Eventscripts: restart lockd in the background when going unhealthy
authorMartin Schwenke <martin@meltin.net>
Wed, 16 May 2012 03:29:58 +0000 (13:29 +1000)
committerMartin Schwenke <martin@meltin.net>
Wed, 16 May 2012 07:19:55 +0000 (17:19 +1000)
commitffbe59bd44fee532e0167b56f4f031dc3d673380
tree96751dc06832deb4061a54039dfd0b1aebfde9fb
parent92eb004162ac1729a3ca5210a0c97257e86cd0c6
Eventscripts: restart lockd in the background when going unhealthy

Sometimes the restart can hang when there are I/O problems.  Then the
eventscript times out and gets killed so the node never marked as
unhealthy.

Restarting in the background avoids this.

Signed-off-by: Martin Schwenke <martin@meltin.net>
(This used to be ctdb commit 13acd58c41fba1a33894fbd654fed69ea0eac322)
ctdb/config/events.d/60.nfs