[Pacemaker] "No respawn" for cluster resources

[Stallmann, Andreas]

Hi there!

I just asked a few hours ago:


Ø  Is there a way to stop crm/pacemaker from doing that (an automatic respawn of a resource)?

Thing is: This “automatic respawn” does not happen for our mysql m/s resource (using the MySQL resource Script from Yves / Percona). We tried several things, but anything we did was futile, yet.

Still, if tomcat is killed, crm tries to restart tomcat on the same node. We don’t want that.

Perhaps you might want to have a look at our configuration and at the script itself. It’s an extension to the regular tomcat resource agent found in the distributions.

We modified it in a way, that it is able to control (promote / demote) a java web application. This is done through “wget” to a certain URL. The monitoring is done by a script that simulates direct access to the API of our application (javaapp-ocf-monitor.sh).

If you have any ideas or questions (sometimes questions are equally helpful), fire away and let us know. I’ll try to answer any questions arising, as far as they concern our configuration and the resource agent (but I’m useless as soon as the application itself is concernend).

Because of the length of script and config, I pasted them to pastebin. You’ll find them here:

Agent: http://pastebin.com/DgpjbRqM
Config: http://pastebin.com/CXD1BDsq

Cheers, thanks in advance and have a nice weekend,

Andreas
PS: If the tomcat m/s-agent will ever be finished, would it be relevant enough to be contributed to the community, what do you think? If yes, let me know how.



----------------------------
CONET Solutions GmbH, Theodor-Heuss-Allee 19, 53773 Hennef.
Registergericht/Registration Court: Amtsgericht Siegburg (HRB Nr. 9136)
Geschäftsführer/Managing Director: Anke Höfer

 ----------------------------


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20120727/e9a43fad/attachment.htm>