[Linux-HA] BUG or just oddity

[Andrew Beekhof]

On 11/10/06, Eddie C <edlinuxguru at gmail.com> wrote:
> Today I was writing an OCF script on a two node cluster I have.
>
> It started and then tried to run on node A. Which happned to be the DC
> The OCF script was not even on node a.
> Here is the funny part. I put the node A in standby it would not move to B
> I used a place to preference the resource to B
>
> Nothing.
>
> The only way I got it to move was installing the script on the other server
> and 'cleanup resource'


before we start a resource, we check all nodes to see if its running anywhere.

if the RA is not installed, then we cant do this properly and cant be
sure its ok to start it.

the reason it appeared to be "running" on node A is that we couldn't
tell it wasn't and to be safe, we assume it is.