[Pacemaker] Pacemaker failover problem

[Erich Weiler]

I think I may have found an answer.  I had this in my config:

order LDAP-after-IP inf: LDAP-IP LDAP-clone

And, according to the logs, it *looks* like what happens when 
genome-ldap1 goes gown, the IP goes over to genome-ldap2, AND THEN tries 
to start LDAP there, even though LDAP is already started there because 
it is an anonymous clone.  LDAP cannot start (because it is already 
started) and throws an error exit code, and presumably pacemaker freaks 
out because of that and shuts down LDAP on all nodes.  Then the floating 
IP disappears because of the line:

colocation LDAP-with-IP inf: LDAP-IP LDAP-clone

which is expected at that point.  It seems that when I tested this with 
older versions of pacemaker, this didn't happen.  Should 'order' 
statements be avoided entirely when dealing with anonymous clones?  Is 
that behavior expected?

Thanks,
erich

Erich Weiler wrote:
> I tried that (after setting 'property symmetric-cluster="true"'), didn't 
> seem to make a difference...
> 
> Thanks for the suggestion though!
> 
> hj lee wrote:
>> I don't think it's a good idea to put location on clone. The clone is 
>> designed to run equally on every node. Please remove these and see 
>> that helps.
>>
>> location LDAP-IP-placement-2 LDAP-IP 50: genome-ldap2
>> location LDAP-placement-1 LDAP-clone 100: genome-ldap1
>> location LDAP-placement-2 LDAP-clone 100: genome-ldap2
>> colocation LDAP-with-IP inf: LDAP-IP LDAP-clone
>>
>> hj
>>
>> On Mon, Mar 8, 2010 at 1:34 PM, Erich Weiler <weiler at soe.ucsc.edu 
>> <mailto:weiler at soe.ucsc.edu>> wrote:
>>
>>     Hi All,
>>
>>     I have a (hopefully) simple problem that I need to fix, but I feel
>>     like I'm missing a key concept here that is causing problems.  I
>>     have 2 nodes, genome-ldap1 and genome-ldap2.  Using latest corosync,
>>     pacemaker and openais from the epel and clusterlabs repos, CentOS 
>> 5.4.
>>
>>     Both nodes are running OpenLDAP daemons, genome-ldap1 is the primary
>>     and genome-ldap2 is the replication slave.  Both services are simple
>>     LSB services, and I want them both active all the time.  Here's my
>>     CRM config:
>>
>>     node genome-ldap1
>>     node genome-ldap2
>>     primitive LDAP lsb:ldap \
>>            op monitor interval="10s" timeout="15s" \
>>            meta target-role="Started"
>>     primitive LDAP-IP ocf:heartbeat:IPaddr2 \
>>            params ip="10.1.1.83" nic="eth0" cidr_netmask="16" \
>>            op monitor interval="30s" timeout="20s" \
>>            meta target-role="Started"
>>     clone LDAP-clone LDAP \
>>            meta clone-max="2" clone-node-max="1" globally-unique="false"
>>     location LDAP-IP-placement-1 LDAP-IP 100: genome-ldap1
>>     location LDAP-IP-placement-2 LDAP-IP 50: genome-ldap2
>>     location LDAP-placement-1 LDAP-clone 100: genome-ldap1
>>     location LDAP-placement-2 LDAP-clone 100: genome-ldap2
>>     colocation LDAP-with-IP inf: LDAP-IP LDAP-clone
>>     order LDAP-after-IP inf: LDAP-IP LDAP-clone
>>     property $id="cib-bootstrap-options" \
>>            dc-version="1.0.7-d3fa20fc76c7947d6de66db7e52526dc6bd7d782" \
>>            cluster-infrastructure="openais" \
>>            expected-quorum-votes="2" \
>>            stonith-enabled="false" \
>>            symmetric-cluster="false" \
>>            no-quorum-policy="ignore" \
>>            last-lrm-refresh="1268024522"
>>
>>
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> Pacemaker mailing list
>> Pacemaker at oss.clusterlabs.org
>> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>