[Pacemaker] why does ressource go down on both nodes

Dejan Muhamedagic dejanmm at fastmail.fm
Wed Jun 8 16:45:02 UTC 2011


Hi,

On Mon, Jun 06, 2011 at 05:23:34AM +0000, Guido Schmidt wrote:
> Hi to all,
> please can someone explain to me why resource OCFS2 in my configuration stopped 
> on node#2 by pacemakers-gui will also after a while stop upon node#1.

Can you please rephrase. I'm not sure I understand what happened.
If you ordered a resource to stop in the GUI, then it is going to
stop everywhere.

> For me for maintenance purpose it will be helpful to stop OCFS2, CLVG01 and 
> CLVM on one node only.
> So the whole applications (4 tomcats, apache, postfix, cyrus) will be serviced 
> on the other node (where the VIP1 will point to) during maintenance.

For this you want to set is-managed to false on the resource you
want to upgrade.

Thanks,

Dejan

> It seems to me, that it has worked as I described above some time ago;-)  I 
> really could not find the mistake in my configuration.
>  
> Following the crm-configuration I am using now (based on SLES-HA v11.1-64Bit):
>  
> node srv31084 \
>                 attributes standby="off"
> node srv31085 \
>                 attributes standby="off"
> primitive CLVG01 ocf:heartbeat:LVM \
>                 operations $id="CLVG01-operations" \
>                 op monitor interval="60" timeout="60" \
>                 params volgrpname="cl-vg01"
> primitive CLVM ocf:lvm2:clvmd \
>                 operations $id="CLVM-operations" \
>                 op monitor interval="60" timeout="60" \
>                 params daemon_timeout="30"
> primitive DLM ocf:pacemaker:controld \
>                 operations $id="DLM-operations" \
>                 op monitor interval="60" timeout="60" start-delay="0" \
>                 meta target-role="Started"
> primitive O2CB ocf:ocfs2:o2cb \
>                 operations $id="O2CB-operations" \
>                 op monitor interval="60" timeout="60"
> primitive OCFS2 ocf:heartbeat:Filesystem \
>                 operations $id="OCFS2-operations" \
>                 op monitor interval="20" timeout="40" \
>                 params device="/dev/cl-vg01/cl-b2b" directory="/opt/b2b" 
> fstype="ocfs2" options="acl"
> primitive SBD-STONITH stonith:external/sbd \
>                 operations $id="SBD-STONITH-operations" \
>                 op monitor interval="30" timeout="30" start-delay="15" \
>                 params sbd_device="/dev/sdb5"
> primitive VIP1 ocf:heartbeat:IPaddr2 \
>                 operations $id="VIP1-operations" \
>                 op monitor interval="20s" timeout="30s" \
>                 params ip="10.160.240.13" nic="eth0" cidr_netmask="32"
> group DLM-O2CB-CLVM-CLVG01-OCFS2 DLM O2CB CLVM CLVG01 OCFS2 \
>                 meta target-role="Started"
> group VIP-VTCX VIP1 \
>                 meta target-role="Started"
> clone FENCING SBD-STONITH \
>                 meta clone-max="2" target-role="Started"
> clone VINDEX DLM-O2CB-CLVM-CLVG01-OCFS2 \
>                 meta interleave="true" clone-max="2" clone-node-max="1"
> location cli-prefer-VIP1 VIP1 \
>                 rule $id="cli-prefer-rule-VIP1" inf: #uname eq srv31084
> location prefer-VIP-VTCX VIP-VTCX 50: srv31084
> property $id="cib-bootstrap-options" \
>                 dc-version="1.1.5-5ce2879aa0d5f43d01629bc20edc6868a9352002" \
>                 cluster-infrastructure="openais" \
>                 no-quorum-policy="ignore" \
>                 default-resource-stickiness="100" \
>                 last-lrm-refresh="1306977984" \
>                 stonith-timeout="30s" \
>                 cluster-recheck-interval="0" \
>                 expected-quorum-votes="2" \
> op_defaults $id="op_defaults-options" \
>                 record-pending="false"
>  
> Best regrads from Guido Schmidt
> 
> 
> 
> _______________________________________________
> 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