[Linux-cluster] rgmanager timeout of script.sh
jakub.suchy at enlogit.cz
Thu Aug 14 08:42:59 UTC 2008
i am trying to solve a problem with timeouting a status script.
We are using custom init script for our service, which is doing some
operations in status section. However, one of the options when this
service is down is, that it hangs. Then, a status script may hang too
because it is waiting for an operation to complete.
Is there a way to instruct rgmanager to automatically treat the service as
failed if statut script doesn't return OK in say 30 seconds?
I've found /usr/share/cluster/script.sh with:
<action name="status" interval="30s" timeout="0"/>
but i don't really want to mess with this as it's a part of cluster
scripts (means red hat support). anyway, i tried timeout="30" but it
didn't work (guess "30s" is correct?)
Also, I don't like doing something like:
status: nohup operation &;
test if completed ...
More information about the Linux-cluster