[Pacemaker] [Problem]colocation condition does not become effective.

[renayama19661014 at ybb.ne.jp]

Hi All,

We put pgsql resource of Master/Slave and original RA together and constituted
a cluster.

However, the problem that colocation condition did not become effective
happened.
Therefore promote is carried out.

Step1) Start a node not to be used to Master earlier.
Step2) VIPCheck resources start.
Step3) Start a node to fit a master next.

Because there are VIPCheck resource and colocation condition, promote cannot be
carried out.

(snip)
      <rsc_colocation id="rsc_colocation-9" rsc="msPostgresql"
rsc-role="Master" score="INFINITY" with-rsc="vipCheck"/>
(snip)

However, in violation of colocation condition, promote is carried out.

---crm_mon-------------
============
Last updated: Mon Jun 11 18:58:35 2012
Stack: Heartbeat
Current DC: 03-sl6 (fa2f3fed-c63f-4dfb-b0c1-bd02a49ef9b9) - partition with
quorum
Version: 1.0.12-066152e
2 Nodes configured, unknown expected votes
7 Resources configured.
============

Online: [ 02-sl6 03-sl6 ]

 vip-slave    (ocf::heartbeat:IPaddr2):    Started 02-sl6
 vipCheck    (ocf::heartbeat:VIPcheck):    Started 03-sl6
 Master/Slave Set: msPostgresql
     Masters: [ 02-sl6 ]
     Slaves: [ 03-sl6 ]
(snip)

---ptest--------
[root at rh62-test1 ~]# ptest -x pe-input-778 -VVV
ptest[5261]: 2012/06/12_22:48:56 notice: unpack_config: On loss of CCM Quorum:
Ignore
ptest[5261]: 2012/06/12_22:48:56 WARN: unpack_nodes: Blind faith: not fencing
unseen nodes
ptest[5261]: 2012/06/12_22:48:56 notice: native_print: vip-slave       
(ocf::heartbeat:IPaddr2):       Stopped 
ptest[5261]: 2012/06/12_22:48:56 notice: native_print: vipCheck
(ocf::heartbeat:VIPcheck):      Started 03-sl6
ptest[5261]: 2012/06/12_22:48:56 notice: group_print:  Resource Group:
master-group
ptest[5261]: 2012/06/12_22:48:56 notice: native_print:      vip-master 
(ocf::heartbeat:IPaddr2):       Stopped 
ptest[5261]: 2012/06/12_22:48:56 notice: native_print:      vip-rep    
(ocf::heartbeat:IPaddr2):       Stopped 
ptest[5261]: 2012/06/12_22:48:56 notice: clone_print:  Master/Slave Set:
msPostgresql
ptest[5261]: 2012/06/12_22:48:56 notice: short_print:      Slaves: [ 03-sl6
02-sl6 ]
ptest[5261]: 2012/06/12_22:48:56 notice: clone_print:  Clone Set: clnDiskd1
ptest[5261]: 2012/06/12_22:48:56 notice: short_print:      Started: [ 03-sl6
02-sl6 ]
ptest[5261]: 2012/06/12_22:48:56 notice: clone_print:  Clone Set: clnDiskd2
ptest[5261]: 2012/06/12_22:48:56 notice: short_print:      Started: [ 03-sl6
02-sl6 ]
ptest[5261]: 2012/06/12_22:48:56 notice: clone_print:  Clone Set: clnPingd
ptest[5261]: 2012/06/12_22:48:56 notice: short_print:      Started: [ 03-sl6
02-sl6 ]
ptest[5261]: 2012/06/12_22:48:56 notice: RecurringOp:  Start recurring monitor
(9s) for postgresql:1 on 02-sl6
ptest[5261]: 2012/06/12_22:48:56 notice: RecurringOp:  Start recurring monitor
(9s) for postgresql:1 on 02-sl6
ptest[5261]: 2012/06/12_22:48:56 notice: LogActions: Leave   resource vip-slave
(Stopped)
ptest[5261]: 2012/06/12_22:48:56 notice: LogActions: Leave   resource vipCheck 
(Started 03-sl6)
ptest[5261]: 2012/06/12_22:48:56 notice: LogActions: Leave   resource
vip-master        (Stopped)
ptest[5261]: 2012/06/12_22:48:56 notice: LogActions: Leave   resource vip-rep  
(Stopped)
ptest[5261]: 2012/06/12_22:48:56 notice: LogActions: Leave   resource
postgresql:0      (Slave 03-sl6)
ptest[5261]: 2012/06/12_22:48:56 notice: LogActions: Promote postgresql:1      
(Slave -> Master 02-sl6)
ptest[5261]: 2012/06/12_22:48:56 notice: LogActions: Leave   resource
prmDiskd1:0       (Started 03-sl6)
ptest[5261]: 2012/06/12_22:48:56 notice: LogActions: Leave   resource
prmDiskd1:1       (Started 02-sl6)
ptest[5261]: 2012/06/12_22:48:56 notice: LogActions: Leave   resource
prmDiskd2:0       (Started 03-sl6)
ptest[5261]: 2012/06/12_22:48:56 notice: LogActions: Leave   resource
prmDiskd2:1       (Started 02-sl6)
ptest[5261]: 2012/06/12_22:48:56 notice: LogActions: Leave   resource
pingCheck:0       (Started 03-sl6)
ptest[5261]: 2012/06/12_22:48:56 notice: LogActions: Leave   resource
pingCheck:1       (Started 02-sl6)
----------------------------------------

We tried the same movement in Pacemaker1.1.7, but the problem occurred in the
same way.

---crm_mon-------------
============
Last updated: Mon Jun 11 19:52:59 2012
Last change: Mon Jun 11 19:49:30 2012 via crm_attribute on 16-sl6
Stack: Heartbeat
Current DC: 17-sl6 (706ba487-030a-41f8-bcb0-b8c6371bfb89) - partition with
quorum
Version: 1.1.7-512f868
2 Nodes configured, unknown expected votes
6 Resources configured.
============

Online: [ 16-sl6 17-sl6 ]

vip-slave       (ocf::heartbeat:IPaddr2):       Started 16-sl6
vipCheck        (ocf::heartbeat:VIPcheck):      Started 17-sl6
 Master/Slave Set: msPostgresql [postgresql]
     Masters: [ 16-sl6 ]
     Slaves: [ 17-sl6 ]
(snip)


 I registered these contents with Bugzilla.
  * http://bugs.clusterlabs.org/show_bug.cgi?id=5070

Best Regards,
Hideo Yamauchi.