[Pacemaker] [Problem]It does not start by single node configuration.

Andrew Beekhof andrew at beekhof.net
Thu Feb 4 07:31:25 EST 2010


2010/1/29  <renayama19661014 at ybb.ne.jp>:
> Hi Andrew,
>
>> > Could you possibly retest with debug on?
>> Please wait....
>
> I confirmed it in Pacemaker-1-0-9287969750e1 again.
> I attached the debug log.

Could you perhaps try with only a single ring?
It shouldn't make a difference but one never knows...

Almost certainly its a network issue... perhaps VMware's host-only
NICs don't do broadcast properly?
It might be worth trying a different virtual NIC type.

SELinux could conceivably be doing something too. Perhaps try
disabling that and see if it helps.

If you start corosync on another node will the two nodes see each other?

Here's one of my clusters with a single-node:

============
Last updated: Thu Feb  4 13:24:18 2010
Stack: openais
Current DC: pcmk-1 - partition WITHOUT quorum
Version: 1.1.0-4cdb459f017f10531226ba7b5bf56cbc774fb1f7
4 Nodes configured, 4 expected votes
10 Resources configured.
============

Online: [ pcmk-1 ]
OFFLINE: [ pcmk-2 pcmk-3 pcmk-4 ]

 Clone Set: Fencing
     Started: [ pcmk-1 ]
     Stopped: [ FencingChild:1 FencingChild:2 FencingChild:3 ]




More information about the Pacemaker mailing list