Hi,
I have configured a cluster script resource and in terms of scripting everything seems to be ok (see bellow).
The cluster has two nodes configured with following properties:
- Possible owners: node1, node2
- health check for 10s
- thorough health check 1m
- Persistence mode enabled and auto start also enabled
I can move nodes using move this Service or Application to another node: it send first node offline and starts the second one it turns green and online and doing the reverse again also works.
If I have an error specified on the isAlive or LooksAlive cluster goes to failed state.
Now the issue when it gets to failled state if I have set to restart on Respond to resource failure the resource will attempt to restart the current node and if I make a second failure on the period of restarts it moves to second node.
If I disable this option (by setting If resource fails, do not restart) once the resource enters the failled state it won't do nothing just stays in failled state!
I believe the correct behaviour should be to move to second node? Is this a bug or am I missing some config?
Some more info the last entry I can see in the cluster log is:
[Terminating to Failed] --> Failed
rcm::RcmGroup::UpdateStateIfChanged: (node, Pending --> Failed)
Thanks for you're time.
Regards,
Luis Mesquita