:36 srv008 sshd[28118]: Server listening on 0.0.0.0 port 22.
Apr 4 08:30:01 srv008 crond[29861]: PAM audit_open() failed: Cannot allocate memory

====================================================

Thanks,

Report message to a moderator

Re: Problem dcachesize [message #29077 is a reply to message #29075]

Fri, 04 April 2008 11:44

maratrus
Messages: 1495
Registered: August 2007
Location: Moscow

Senior Member

And also "cat /proc/user_beancounters". I wasn't be able to find it.

Report message to a moderator

Re: Problem dcachesize [message #29079 is a reply to message #29077]

Fri, 04 April 2008 11:48

afrare
Messages: 11
Registered: November 2007

Junior Member

Please, see below.

108: kmemsize 5301822 5807381 22020096 23488096 0
lockedpages 0 0 32 32 0
privvmpages 60235 68152 264144 524288 0
shmpages 59 715 8192 8192 0
dummy 0 0 0 0 0
numproc 53 68 200 200 0
physpages 19571 22052 0 2147483647 0
vmguarpages 0 0 65536 2147483647 0
oomguarpages 19571 22052 6144 2147483647 0
numtcpsock 14 16 280 280 0
numflock 6 9 100 110 0
numpty 0 0 16 16 0
numsiginfo 0 11 256 256 0
tcpsndbuf 125216 0 3194880 5242880 0
tcprcvbuf 229376 0 3194880 5242880 0
othersockbuf 213768 230340 1320960 3368960 0
dgramrcvbuf 0 8380 132096 132096 0
numothersock 133 144 280 280 0
dcachesize 1047132 1052088 1050000 1100000 634
numfile 1401 1672 4096 4096 0
dummy 0 0 0 0 0
dummy 0 0 0 0 0
dummy 0 0 0 0 0
numiptent 6 6 128 128 0

Report message to a moderator

Re: Problem dcachesize [message #29081 is a reply to message #29079]

Fri, 04 April 2008 12:07

maratrus
Messages: 1495
Registered: August 2007
Location: Moscow

Senior Member

Hi,

Quote:

I incremented parameter dcachesize, but I continue receive failcnt.

If you continue receive failcnt increase dcachesize parameter until failcnt stops increasing.
Try to set a huge value of dcachesize then look at its consumption and then decrease it to appropriate value.
If it is not help we'll continue our investigations.

Report message to a moderator

Re: Problem dcachesize [message #29083 is a reply to message #29081]

Fri, 04 April 2008 12:24

afrare
Messages: 11
Registered: November 2007

Junior Member

Dear Friend,

See -> cat /proc/user_beancounters

dcachesize 1049256 1054920 1050000 2147483647 660

I don't understand why it occour.

Thanks again

Adriano

Report message to a moderator