[Pacemaker] understanding resource restarts through pengine

Oualid Nouri o.nouri at computer-lan.de
Sat Sep 24 04:22:26 EDT 2011

> Sep 20 13:18:41 tnode2 pengine: [3116]: notice: unpack_rsc_op:
> Operation res_drbd_1:1_monitor_0 found resource res_drbd_1:1 active on
> tnode1
> When tnode1 came back online, the cluster found that drbd was already
> running.
> Do you have it configured to start at boot time?

Yes ...  :-(
I'm testing now with disabled start at boot time.
hm...it seems to work better :-)

I didn't recognized the notice of the pengine as "important" for my failure scenario.

When there is a clean state (no previous failures or from standby mode) before the cluster starts all resources starts without complains and making problems.

Is it right to say that the activated DRBD-device (after a failure) is interpreted as an undefined resource state and pacemaker tries to get an defined state on depending resources?

Thank you!

More information about the Pacemaker mailing list