Hi, On Fri, Sep 24, 2010 at 11:34:23AM -0600, Greg Woods wrote: > I've got two two-node clusters. They are not identical hardware but they > are both running CentOS 5.5 with the latest heartbeat and pacemaker from > the clusterlabs repo. I ran into a minor issue with a script I have that > does a lot of management for me. One of the things it does is parse the > output of "crm node show" and report on the health of the cluster. To my > surprise, the output of this command is different on the two machines. > The first cluster has an additional line per node reporting the standby > attribute. On the first cluster: > > # crm node show > vmserve2.scd.ucar.edu(16fde08d-b4b6-4550-adfb-b3aab83f706f): normal > standby: off > vmserve.scd.ucar.edu(6f5ced83-a790-4519-8449-3d4cf43275b0): normal > standby: off
Because the shell found these attributes in the node definition. Perhaps we shouldn't print them in case they match the default. > On the second cluster: > > # crm node show > vmx1.ucar.edu(62cf0a44-5d0f-475e-a0ac-689537f98f58): normal > vmx2.ucar.edu(8ad9076e-c571-499b-91e9-4d513fd5be61): normal > > This messes up the parsing, which wants to check to see if either node > is in standby mode. That is really only a minor problem as the second > cluster seems to be working fine, but I would like to know why is the > output different, so I checked "crm configure show" and see a similar > difference: > > node $id="16fde08d-b4b6-4550-adfb-b3aab83f706f" vmserve2.scd.ucar.edu \ > attributes standby="off" > node $id="6f5ced83-a790-4519-8449-3d4cf43275b0" vmserve.scd.ucar.edu \ > attributes standby="off" > > > node $id="62cf0a44-5d0f-475e-a0ac-689537f98f58" vmx1.ucar.edu > node $id="8ad9076e-c571-499b-91e9-4d513fd5be61" vmx2.ucar.edu > > So I figure, OK, I'll just add the standby="off" attribute with "crm > configure edit", but when I try to do that, I get this error when trying > to commit: > > > element instance_attributes: Relax-NG validity error : Type ID doesn't > allow value '62cf0a44-5d0f-475e-a0ac-689537f98f58-instance_attributes' > element nvpair: Relax-NG validity error : Element instance_attributes > failed to validate content > element instance_attributes: Relax-NG validity error : Element node has > extra content: instance_attributes > element node: Relax-NG validity error : Element nodes has extra content: > node > element configuration: Relax-NG validity error : Invalid sequence in > interleave > element cib: Relax-NG validity error : Element cib failed to validate > content > crm_verify[23031]: 2010/09/24_11:26:46 ERROR: main: CIB did not pass > DTD/schema validation > Errors found during check: config not valid The schema doesn't allow ids to start with a number. There's an exception for node ids, to preserve compatibility with CIBs coming from Heartbeat. The shell uses the id to build the id for the instance attributes set. Changed now to use the uname instead, so this won't happen anymore. Thanks, Dejan > So why is this behavior different between the two clusters? Is it > because the hardware is not identical? Does it indicate that anything is > wrong? Is there something I can do to get the second cluster to display > the standby attribute? > Thanks, > --Greg > > > _______________________________________________ > Linux-HA mailing list > [email protected] > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
