The heartbeat scenario goes as follows.
In a simple v1 config setup where server A has resources and server B dies.
Server A re-runs all the resources allocated to it.
When openvpn is re-run the exit code is non-zero and heartbeat thinks its resources have failed.
When this non-zero exit code is detected the Server A thinks its resources have failed and then passes them over to Server B.
Since server B is dead all the resources are now turned off and I'm up at 2am fixing it.
You can then understand why I feel so strongly about this seemingly minor issue.
Thanks for the info. I'll follow that up.
The heartbeat scenario goes as follows.
In a simple v1 config setup where server A has resources and server B dies.
Server A re-runs all the resources allocated to it.
When openvpn is re-run the exit code is non-zero and heartbeat thinks its resources have failed.
When this non-zero exit code is detected the Server A thinks its resources have failed and then passes them over to Server B.
Since server B is dead all the resources are now turned off and I'm up at 2am fixing it.
You can then understand why I feel so strongly about this seemingly minor issue.