[Pacemaker] re-manage a resource

[andreas graeper]

(in addition)

i tried
 pcs resource start ms_drbd       # rc=0     ( but afaik this only removes
is-managed="false" if exists. but this meta-attribute does not exist)
 pcs resource manage ms_drbd  #  ms_drbd does not exist
 pcs resource manage drbd        # already managed

what kind of state is 'unmanaged'
 i thought there is a meta-attribute for each resource and it tells only
whether the resource shall be started / stopped by pacemaker.
 i could imagine
  failed_to_(start|stop|promote|demote|...) and i am able to intervene and
tell corosync/pacemaker about the new state
  and ask corosync/pacemaker to monitor that resource again and start all
the other pending resources after check, whether that resource actually
  in the state, that was reported ?!


what is corosync waiting for when i try to stop it while drbd is not
managed and all the other resources are stopped ?
log: pcmk_shutdown: Still waiting for crmd (pid=xxx, seq=6) to terminate
...

( man crm(7) ...
  This is a fake resource that details the options that can be configured
for the CRM Daemon. )
a fake resource that holds the global meta-attributes (options) ?
what is crmd waiting for ?
now corosync has gone, but all the pacemaker-services are still there ?!

is there a more gentle way than kill -9 and kill all that
pacemaker-serivces manually and stop drbd before starting corosync again.
which is the most central service ?
when i kill some of the services they get started again ? nice, but not in
this situation.


thanks
andreas


2013/6/21 andreas graeper <agraeper at googlemail.com>

> maybe i asked this before, but i could not find message + answer.
>
> when a resource gets unmanaged and the problems has gone, i want the
> resource get managed by pacemaker again. what is to do ?
>
> situation: only on node left (other ill)
>  drbd could not get promoted
>  now it is standalone,primary,uptodate but not managed and all other
> resource stopped (cause depend on drbd:master )
>
> thanks in advance
> andreas
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20130621/3a36f276/attachment-0003.html>