OpenVZ Forum


Home » General » Support » KMEMSIZE high failcount - why?
KMEMSIZE high failcount - why? [message #2006] Mon, 13 March 2006 15:56 Go to next message
bjmg is currently offline  bjmg
Messages: 32
Registered: December 2005
Location: Puettlingen, Germany
Member

From: *dip.t-dialin.net
Hi,

I am using OpenVZ since about 3 months now. On a new host server (32bit system, 1 GB RAM, CentOS 4.2) I set up a debian VPS (based on the template provided on your FTP server).
Now the problem is that I get a REALLY high failcount for kmemsize. It is completely irrelevant what values for KMEMSIZE I use I always get nearly the same values. The problem is that the value is already that high about 1 second after start.
I really hope someone of the developers can help me.

First, here is the output of /proc/user_beancounters:
Version: 2.5
       uid  resource           held    maxheld    barrier      limit    failcnt
         0: kmemsize        1859563    2704948 2147483647 2147483647          0
            lockedpages           0          0 2147483647 2147483647          0
            privvmpages        2726      10843 2147483647 2147483647          0
            shmpages            668       1324 2147483647 2147483647          0
            dummy                 0          0 2147483647 2147483647          0
            numproc              31         48 2147483647 2147483647          0
            physpages          1460       6779 2147483647 2147483647          0
            vmguarpages           0          0 2147483647 2147483647          0
            oomguarpages       1488       6779 2147483647 2147483647          0
            numtcpsock            2          7 2147483647 2147483647          0
            numflock              1          5 2147483647 2147483647          0
            numpty                2          4 2147483647 2147483647          0
            numsiginfo            0          4 2147483647 2147483647          0
            tcpsndbuf          2228     138136 2147483647 2147483647          0
            tcprcvbuf             0     122540 2147483647 2147483647          0
            othersockbuf       5148      75752 2147483647 2147483647          0
            dgramrcvbuf           0       8912 2147483647 2147483647          0
            numothersock          8         20 2147483647 2147483647          0
            dcachesize      1025164    1076824 2147483647 2147483647          0
            numfile             140        328 2147483647 2147483647          0
            dummy                 0          0 2147483647 2147483647          0
            dummy                 0          0 2147483647 2147483647          0
            dummy                 0          0 2147483647 2147483647          0
            numiptent            10         10 2147483647 2147483647          0
     22114: kmemsize        6886100    7882846    9752512    9936012     773421
            lockedpages           0          0        128        128          0
            privvmpages       73406     112592     256000     262140          0
            shmpages            668       1978       8192       8192          0
            dummy                 0          0          0          0          0
            numproc             100        116        127        192          0
            physpages         25985      65609          0 2147483647          0
            vmguarpages           0          0     131072 2147483647          0
            oomguarpages      25986      66342       6144 2147483647          0
            numtcpsock           27         64        127        255          0
            numflock             17         23        100        110          0
            numpty                2          4         16         16          0
            numsiginfo            0         22        256        256          0
            tcpsndbuf         26736     188536     519488     924288          0
            tcprcvbuf             0     222800     519488     924288          0
            othersockbuf     181408     285344     532096     936896          0
            dgramrcvbuf           0      60156     332096     332096          0
            numothersock        147        171        255        300          0
            dcachesize       381207     445495    1048576    1097728          0
            numfile            1068       1377       3072       4096          0
            dummy                 0          0          0          0          0
            dummy                 0          0          0          0          0
            dummy                 0          0          0          0          0
            numiptent           143        143        128        256          0


Here is the output of uname -a:
Linux master.my-domain.net 2.6.8-022stab070.1-enterprise #1 SMP Mon Feb 20 19:31:28 MSK 2006 i686 athlon i386 GNU/Linux


Here is the config file of VPS 22114:
ONBOOT="yes"

# UBC parameters (in form of barrier:limit)
# Primary parameters
AVNUMPROC="70:80"
NUMPROC="127:192"
NUMTCPSOCK="127:255"
NUMOTHERSOCK="255:300"
VMGUARPAGES="131072:2147483647"
# Secondary parameters
KMEMSIZE="9752512:9936012"
TCPSNDBUF="519488:924288"
TCPRCVBUF="519488:924288"
OTHERSOCKBUF="532096:936896"
DGRAMRCVBUF="332096:332096"
OOMGUARPAGES="6144:2147483647"
# Auxiliary parameters
LOCKEDPAGES="128:128"
SHMPAGES="8192:8192"
PRIVVMPAGES="256000:262140"
NUMFILE="3072:4096"
NUMFLOCK="100:110"
NUMPTY="16:16"
NUMSIGINFO="256:256"
DCACHESIZE="1048576:1097728"
PHYSPAGES="0:2147483647"
NUMIPTENT="128:256"
# Disk quota parameters (in form of softlimit:hardlimit)
DISKSPACE="32048576:32153434"
DISKINODES="200000:220000"
QUOTATIME="0"

# CPU fair sheduler parameter
CPUUNITS="1000"
VE_ROOT="/vz/root/$VEID"
VE_PRIVATE="/vz/private/$VEID"
ALLOWREBOOT="yes"
OSTEMPLATE="debian-3.1"
ORIGIN_SAMPLE="vps.basic"
IP_ADDRESS="10.10.221.14"
HOSTNAME="web.my-domain.org"
QUOTAUGIDLIMIT="150"


I really hope this is enough information such that someone of the developers/supporters can help me.

Thanks

Bernhard

[Updated on: Mon, 13 March 2006 15:58]

Report message to a moderator

Re: KMEMSIZE high failcount - why? [message #2010 is a reply to message #2006] Mon, 13 March 2006 17:41 Go to previous messageGo to next message
kir is currently offline  kir
Messages: 1645
Registered: August 2005
Location: Moscow, Russia
Senior Member

From: *sw.ru
It happens...

Most probable reason is some application is trying to grab some high amount of memory. So, look for some strange memory-related messages from logs.

Another reason could be a stateful firewall inside a VPS -- if there are too many connections, netfilter can allocate quite a big amounts of kernel memory. Are there a lot of records in /proc/net/ip_conntrack file inside a VPS? Does it help to run a VPS without any iptables rules (you can actually clone a VPS and run its clone without iptables to check)?


Kir Kolyshkin
http://static.openvz.org/userbars/openvz-developer.png
Re: KMEMSIZE high failcount - why? [message #2023 is a reply to message #2006] Wed, 15 March 2006 11:27 Go to previous messageGo to next message
bjmg is currently offline  bjmg
Messages: 32
Registered: December 2005
Location: Puettlingen, Germany
Member

From: *dip.t-dialin.net
Thanks kir,

The problem was that some one my parameters were too low and mysql tried to allocate much more memory. Now I used vzsplit to generate the config and everything seems to work fine.

Bernhard
Re: KMEMSIZE high failcount - why? [message #24646 is a reply to message #2023] Fri, 07 December 2007 08:00 Go to previous message
powervps is currently offline  powervps
Messages: 54
Registered: December 2007
Member
From: 80.72.152*
vzsplit ? how
Previous Topic: CPU usage/load on host
Next Topic: big problem
Goto Forum:
  


Current Time: Sat Jul 11 04:45:24 GMT 2020