[Pacemaker] Newbie: No automatic resource failover but can move resources

[Ben Schmidt]

Hello Mailing List,

Problem solved , the location "rule?" caused it, I used "crm configure
edit" to remove the lines, that did the tick. So I really know that I
must do some reading to understand the tools and infrastructure of
pacemaker&corosync.

Thanks a lot anyway, I'm sure that I'll ask for help in the near future,

Ben

On Sat, May 7, 2011 at 3:39 PM, Ben Schmidt <crackhd2 at gmail.com> wrote:
> Hello,
>
> I just started with Pacemaker + corosync a couple hours ago so please
> excuse my major screw ups.
> What I have achieved so far is a two node Cluster that talks to each
> other and one resource on that Cluster. My Problem is that the
> resource, a failover-ip, doesn't get failed to the other node if I
> take down the node the resource is started on but I can move the
> resource.
>
> Status:
> ###############
> deb-cluster01:~# crm_mon --one-shot
> ============
> Last updated: Sat May  7 15:15:56 2011
> Stack: openais
> Current DC: deb-cluster02 - partition with quorum
> Version: 1.0.9-74392a28b7f31d7ddc86689598bd23114f58978b
> 2 Nodes configured, 2 expected votes
> 1 Resources configured.
> ============
>
> Online: [ deb-cluster01 deb-cluster02 ]
>
>  failover-ip    (ocf::heartbeat:IPaddr2):       Started deb-cluster01
> ###############
>
> This happens when I takedown deb-cluster01
> ###############
> deb-cluster02:/var/log# crm_mon --one-shot
> ============
> Last updated: Sat May  7 15:16:40 2011
> Stack: openais
> Current DC: deb-cluster02 - partition WITHOUT quorum
> Version: 1.0.9-74392a28b7f31d7ddc86689598bd23114f58978b
> 2 Nodes configured, 2 expected votes
> 1 Resources configured.
> ============
>
> Online: [ deb-cluster02 ]
> OFFLINE: [ deb-cluster01 ]
> ###############
>
>
> Here is my config:
> ###############
> deb-cluster02:/var/log# crm configure show
> node deb-cluster01 \
>        attributes standby="off"
> node deb-cluster02
> primitive failover-ip ocf:heartbeat:IPaddr2 \
>        params ip="10.0.2.250" cidr_netmask="22" \
>        op monitor interval="1s" \
>        meta is-managed="true"
> location cli-standby-failover-ip failover-ip \
>        rule $id="cli-standby-rule-failover-ip" -inf: #uname eq deb-cluster02
> property $id="cib-bootstrap-options" \
>        dc-version="1.0.9-74392a28b7f31d7ddc86689598bd23114f58978b" \
>        cluster-infrastructure="openais" \
>        expected-quorum-votes="2" \
>        stonith-enabled="false"
> rsc_defaults $id="rsc-options" \
>        resource-stickiness="100"
> ###############
>
> A Failover to deb-cluster02
> ###############
> deb-cluster01:~# crm_mon --one-shot  | grep Started
>  failover-ip    (ocf::heartbeat:IPaddr2):       Started deb-cluster01
> deb-cluster01:~# crm resource move failover-ip
> WARNING: Creating rsc_location constraint 'cli-standby-failover-ip'
> with a score of -INFINITY for resource failover-ip on deb-cluster01.
>        This will prevent failover-ip from running on deb-cluster01 until the
> constraint is removed using the 'crm_resource -U' command or manually
> with cibadmin
>        This will be the case even if deb-cluster01 is the last node in the cluster
>        This message can be disabled with -Q
> deb-cluster01:~# crm_mon --one-shot  | grep Started
>  failover-ip    (ocf::heartbeat:IPaddr2):       Started deb-cluster02
> ###############
>
> Here is the log of deb-cluster01 when failover-ip was running on the
> other node and I restarted the node: http://pastebin.com/ftLXLBfT
>
> I'm using a plain debian squeeze amd64, everything comes from the distribution.
>
>
> Could somebody please point me in a direction on where to look at.
>
> Thanks a lot,
>
> Ben
>