[Pacemaker] Question about behavior when pacemakerd was terminated
Kazunori INOUE
inouekazu at intellilink.co.jp
Mon May 7 09:53:07 UTC 2012
Hi,
I am using Pacemaker-1.1 (devel: db5e167).
When stonith-enabled is "false", and pacemakerd is terminated abnormally,
the states differ by whether a node is DC.
- (1-a) pacemakerd of DC node was terminated -> OFFLINE
- (1-b) pacemakerd of non-DC node was terminated -> UNCLEAN (offline)
Is this the designed behavior?
* 0. set "false" to stonith-enabled.
# crm configure show
node $id="1" vm1
node $id="2" vm2
primitive prmDummy ocf:pacemaker:Dummy \
op start interval="0s" timeout="60s" on-fail="restart" \
op monitor interval="10s" timeout="60s" on-fail="restart" \
op stop interval="0s" timeout="60s" on-fail="stop"
property $id="cib-bootstrap-options" \
dc-version="1.1.8-1.el6-db5e16736cc2682fbf37f81cd47be7d17d5a2364" \
cluster-infrastructure="corosync" \
no-quorum-policy="ignore" \
startup-fencing="false" \
stonith-enabled="false"
* 1-a. terminate pacemakerd on DC node (vm1) forcibly.
# crm_mon
Stack: corosync
Current DC: vm1 (1) - partition WITHOUT quorum
Version: 1.1.8-1.el6-db5e16736cc2682fbf37f81cd47be7d17d5a2364
2 Nodes configured, unknown expected votes
1 Resources configured.
============
Online: [ vm1 vm2 ]
prmDummy (ocf::pacemaker:Dummy): Started vm1
# ssh vm1 'pkill -9 pacemakerd'
# crm_mon
Current DC: vm2 (2) - partition WITHOUT quorum
============
Online: [ vm2 ]
OFFLINE: [ vm1 ]
prmDummy (ocf::pacemaker:Dummy): Started vm2
* 1-b. terminate pacemakerd on non-DC node (vm2) forcibly.
# crm_mon
Current DC: vm1 (1) - partition WITHOUT quorum
============
Online: [ vm1 vm2 ]
prmDummy (ocf::pacemaker:Dummy): Started vm2
# ssh vm2 'pkill -9 pacemakerd'
# crm_mon -r
Current DC: vm1 (1) - partition WITHOUT quorum
============
Node vm2 (2): UNCLEAN (offline)
Online: [ vm1 ]
Full list of resources:
prmDummy (ocf::pacemaker:Dummy): Stopped
----
Best regards,
Kazunori INOUE
More information about the Pacemaker
mailing list