[Linux-cluster] How to try forever to start a service?

[Lon Hohberger]

On Mon, 2008-03-24 at 02:15 +0200, Theophanis Kontogiannis wrote:
> Hello All,
> 
> I have RHEL5 and what I try to do is make rgmanager to try FOREVER,
> until it succeeds, to start a service that is based on script
> resource.
> 
> Is there any such option either through cluster.conf or through
> the /usr/share/cluster/script.sh or any other file?
> 
> What is the option to do that?

Currently, there isn't one - but you can do it within your script itself
fairly easily.

Rgmanager will "restart" forever (if you want it to; depending on what
version you have), but the initial start needs to succeed or it will
temporarily give up when all nodes are exhausted.

-- Lon