Hi, On Mon, Sep 09, 2013 at 02:35:35PM +0200, Andreas Mock wrote: > Hi Dejan, > > > > thank you for spending your time. > > > > To your question: > > Sorry, but it was not intended to have two meta-attributes sections. I just > overlooked it even after looking at the config snippet for many time. > The reason for that was, that I assumed (yes, yes, IT and assumptions, > two things which don't get together well) crm will aggregate > several 'meta' commands to one meta-attributes section. This > assumption is wrong as I have learned now.
Obviously unexpected too. This is an unusual case which hasn't so far been reported. > It also didn't come to awareness as 'crm resource start' and > 'crm resource stop' seem to work as expected. They just put > a <nvpair name="target-role" value="Started" id="blabla"/> > to all meta-attribute-sections. Besides me being wrong, > this part of the behaviour of crm seems not totally ok > for me. The initial loading of the config inserted one > line <nvpair name="target-role" value="Started" id="blabla"/> > into the cib whereas "start" and "stop" insert this line > in any meta-attributes section. How would you choose into which of the meta-attributes section to insert the attribute? If you say: crm resource stop rsc, then we can assume that you want the resource stopped. Note that the shell doesn't have enough knowledge to figure out if only one of the meta attributes sets is active and if so which one. > Arguing that crm shall hide the cib innards and shall hinder > the user inserting crap I would expect something different. Me too :) > I'm not sure what I would expect because I don't know which > seems valid vor the cib, pacemaker and commandline tools: > > a) When pacemaker and all other commandline tool can live > nicely with multiple meta-attributes sections (it seems to > be allowed by the xml definition) and address all nvpairs > just by name beneath this tag, than crm should insert or > change only one instance of that nvpair. Demoting via crm_resource > should have worked with this, shouldn't it? If there are multiple attributes sets, then you need to specify the id if you want crm_resource to operate on any of these. > b) When the duplication of the meta-attriutes section by > itself is problematic I would expect crm refusing or > at least warning about that construct. Right, it should do something about it. > What do you think? Since multiple attribute sets are allowed, then we need to support them too. However, in a case like this one, I guess that we should've put together the two attribute sets. Could you please open a bug report. It seems like we'd need to track the discussion and possible solutions. Cheers, Dejan > Best regards > > Andreas Mock > > > > -----Ursprüngliche Nachricht----- > Von: linux-ha-boun...@lists.linux-ha.org > [mailto:linux-ha-boun...@lists.linux-ha.org] Im Auftrag von Dejan > Muhamedagic > Gesendet: Montag, 9. September 2013 13:26 > An: 'General Linux-HA mailing list' > Betreff: Re: [Linux-HA] Problem with crm 1.2.6 rc 3 and older > > > > Hi Andreas, > > > > On Fri, Sep 06, 2013 at 02:20:55PM +0200, Andreas Mock wrote: > > > Hi all, hi Kristoffer, hi Dejan, > > > > > > I do have the following problem with crm 1.2.6 rc3 and older > > > in the context of ms resources. > > > > > > When I declare the following config: > > > > > > primitive r_drbd_fodb ocf:linbit:drbd \ > > > params drbd_resource="fodb" drbdconf="/usr/local/etc/drbd.conf" \ > > > op monitor interval="15s" timeout="120s" role="Master" \ > > > op monitor interval="45s" timeout="120s" role="Slave" \ > > > op start timeout="240" \ > > > op stop timeout="240" > > > ms ms_drbd_fodb r_drbd_fodb \ > > > meta master-max="1" master-node-max="1" \ > > > clone-max="2" clone-node-max="1" \ > > > notify="true" \ > > > meta target-role="Stopped" migration-threshold="2" > > > > > > then two meta-attributes sections are inserted into the cib: > > > > Is this what you really intented, i.e. to have two meta > > attributes sets? Or was it accidental? > > > > > <meta_attributes id="ms_drbd_fodb-meta_attributes"> > > > <nvpair name="master-max" value="1" > > > id="ms_drbd_fodb-meta_attributes-master-max"/> > > > <nvpair name="master-node-max" value="1" > > > id="ms_drbd_fodb-meta_attributes-master-node-max"/> > > > <nvpair name="clone-max" value="2" > > > id="ms_drbd_fodb-meta_attributes-clone-max"/> > > > <nvpair name="clone-node-max" value="1" > > > id="ms_drbd_fodb-meta_attributes-clone-node-max"/> > > > <nvpair name="notify" value="true" > > > id="ms_drbd_fodb-meta_attributes-notify"/> > > > </meta_attributes> > > > <meta_attributes id="ms_drbd_fodb-meta_attributes-0"> > > > <nvpair name="target-role" value="Stopped" > > > id="ms_drbd_fodb-meta_attributes-0-target-role"/> > > > <nvpair name="migration-threshold" value="2" > > > id="ms_drbd_fodb-meta_attributes-0-migration-threshold"/> > > > </meta_attributes> > > > > > > You can see one <nvpair name="target-role"/>. > > > > > > After a 'crm resource start ms_drbd_fodb' you see the following: > > > > > > <meta_attributes id="ms_drbd_fodb-meta_attributes"> > > > <nvpair name="master-max" value="1" > > > id="ms_drbd_fodb-meta_attributes-master-max"/> > > > <nvpair name="master-node-max" value="1" > > > id="ms_drbd_fodb-meta_attributes-master-node-max"/> > > > <nvpair name="clone-max" value="2" > > > id="ms_drbd_fodb-meta_attributes-clone-max"/> > > > <nvpair name="clone-node-max" value="1" > > > id="ms_drbd_fodb-meta_attributes-clone-node-max"/> > > > <nvpair name="notify" value="true" > > > id="ms_drbd_fodb-meta_attributes-notify"/> > > > <nvpair id="ms_drbd_fodb-meta_attributes-target-role" > > > name="target-role" value="Started"/> > > > </meta_attributes> > > > <meta_attributes id="ms_drbd_fodb-meta_attributes-0"> > > > <nvpair name="target-role" value="Started" > > > id="ms_drbd_fodb-meta_attributes-0-target-role"/> > > > <nvpair name="migration-threshold" value="2" > > > id="ms_drbd_fodb-meta_attributes-0-migration-threshold"/> > > > </meta_attributes> > > > > > > HAAAAA: Now you have two entries name="target-role". > > > As long as I use a 'crm resource stop ms_drbd_fodb' it's no (?) problem as > > > you get the follwing: > > > > > > <meta_attributes id="ms_drbd_fodb-meta_attributes"> > > > <nvpair name="master-max" value="1" > > > id="ms_drbd_fodb-meta_attributes-master-max"/> > > > <nvpair name="master-node-max" value="1" > > > id="ms_drbd_fodb-meta_attributes-master-node-max"/> > > > <nvpair name="clone-max" value="2" > > > id="ms_drbd_fodb-meta_attributes-clone-max"/> > > > <nvpair name="clone-node-max" value="1" > > > id="ms_drbd_fodb-meta_attributes-clone-node-max"/> > > > <nvpair name="notify" value="true" > > > id="ms_drbd_fodb-meta_attributes-notify"/> > > > <nvpair id="ms_drbd_fodb-meta_attributes-target-role" > > > name="target-role" value="Stopped"/> > > > </meta_attributes> > > > <meta_attributes id="ms_drbd_fodb-meta_attributes-0"> > > > <nvpair name="target-role" value="Stopped" > > > id="ms_drbd_fodb-meta_attributes-0-target-role"/> > > > <nvpair name="migration-threshold" value="2" > > > id="ms_drbd_fodb-meta_attributes-0-migration-threshold"/> > > > </meta_attributes> > > > > > > You see, both incarnations are set to "Stopped". > > > > > > Now the bug "BUT": > > > > > > When I try a 'crm resource demote ms_drbd_fodb' insted of the stop, I get > > > the following: > > > > > > Multiple attributes match name=target-role > > > Value: Started (id=ms_drbd_fodb-meta_attributes-target-role) > > > Value: Started (id=ms_drbd_fodb-meta_attributes-0-target-role) > > > Error performing operation: Invalid argument > > > > > > what is definitely correct as shown by the cib-excerpt. > > > > > > So, I'm not sure whether the initial duplication is the problem or the > > > behaviour > > > in the demote case. > > > > Resource start/stop/manage/unmanage are handled by crmsh > > internally (because of possible children/parent relations; see > > crm options help manage-children), whereas for resource > > promote/demote crmsh invokes crm_resource. That's why the > > behaviour is different, that is crm_resource cannot handle > > multiple attributes sets. > > > > I guess that the two meta attributes sets should have been only > > one in the first place. Otherwise, I'm not quite sure how is this > > to be handled. > > > > Thanks, > > > > Dejan > > > > > Thank you in advance. > > > > > > Best regards > > > Andreas Mock > > > > > > > > > > > > _______________________________________________ > > > Linux-HA mailing list > > > <mailto:Linux-HA@lists.linux-ha.org> Linux-HA@lists.linux-ha.org > > > <http://lists.linux-ha.org/mailman/listinfo/linux-ha> > http://lists.linux-ha.org/mailman/listinfo/linux-ha > > > See also: <http://linux-ha.org/ReportingProblems> > http://linux-ha.org/ReportingProblems > > _______________________________________________ > > Linux-HA mailing list > > <mailto:Linux-HA@lists.linux-ha.org> Linux-HA@lists.linux-ha.org > > <http://lists.linux-ha.org/mailman/listinfo/linux-ha> > http://lists.linux-ha.org/mailman/listinfo/linux-ha > > See also: <http://linux-ha.org/ReportingProblems> > http://linux-ha.org/ReportingProblems > > _______________________________________________ > Linux-HA mailing list > Linux-HA@lists.linux-ha.org > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems _______________________________________________ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
