[Pacemaker] Question on resources' dependency and failover

Andrew Beekhof andrew at beekhof.net
Mon May 17 07:29:08 UTC 2010


The problem is that you have

      <rsc_colocation id="nfs_on_drbd0" rsc="nfs-group"
score="INFINITY" with-rsc="NfsData" with-rsc-role="Master"/>
and
      <rsc_colocation id="nfsclient_not_on_nfs-group" rsc="nfs-group"
score="-INFINITY" with-rsc="nfsclient"/>

This causes a problem when nfsclient can only run on the current drbd master.

Its best to linearize the colocation chain where possible, try this
instead of the second constraint:

      <rsc_colocation id="nfsclient_not_on_nfs-group" rsc="NfsData"
rsc-role="Master" score="-INFINITY" with-rsc="nfsclient"/>


On Wed, May 12, 2010 at 10:22 PM, Gianluca Cecchi
<gianluca.cecchi at gmail.com> wrote:
> On Wed, May 12, 2010 at 2:27 PM, Andrew Beekhof <andrew at beekhof.net> wrote:
>>
>> Attach the output from cibadmin -Ql when the cluster is in this state
>> and I'll take a look.
>>
>>
>
> Ok. here we are with the commands.
> I send attached the output of cibadmin command when the services stop after
> the move operation (cibadmin_stop.log) and after the unmove operation that
> causes them to restart (cibadmin_restart.log).
> Previous message with attachments in zip format seems didn't go to the list,
> perhaps.
>
> # crm_mon -fr1
> ============
> Last updated: Wed May 12 17:31:15 2010
> Stack: openais
> Current DC: ha1 - partition with quorum
> Version: 1.0.8-9881a7350d6182bae9e8e557cf20a3cc5dac3ee7
> 2 Nodes configured, 2 expected votes
> 5 Resources configured.
> ============
>
> Online: [ ha1 ha2 ]
>
> Full list of resources:
>
>  SitoWeb    (ocf::heartbeat:apache):    Started ha1
>  Master/Slave Set: NfsData
>      Masters: [ ha1 ]
>      Slaves: [ ha2 ]
>  Resource Group: nfs-group
>      lv_drbd0    (ocf::heartbeat:LVM):    Started ha1
>      ClusterIP    (ocf::heartbeat:IPaddr2):    Started ha1
>      NfsFS    (ocf::heartbeat:Filesystem):    Started ha1
>      nfssrv    (ocf::heartbeat:nfsserver):    Started ha1
>  nfsclient    (ocf::heartbeat:Filesystem):    Started ha2
>  Clone Set: cl-pinggw
>      Started: [ ha1 ha2 ]
>
> Migration summary:
> * Node ha1:  pingd=200
> * Node ha2:  pingd=200
>
> #crm resource move nfsclient
> WARNING: Creating rsc_location constraint 'cli-standby-nfsclient' with a
> score of -INFINITY for resource nfsclient on ha2.
>     This will prevent nfsclient from running on ha2 until the constraint is
> removed using the 'crm_resource -U' command or manually with cibadmin
>     This will be the case even if ha2 is the last node in the cluster
>     This message can be disabled with -Q
>
> # crm_mon -fr1
> ============
> Last updated: Wed May 12 17:34:14 2010
> Stack: openais
> Current DC: ha1 - partition with quorum
> Version: 1.0.8-9881a7350d6182bae9e8e557cf20a3cc5dac3ee7
> 2 Nodes configured, 2 expected votes
> 5 Resources configured.
> ============
>
> Online: [ ha1 ha2 ]
>
> Full list of resources:
>
>  SitoWeb    (ocf::heartbeat:apache):    Stopped
>  Master/Slave Set: NfsData
>      Masters: [ ha1 ]
>      Slaves: [ ha2 ]
>  Resource Group: nfs-group
>      lv_drbd0    (ocf::heartbeat:LVM):    Stopped
>      ClusterIP    (ocf::heartbeat:IPaddr2):    Stopped
>      NfsFS    (ocf::heartbeat:Filesystem):    Stopped
>      nfssrv    (ocf::heartbeat:nfsserver):    Stopped
>  nfsclient    (ocf::heartbeat:Filesystem):    Stopped
>  Clone Set: cl-pinggw
>      Started: [ ha1 ha2 ]
>
> Migration summary:
> * Node ha1:  pingd=200
> * Node ha2:  pingd=200
>
> # cibadmin -Ql > cibadmin_stop.log 2>&1
>
> # crm resource unmove nfsclient
>
> # crm_mon -fr1
> ============
> Last updated: Wed May 12 17:35:08 2010
> Stack: openais
> Current DC: ha1 - partition with quorum
> Version: 1.0.8-9881a7350d6182bae9e8e557cf20a3cc5dac3ee7
> 2 Nodes configured, 2 expected votes
> 5 Resources configured.
> ============
>
> Online: [ ha1 ha2 ]
>
> Full list of resources:
>
>  SitoWeb    (ocf::heartbeat:apache):    Started ha1
>  Master/Slave Set: NfsData
>      Masters: [ ha1 ]
>      Slaves: [ ha2 ]
>  Resource Group: nfs-group
>      lv_drbd0    (ocf::heartbeat:LVM):    Started ha1
>      ClusterIP    (ocf::heartbeat:IPaddr2):    Started ha1
>      NfsFS    (ocf::heartbeat:Filesystem):    Started ha1
>      nfssrv    (ocf::heartbeat:nfsserver):    Started ha1
>  nfsclient    (ocf::heartbeat:Filesystem):    Started ha2
>  Clone Set: cl-pinggw
>      Started: [ ha1 ha2 ]
>
> Migration summary:
> * Node ha1:  pingd=200
> * Node ha2:  pingd=200
>
> # cibadmin -Ql > cibadmin_restart.log
>
> Thanks in advance.
> Gianluca
> _______________________________________________
> 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
>




More information about the Pacemaker mailing list