Hi,
I'm working on a server alert plugin. For the alert notification I want to:
- execute an operation
- wait a couple of minutes
- collect a snapshot
Originally, I was just collecting the snapshot and that worked fine.
At the moment I am having trouble executing the operation. I am using
the same kind of code I use in our Java remote client to schedule an
operation and then wait for its completion, but the operation is
consistently timing out. Furthermore, the RHQ server just keeps trying
to re-execute the notification logic after each time out, but that's not
my first concern.
In the alert sender send() method I just call
LookupUtil.getResourceManager().scheduleResourceOperation(...) and then
call
LookupUtil.getResourceManager().findResourceOperationHistoriesByCriteria(...)
as we do in our client. Checking my agent log file I see the operation
being performed and apparently completing normally in a few seconds.
Waiting for completion for 4 minutes does not return any histories at
all in the alert. The schedule and the wait are done using overlord as
the Subject.
Is there any special context I need when scheduling operations from a
server alert plugin? Any other ideas?
Thanks in advance,
Steve