ctdb recovery lock file created with wrong permissions?

[Martin Schwenke]

On Tue, 21 Jun 2016 02:35:59 -0500, Steve French <smfrench at gmail.com>
wrote:

> I did an experiment with ctdb using a local recovery lock file ie
> /etc/sysconfig/ctdb had:
> 
> CTDB_RECOVERY_LOCK=/var/lock/.ctdbrecoverylock
> 
> and noticed that on this node the cluster it started unhealthy even
> though it should be the master.  When I looked at the mode of
> /var/lock/.ctdbrecoverylock its permissions were 0600 so I changed
> them to 0777 and ctdb went back to healthy (and the master).  Seemed
> strange.   Do I need to do something special to make sure the mode of
> the file is set?

Nodes need to pass a monitor cycle before they become healthy.  Could
your update of the mode just have happened about the time the monitor
event happened?

peace & happiness,
martin