[Pacemaker] Patch for bugzilla 2541: Shell should warn if parameter uniqueness is violated

Dejan Muhamedagic dejanmm at fastmail.fm
Wed Jun 29 12:35:38 UTC 2011


Hi,

On Tue, Jun 28, 2011 at 06:10:20PM -0700, Bob Haxo wrote:
> Well, never mind ...
> 
> There appears to be an easy fix for this "violate uniqueness for
> parameter" issue with stonith:/external/ipmi.  As noted earlier in this
> thread WRT fence_ipmilan, the parmeters in stonith:external/ipmi have
> all been set to "unique".  Unsetting the uniqueness seems to solve the
> issue.
> 
> For example, setting unique to 0:
> 
> <parameter name="userid" unique="0">

Yes. Alternatively, you can enforce changes using the -F option
(crm -F ...).

> Is proper protocol for me to file a bug with Novell as this is code from
> Novell?

Yes, please, Novell bugzilla is a good place.

Thanks,

Dejan

> Regards,
> Bob Haxo
> 
> 
> On Tue, 2011-06-28 at 17:53 -0700, Bob Haxo wrote:
> > All,
> > 
> > I'm seeing a problem referred to in this thread:
> > 
> > crm(live)configure# primitive stonith-ipmi-1 stonith:external/ipmi \
> > > meta target-role="started" is-managed="true" \
> > > operations $id="stonith-ipmi-1-operations" \
> > > op monitor interval="15" timeout="15" on-fail="restart" start-delay="15" \
> > > op start interval="0" timeout="15" on-fail="restart" \
> > > params hostname="r1lead2-bmc" ipaddr="172.24.0.10" userid="XXXXX" passwd="XXXXX" passwd_method="param" interface="lanplus"
> > 
> > crm(live)configure# commit
> > 
> > crm(live)configure# primitive stonith-ipmi-2 stonith:external/ipmi \
> > > meta target-role="started" is-managed="true" \
> > > operations $id="stonith-ipmi-2-operations" \
> > > op monitor interval="15" timeout="15" on-fail="restart" start-delay="15" \
> > > op start interval="0" timeout="15" on-fail="restart" \
> > > params hostname="r1lead1-bmc" ipaddr="172.24.0.7" userid="XXXXX" passwd="XXXXX" passwd_method="param" interface="lanplus"
> > 
> > crm(live)configure# commit
> > WARNING: Resources stonith-ipmi-1,stonith-ipmi-2 violate uniqueness for parameter "passwd": "XXXXX"
> > WARNING: Resources stonith-ipmi-1,stonith-ipmi-2 violate uniqueness for parameter "passwd_method": "param"
> > WARNING: Resources stonith-ipmi-1,stonith-ipmi-2 violate uniqueness for parameter "interface": "lanplus"
> > WARNING: Resources stonith-ipmi-1,stonith-ipmi-2 violate uniqueness for parameter "userid": "XXXXX"
> > Do you still want to commit? yes
> > crm(live)configure#
> > 
> > The problem is that I want to perform this operation as "crm <
> > commands", which returns the warnings and assumes a "NO" to the
> > question.
> > 
> > I have current Novell updates applied, which includes the cibconfig.py
> > d0472a26eda1 patch. I see no Novell bug for this issue.
> > 
> > Have I missed the fix under a different thread?
> > 
> > Any suggestions will be appreciated.  Thanks.
> > 
> > Bob Haxo
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > On Fri, 2011-04-01 at 10:33 +0200, Andrew Beekhof wrote:
> > > On Tue, Mar 29, 2011 at 6:03 PM, Vladislav Bogdanov
> > > <bubble at hoster-ok.com> wrote:
> > > > Hi,
> > > >
> > > > 29.03.2011 18:33, Dejan Muhamedagic wrote:
> > > >> Hi,
> > > >>
> > > >> On Sat, Mar 26, 2011 at 08:41:45AM +0200, Vladislav Bogdanov wrote:
> > > >>> Oops, this is actually a bug in fence_ipmilan which reports all params
> > > >>> as unique.
> > > >>
> > > >> Well, that's bad, but then we don't want to annoy people until
> > > >> this issue has been resolved. The meta-data stuff usually needs
> > > >> a bit more attention, as far as I could see. I think that we'd
> > > >> need to make this check optional.
> > > >
> > > > I planned to post patch to redhat's bugzilla in nearest days (together
> > > > with bunch of other fixes and features). Should I send it (them)
> > > > somewhere else too?
> > > 
> > > Nah, RHBZ should be enough.
> > > 
> > > >
> > > >>
> > > >> What's your experience? Does it show these messages only once or
> > > >> all the time?
> > > >
> > > > Only once after resource either defined or changed.
> > > > It also cries something about start and stop operations, do not remember
> > > > now, sorry, short on time right now.
> > > >
> > > >
> > > > Vladislav
> > > >
> > > > _______________________________________________
> > > > Pacemaker mailing list: Pacemaker at 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://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
> > > >
> > > 
> > > _______________________________________________
> > > Pacemaker mailing list: Pacemaker at 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://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
> 
> 
> _______________________________________________
> Pacemaker mailing list: Pacemaker at 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://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker




More information about the Pacemaker mailing list