Hi, On Mon, Oct 01, 2012 at 03:12:51PM +0200, Jan Horníček wrote: > Thank you for suggestions. > > Finally I managed to rewrite the script, so it's capable of reloading. > But I have one question. > > Why, when I remove the first parameter definition (parameter > name=status) in metadata section of Dummy RA, the reload isn't performed
The metadata section? The metadata is not part of the resource definition. Otherwise, it should really do a reload. If it doesn't then file a bug report. Thanks, Dejan > and stop/start is applied instead? > > It was also the reason, why my RA wasn't reloading. It started to work > after adding one parameter definition before the original one. > > Thank you. > > Jan Hornicek > > > On 09/26/2012 09:21 AM, pacemaker-requ...@oss.clusterlabs.org wrote: > > Message: 2 > > Date: Tue, 25 Sep 2012 12:36:56 -0400 (EDT) > > From: David Vossel <dvos...@redhat.com> > > To: The Pacemaker cluster resource manager > > <pacemaker@oss.clusterlabs.org> > > Subject: Re: [Pacemaker] Reload implemention for RA > > Message-ID: <1441928510.645759.1348591016551.javamail.r...@redhat.com> > > Content-Type: text/plain; charset=utf-8 > > > > > > > > ----- Original Message ----- > >> > From: "Jan Horn??ek" <jan.horni...@cesnet.cz> > >> > To: pacemaker@oss.clusterlabs.org > >> > Sent: Tuesday, September 25, 2012 6:54:16 AM > >> > Subject: [Pacemaker] Reload implemention for RA > >> > > >> > Hello, > >> > I'm trying to implement reload functionality to ocf:sgi:cxfs-client > >> > script. The rewritten script is working if called from shell, it pass > >> > ocf-tester. When I change the parameter of the primitive based on > >> > that > >> > RA in the crm shell and commit the change I get restart (stop and > >> > then > >> > start) actions. When I try ocf:heartbeat:Dummy based primitive and > >> > change it's parameter I get reload action. > >> > unique="0" set to both parameters in meta-data > >> > pacemaker version 1.1.5 > > Pacemaker determines if a ocf resource is reloadable by reading the > > resource script's meta_data() output. Take a look at the pacemaker/Dummy > > script in a text editor. The meta_data() function returns documentation on > > what the agent is capable of in xml. Pacemaker is looking through the xml > > output of that meta_data() function for <actions><action > > name='reload'/></actions> to determine if the resource can reload. Then > > pacemaker looks through the <parameters/> tags for parameters with > > unique="0" to determine what parameters are allowed to be reloaded. > > > > Hope that helps. > > > > -- Vossel > > > >> > > >> > My questions are: > >> > > >> > 1) Is it possible to somehow debug the decision of the pengine for > >> > the > >> > task? (Something like reload not possible because ...) > >> > > >> > 2) Is there any way how to force the reload of changed primitive's > >> > parameters? > >> > > >> > 3) I'm not able to find more restrictions allowing reload to be > >> > functional except setting parameter unique="0" in it's meta-data > >> > definition of RA. Are there others? > >> > > >> > Thank you for your suggestions. > >> > > >> > Jan Hornicek > >> > > >> > > >> > > >> > > >> > _______________________________________________ > >> > Pacemaker mailing list: Pacemaker@oss.clusterlabs.org > >> > http://oss.clusterlabs.org/mailman/listinfo/pacemaker > >> > > >> > Project Home: http://www.clusterlabs.org > >> > Getting started: > >> > http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf > >> > Bugs: http://bugs.clusterlabs.org > >> > > > _______________________________________________ > Pacemaker mailing list: Pacemaker@oss.clusterlabs.org > http://oss.clusterlabs.org/mailman/listinfo/pacemaker > > Project Home: http://www.clusterlabs.org > Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf > Bugs: http://bugs.clusterlabs.org _______________________________________________ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org