Hi,

I noticed a strange behavior in heartbeat 1.2.5 during resource
transition from passive node to active node (passive/active cluster) -
when the passive node releases its resources.
Normally (as I understand) the passive node doesn't have any resources
(it is _passive_) to release and therefore the active node doesn't need
to aquire any. But it seems that the active node starts all resources it
finds in its haresources. This leads to a broken/cancelled startup
process (the resources are alreay started -> startup scripts return
with exit code 1) and finally to a shutdown of all resources. In the
end I have two passive nodes and no running resources at all just
because of this resource aquistion. 
I'm able to reproduce this at any time by invoking hb_standby on the
the passive node.

So, is this intended behaviour or a bug or perhaps just a configuration
problem? I couldn't find any clue on the linux-ha.org page that's why
I'm asking here.


haresources:
> node1   drbddisk::drbd0 \
>         Filesystem::/dev/drbd0::/home/share::ext3 \
>         quagga

ha.cf:
> debugfile /var/log/ha/ha-debug
> logfile /var/log/ha/ha-log
> logfacility daemon
> keepalive 500ms
> warntime 15
> deadtime 30
> deadping 30
> initdead 30
> udpport 694
> ucast eth3 10.10.10.2
> ping 192.168.0.1
> auto_failback off
> node node1
> node node2
> respawn hacluster /usr/local/heartbeat/lib/heartbeat/ipfail
> debug 0


Thanks in advance,

Alex


-- 
Artfiles New Media GmbH | Spaldingstr. 160 b | 20097 Hamburg
Tel: 040 - 32 02 72 90 | Fax: 040 - 32 02 72 95
E-Mail: [EMAIL PROTECTED] | Web: http://www.artfiles.de
Geschäftsführer Carsten Bals | Handelsregister Hamburg - HRB 81478
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to