Image by Hil via Flickr
if you’re using the native exchange 2007 mp with clustered mailbox servers, you may have noticed that your synthetic transaction executions are timing out.
you should be able to recognize the alerts. they look something like this:
Some of the MAPI connectivity transactions failed. Detailed information:
Target: System mailbox for XYZ
Error: The transaction did not complete in the alotted time (20 seconds).
here is an explanation from microsoft:
“In the core OpsMgr code, there is a “cluster override” that disables all workflows for objects that are contained by a cluster virtual server, unless the workflows are running on the Active Node. When we create the relationship between the Synthetic Transaction object and the Cluster Virtual Node for the Mailbox Server, the CAS Server is now subject to the "cluster override" even though it isn't a member of the cluster, because now the Synthetic Transaction hosted on the CAS server is contained by the Cluster virtual server.”
the workaround is to disable the discovery that builds this relationship. when you do this, be cognizant that the maintenance mode model changes. the discovery mentioned above that requires disabling is:
name: rms target relationship discovery
target: root management server
this is the net effect of disabling this discovery:
- Mailbox Server is put into Maintenance Mode and taken offline
- The CAS Synthetic Transactions that target that Mailbox Server will not be put in MM and will continue to run
- The transactions will fail and will go into a critical state
oh well! at least your synthetic transactions will run. by the way, since this is a bug, it’s being worked on. :)