Over the last couple of days I have been playing with the OpsMgr connector to get the alerts raised by OpsMgr as incidents in Service Manager. Now, while my ServiceManager environment is a POC environment, I did the unthinkable, and connected the environment to our production OpsMgr environment. I did this intentionally, as it is a fairly busy OpsMgr environment, and I need to understand what the impact of this will be on our eventual production SCSM environment.
In the first couple of days, I got well over 15000 incidents raised in ServiceManager, just from OpsMgr alerts. As this is a multisourced environment, with various different companies and teams responsible for each different technology/application within the environment, assignment rules need to be created to ensure the alerts are routed correctly. I realised that possibly the big-bang approach is not the right one, and I will need to approach this beast from a different angle. As part of our POC, we decided to focus on two very specific technology areas first for testing Service Manager’s functionality.
So, I removed the connector, and decided to take on the alerts one group at a time. I created a new connector, but this time, when it got to the configuration on the OpsMgr side, I created very specific, targeted subscriptions.
To do this, navigate to the Administration tab in the OpsMgr console, and locate the SCSM connector under the Internal Connectors section, and open the connector’s properties. You can from here create the big-bang subscription (see the Administrator’s guide), or you can get really specific. In this case, I created P2 and P3 subscriptions for AD and SQL, using the groups available in OpsMgr:
And specified the severity and priority I wanted to be sent through:
On the SCSM side, I used the standard Operations Manager Incident Template as a template, duplicated it and created one for each of the Incident priority levels that we will be using:
In these templates, I could then specify the assignment group (Assigned To) and a couple of other things specific to that technology. Back in the connector’s settings, I created routing rules for each condition relating back to the subscriptions in OpsMgr, and specified the corresponding template.
Lastly, I also created another, general OpsMgr incident template with the helpdesk as assignee for any alerts that may come through and not fit any of the rules created.
While being a bit time consuming to configure, it saves a lot of time in the long run, as you don’t have to sit and assign the incidents to the relevant teams (or employ an army of helpdesk people to do that all day long).