Home > Operations Manager > Exchange 2010 Management Pack for SCOM Released

Exchange 2010 Management Pack for SCOM Released

Having been on the TAP program for Exchange 2010 I have had some experience with the beta version of this management pack.  MSFT released the Exchange 2010 management pack yesterday and you can get it here.

There are a few pre-requisites before installing this management pack.  Basically they are hotfixes that need to be applied to either your SCOM 2007 SP1 or R2 environments.  These are documented in the MP guide and information on how to obtain them are listed as well.  Be sure you get these and apply them to your environment before installing this MP.

MSFT also has a list of recommended additional management packs that should be installed, but not necessary for all environments (I.E. Labs, etc).

One thing that you will notice that is different for this management pack is that it will install a service on one of your management servers.  The RMS is recommended, but that decision is up to you.  This new service is called the Correlation Engine (reminds me of what HP does with it’s EVA systems using something called WEBES).

So when you install this MP you will see two directories for data to be extracted in.  After the install you simply import the MP (there are two mp files one for monitoring and one for reporting) and wait for discovery to be completed.

One thing to note, is that you must enable agent proxying on all Exchange 2010 agents in order for this MP to work (as well as have the service run as local system).  It might be a good idea also to verify that those agents did receive the hotfixes necessary before importing the MP as well as proxy enabled (both of these things can easily be checked by creating state views to display this data).

In the guide there is a warning that if you have the Cluster Management pack installed, you will be told how to exclude Exchange 2010 clusters from this management pack.  I looked through the guide and didn’t find those instructions.  I contacted the MSFT party responsible for the MP and he said that note should have been removed prior to it releasing, so if you see this and get concerned, then don’t worry it wasn’t meant to be in the guide.

As always, read the guide carefully before installing this MP and ensure you understand how alerts bubble up, and how components are discovered.  Overall the deployment was relatively simple and I am looking forward to seeing just how well this MP will manage our Exchange 2010 infrastructure.

Categories: Operations Manager
  1. Jonathan
    November 20, 2009 at 12:39 am

    "…ensure you understand how alerts bubble up…"There is no way to know how the alerts will bubble up, because the correlation engine determines that using some secret algorith. We\’ll see how this flies. That is, taking the control and flexibility away from the SCOM administrator and allowing the correlation engine to keep it a "hands-off" monitoring solution. Microsoft does toute in the guide that very little tuning is necessary with the advent of this great new correlation engine.

  2. Blake
    November 30, 2009 at 6:45 pm

    We shall see Jonathan..We shall see, but of course I agree with you they should stick to best practices.

  3. Jon
    December 4, 2009 at 4:39 pm

    I put it all in the MP guide – the correlation engine uses the dependency relationships tagged with the "virtual" property in the health model. These are the somewhat complex diagrams in the appendix. Unfortunately OpsMgr has some limitations in terms of making things both flexible and upgradable, and removing an MP is a lot more invasive than potentially not being able to tweak somthing.Cheers,Jon

  4. Jonathan
    December 7, 2009 at 6:21 pm

    "…the correlation engine uses the dependency relationships tagged with the "virtual" property in the health model…"I must have missed this in the guide. Which page is that on? Even still, just knowing which dependency relationships are tagged with the vitual property does not help to deduce when an alert should or should not be generated. There are rules around any type of correllation that need to be explained before one can understand how it will process.

  5. kamal
    January 8, 2010 at 12:36 am

    Nice article, For more details on Exchange 2010 MP, see the article as "The new Exchange 2010 Management Pack for OpsMgr" at http://ramratankhoj.blogspot.com/2009/01/new-exchange-2010-management-pack-for.html

  6. Murad
    March 3, 2010 at 7:11 am

    Blake I have two questions,First, in the "Exchange 2010 MP Guide on page 6 it talks about "Supported Configurations" and there is a two column table that lists the "Configuration and Support" information. Would you be able to explain what Microsoft means by "Virtual environment" on line 3 indicating as "Not Supported"? It\’s not clear to me if they are referring that if my Exchange 2010 is running in a virtual environment it\’s not supported by this correlation engine or if my RMS/SCOM 2007 R2 is running in virtual is not supported. Second, my RMS is clustered as I am pretty sure many others are running the configuration, though Microsoft talks about installing this correlation engine on RMS for better access to SDK service, no where in this guide talk about clustering this engine and if it can be done/supported. Would you be able to enlighten us regarding installing this engine on a clustered RMS (Win 2008)?As mentioned in this guide, I\’ve tried sending an email to exguide@microsoft.com, but it bounced back as "BAD ADDRESS", go figure :-)ThanksMurad

  7. Blake
    March 5, 2010 at 11:19 am

    Murad, We have Exchange 2010 running on physical hardware so I am not certain and my RMS is not clustered. I will contact the PM and get these answers for you.

  8. Blake
    March 8, 2010 at 6:07 pm

    Murad, I got a response from Jon LeCroy of MSFT in regards to your questions. Here are his answers:1. Virtualization support: Any platforms supported by both Operations Manager 2007/2007 R2 and Exchange 2010 are supported by the MP. The MP guide verbiage was a copy/paste issue from the templates we use to aid consistency with MP guides. Sorry for that.2. Clustering the Correlation Engine: I expect this is straightforward (installing it on each node and adding it to the cluster config) but we have yet to test it. It will be tested & documented for our SP1 release.3. Thanks for letting me know about the exguide address bounce – we have groups set up not to allow external mail and I neglected to fix that for this one. Should start working shortly.Thanks for the questions Murad, and of course, thanks Jon for the answers!

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: