[Pacemaker] colocation and ordering for groups/clones/instances

[Andrew Beekhof]

On Fri, Oct 21, 2011 at 9:30 AM, gustavo panizzo <gfa> <gfa at zumbi.com.ar> wrote:
> still having issues
>
>
> this is the correct behavior
> Filesystem[3785]: INFO: Running stop for /dev/vgoradb/d208i0 on /oracle/d208i0
> Filesystem[3786]: INFO: Running stop for /dev/vgoradb/d209i0 on /oracle/d209i0
> Filesystem[3785]: INFO: Trying to unmount /oracle/d208i0
> Filesystem[3786]: INFO: Trying to unmount /oracle/d209i0
> Filesystem[3786]: INFO: unmounted /oracle/d209i0 successfully
> Filesystem[3785]: INFO: unmounted /oracle/d208i0 successfully
> Filesystem[3980]: INFO: Running start for /dev/vgoradb/d208i0 on /oracle/d208i0
> Filesystem[4017]: INFO: Running start for /dev/vgoradb/d209i0 on /oracle/d209i0
> Filesystem[4113]: INFO: Running start for /dev/vgoralog/d208i0 on /oracle/d208i0/archiveredo
> Filesystem[4216]: INFO: Running start for /dev/vgoralog/d209i0 on /oracle/d209i0/archiveredo
>
>
> incorrect behavior
> Filesystem[15460]: INFO: Running start for /dev/vgoradb/d209i0 on /oracle/d209i0 -> d209i0-fs
> Filesystem[15544]: INFO: Running start for /dev/vgoralog/d209i0 on /oracle/d209i0/archiveredo -> d209i0-oraarch
> Filesystem[15544]: ERROR: Couldn't mount filesystem /dev/vgoralog/d209i0 on /oracle/d209i0/archiveredo
> Filesystem[16523]: INFO: Running stop for /dev/vgoralog/d209i0 on /oracle/d209i0/archiveredo
> Filesystem[16574]: INFO: Running stop for /dev/vgoradb/d209i0 on /oracle/d209i0
> Filesystem[16574]: INFO: Trying to unmount /oracle/d209i0
> Filesystem[16574]: INFO: unmounted /oracle/d209i0 successfully
>
>
> group d208i0_storage d208i0-fs d208i0-oraarch
> clone d208i0-storage d208i0_storage
> colocation col_d208i0 +inf: HASI ora-bkp vgoradb vgoralog
> d208i0-storage d208i0-ip
> order ord_d208i0 +inf: HASI ora-bkp vgoradb vgoralog d208i0-storage
> d208i0-ip --> adding symmetrical=false does not change the behavior

The shell syntax for resources sets is broken, you'd have to show me the xml.
You need to specify sequential=true for the set itself.

>
>
> any clue? HASI, ora-bkp, vgoradb and vgoralog are clones of
> groups/resources
>
> this is a huge cluster (at least to me), 160 resources and growing
>
>
> thanks
>
> On Tue, Oct 04, 2011 at 08:21:57PM -0300, gustavo panizzo <gfa> wrote:
>> hi list
>>     what should i use to create dependencies? to a resource that run on
>> all the nodes?
>>
>> (resumed) required order is
>> clone     -> clone2    ->  RG
>> all boxes    all boxes     one box
>>
>> colocation col_name +inf: clone2 clone
>> colocation col_name +inf: group clone2
>> order ord_name +inf: clone2 group symmetrical=false
>> order ord_name +inf: clone clone2 symmetrical=false
>>
>>
>> when i restart a machine (not the one with RG on it), clone starts on it
>> RG gets stopped where is running until clone is fully up
>>
>> what am i missing? i've tried to use group:instance for colocation and
>> order rules but does not pass the syntax check
>>
>> 4 nodes cluster, suse 11 sp1
>> clone and clone2 are clones from (different) groups. RG is a resource
>> group
>>
>>
>>
>> thanks!
>>
>> --
>> 1AE0 322E B8F7 4717 BDEA  BF1D 44BB 1BA7 9F6C 6333
>>
>
> --
> 1AE0 322E B8F7 4717 BDEA  BF1D 44BB 1BA7 9F6C 6333
>
>
> _______________________________________________
> 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
>