[Pacemaker] Problem: No space left on device.

[Andrew Beekhof]

On Fri, Mar 5, 2010 at 3:38 PM, Kees <chkoehoorn at live.nl> wrote:
> Hi,
>
> When i start the cluster software with /etc/init.d/corosync start, i see the
> whole stack in my processlist:
>
> 31838 ?        Ssl    0:06 /usr/sbin/corosync
> 31849 ?        SLs    0:00  \_ /usr/lib/heartbeat/stonithd
> 31850 ?        S      0:02  \_ /usr/lib/heartbeat/cib
> 31851 ?        S      0:01  \_ /usr/lib/heartbeat/lrmd
> 31852 ?        S      0:00  \_ /usr/lib/heartbeat/attrd
> 31853 ?        S      0:00  \_ /usr/lib/heartbeat/pengine
> 31854 ?        S      0:00  \_ /usr/lib/heartbeat/crmd
>
> I looks like everything is running, but there is a problem:
>
> daemon.log:Mar  5 11:54:24 test1 cib: [23150]: ERROR: write_xml_file: Cannot
> open /var/lib/heartbeat/crm/cib.qFnnLt for writing: No space left on device
> (28)
> daemon.log:Mar  5 11:55:27 test1 pengine: [23145]: ERROR: write_xml_file:
> Cannot open /var/lib/pengine/pe-warn-418392.bz2 for writing: No space left
> on device (28)

You might want to set the pe-*-series-max options to limit the amount
of space used to store old PE inputs (used for debugging)
Looks like you have quite a few.

    </parameter>
    <parameter name="pe-error-series-max" unique="0">
      <shortdesc lang="en">The number of PE inputs resulting in ERRORs
to save</shortdesc>
      <content type="integer" default="-1"/>
      <longdesc lang="en">Zero to disable, -1 to store unlimited.</longdesc>
    </parameter>
    <parameter name="pe-warn-series-max" unique="0">
      <shortdesc lang="en">The number of PE inputs resulting in
WARNINGs to save</shortdesc>
      <content type="integer" default="-1"/>
      <longdesc lang="en">Zero to disable, -1 to store unlimited.</longdesc>
    </parameter>
    <parameter name="pe-input-series-max" unique="0">
      <shortdesc lang="en">The number of other PE inputs to save</shortdesc>
      <content type="integer" default="-1"/>
      <longdesc lang="en">Zero to disable, -1 to store unlimited.</longdesc>
    </parameter>


> daemon.log:Mar  5 11:55:28 test1 pengine: [23145]: ERROR: write_xml_file:
> Cannot open /var/lib/pengine/pe-warn-418393.bz2 for writing: No space left
> on device (28)
> daemon.log:Mar  5 11:55:28 test1 pengine: [23145]: ERROR: write_xml_file:
> Cannot open /var/lib/pengine/pe-warn-418394.bz2 for writing: No space left
> on device (28)
> daemon.log:Mar  5 11:55:28 test1 lrmd: [23143]: info: RA output:
> (ip_storage:start:stderr) info: Could not open pid-file
> [/var/run/heartbeat/rsctmp/send_arp/send_arp-172.16.0.4]: No space left on
> device
> daemon.log:Mar  5 11:55:28 test1 send_arp: [23358]: info: Could not open
> pid-file [/var/run/heartbeat/rsctmp/send_arp/send_arp-172.16.0.4]: No space
> left on device
> daemon.log:Mar  5 12:13:18 test1 cib: [24900]: ERROR: write_xml_file: Cannot
> open /var/lib/heartbeat/crm/cib.2rfyDF for writing: No space left on device
> (28)
> daemon.log:Mar  5 12:19:11 test1 lrmd: [24894]: info: RA output:
> (ip_storage:start:stderr) info: Could not open pid-file
> [/var/run/heartbeat/rsctmp/send_arp/send_arp-172.16.0.4]: No space left on
> device
> daemon.log:Mar  5 12:19:11 test1 send_arp: [26746]: info: Could not open
> pid-file [/var/run/heartbeat/rsctmp/send_arp/send_arp-172.16.0.4]: No space
> left on device
> daemon.log:Mar  5 12:25:47 test1 lrmd: [24894]: info: RA output:
> (drbd_websites:0:start:stderr) symlink(/etc/drbd.conf,
> /var/lib/drbd//drbd-minor-0.conf): No space left on device
> daemon.log:Mar  5 12:25:47 test1 lrmd: [24894]: info: RA output:
> (drbd_websites:0:start:stderr) symlink(/etc/drbd.conf,
> /var/lib/drbd//drbd-minor-0.conf): No space left on device
> daemon.log:Mar  5 12:25:47 test1 lrmd: [24894]: info: RA output:
> (drbd_websites:0:start:stderr) symlink(/etc/drbd.conf,
> /var/lib/drbd//drbd-minor-0.conf): No space left on device
> daemon.log:Mar  5 12:25:47 test1 lrmd: [24894]: info: RA output:
> (drbd_websites:0:notify:stderr) symlink(/etc/drbd.conf,
> /var/lib/drbd//drbd-minor-0.conf): No space left on device
> daemon.log:Mar  5 12:25:47 test1 lrmd: [24894]: info: RA output:
> (drbd_websites:0:notify:stderr) symlink(/etc/drbd.conf,
> /var/lib/drbd//drbd-minor-0.conf): No space left on device
> daemon.log:Mar  5 12:25:48 test1 lrmd: [24894]: info: RA output:
> (drbd_websites:0:notify:stderr) symlink(/etc/drbd.conf,
> /var/lib/drbd//drbd-minor-0.conf): No space left on device
> daemon.log:Mar  5 12:25:48 test1 lrmd: [24894]: info: RA output:
> (drbd_websites:0:promote:stderr) symlink(/etc/drbd.conf,
> /var/lib/drbd//drbd-minor-0.conf): No space left on device
> daemon.log:Mar  5 12:25:48 test1 lrmd: [24894]: info: RA output:
> (drbd_websites:0:promote:stderr) symlink(/etc/drbd.conf,
> /var/lib/drbd//drbd-minor-0.conf): No space left on device
> daemon.log:Mar  5 12:25:48 test1 lrmd: [24894]: info: RA output:
> (drbd_websites:0:notify:stderr) symlink(/etc/drbd.conf,
> /var/lib/drbd//drbd-minor-0.conf): No space left on device
> daemon.log:Mar  5 12:25:48 test1 lrmd: [24894]: info: RA output:
> (drbd_websites:0:monitor:stderr) symlink(/etc/drbd.conf,
> /var/lib/drbd//drbd-minor-0.conf): No space left on device
>
> Somehow my /var partion is not writeable anymore. When i try it myself with
> a 'touch testfile' i get the same error:
>
> touch: cannot touch `testfile': No space left on device
>
> When i stop the cluster, i can write again to /var. I can't find the
> problem, what is going wrong here?
>
> Debian leny
>
> Filesystem            Size  Used Avail Use% Mounted on
> /dev/sda5             942M  116M  779M  13% /
> /dev/sda1             942M   38M  857M   5% /boot
> /dev/sda6             942M   18M  877M   2% /home
> /dev/sda10            1.9G   35M  1.8G   2% /tmp
> /dev/sda7             1.9G  593M  1.2G  34% /usr
> /dev/sda8             1.9G  894M  888M  51% /var
> /dev/sda9             1.9G   57M  1.7G   4% /var/log
> /dev/drbd0            102G  188M   97G   1% /websites
>
> corosync_1.2.0-1_i386.deb
> pacemaker_1.0.7+hg20100203-1_i386.deb
>
> I use the Debian-packages from madkiss.
>
>
> Greeting,
>
> Kees
>
>
>
>
> Thanks in advance for your help.
>
> Kees Koehoorn
>
> _______________________________________________
> Pacemaker mailing list
> Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>