Warning: Part of this post describes navigation through Grid Control's UI. That means screenshots. Fair warning delivered.
Part of my recent Grid Control deployment project involved monitoring several E-Business Suite instances via the Applications Management Pack. On the test and development systems, the Workflow Notification Mailer is usually shut down, to prevent notification messages from those systems from confusing users -- or worse, customers. Of course, shutting down the Workflow Notification Mailer results in the service being listed as down in Grid Control. An additional side effect is that a number of dependent services will also be reported as down (or, amusingly enough, "partially up"), when, from the admin's perspective, they should be considered up. And then what happens? Well, people get notification emails, ugly red slices appear on the Grid Control dashboard's target pie-chart thingy, and the true status of some of the services is obscured. Minor annoyance, all around.
One solution would be to black out the Workflow Notification Mailer. Since the mailer is component of several tiers of sub-service definitions, however, this just sets the status of those sub-services to "Unknown/Status pending". These indeterminate states might be acceptable for Schrödinger's E-Business Suite instance, but the preferred outcome is to have these services show as "up."
A better solution, described below, is to remove the Workflow Notification Mailer from the list of components that represent availability of the affected Oracle Applications targets. This results in only the Mailer itself being listed as down, without impacting the status of other services. In a production environment, this monitoring configuration is not recommended, since people usually do care whether the Workflow Mailer is up. In the case of the development environment in this example, however, it's less of an issue.
Screenshot time, woohoo!
First, get to the main Grid Control target page for the Oracle Applications system in question, and find the Workflow Service, which should be a sub-service under Oracle Applications Infrastructure Service, itself a sub-service of Oracle Applications Service. Follow the Workflow Service link, as shown at the bottom of this screenshot:
On the Workflow Service page, follow the Monitoring Configuration link:
On the Monitoring Configuration tab, follow the Availability Definition link:
Under the System list, uncheck the Key Component checkbox for the Workflow Notification Mailer service:
Click the OK button to save the configuration change, and the Yes button on the subsequent confirmation page. Shortly thereafter, the status of the Workflow Service should change, as should the status of the services containing the Workflow Service.
But wait, there's more!
A down Workflow Notification Mailer also affects the status of the Workflow Infrastructure target, which in turn affects the overall availability status of the E-Business Suite system. The easiest remedy is to remove the Workflow Infrastructure target from the availability list for the entire instance, via the Monitoring configuration link at the bottom of the main EBS dashboard, under Related Links:
In the Edit Monitoring Configuration view, click the Next button twice to arrive at Step 3: Availability Definition. Find the Workflow Infrastructure entry in the table, uncheck the checkbox, and click the Next button twice more to reach the Summary page.
On the Summary page, click Submit to save the configuration changes. This should return you to the main E-Business Suite dashboard page, which should now read "Up" instead of "Partially Up," unless there are other, unexpectedly down components. If it ever becomes necessary to reintegrate the Workflow Notification Mailer into instance monitoring, it's easy to reverse the changes to the availability definitions.
I hope this is useful to anyone else out there who's had a similar need to alter the default Grid Control monitoring for an E-Business Suite instance. A final disclaimer, however, just to be clear: I've performed these steps on the monitoring configuration for test and development systems. If you really want to do this in production, I advise consulting documentation or Oracle Support beforehand. "I saw some dude on the Internet do this once" is rarely adequate justification for a change to a production service.