Thursday, December 3, 2009

Exchange 2010, OpsMgr and other System Center Products

Yesterday I attended a Influencer Live Meeting about System Center and Exchange 2010. Very interesting it was. During this meeting much was told and demonstrated about how the System Center Products like OpsMgr and DPM 2010 can aid in making an Exchange 2010 more robust and disaster proof.


OpsMgr MP for Exchange 2010
Jon LeCroy, Program Manager of the Exchange Team told the audience about this part of the Meeting.

Like the R2 MP for Exchange 2007 has already shown, the newly released MPs have become way much better and far less noisy. This new approach is to be found in the Exchange 2010 MP as well. For noise reduction this MP is shipped with a Windows Service called the Correlation Engine. What it does? This engine uses dependencies encoded in the Health Model in order to determine the most likely root cause Alert when a problem occurs.

Duh! But what does that mean exactly? Well, take a look at this picture:

Suppose you have an problem with the AD Driver. This has its effect on many Exchange services like EWS, EAS, IMAP, POP and OWA. In the old days this would fill up the OpsMgr Console with a whole lot of Alerts. But now only the Alert for the lowest level failure will be raised and that within a 90 second window! That is a huge improvement isn’t it?

Also the Alert has been reclassified into 3 categories:

  1. Key Health Indicator
    Important service impacting issue (most alerts).

  2. Non Service Impacting
    Issues that affect specific mailboxes but do not impact the larger system.

  3. Forensic
    Monitors diagnostics that may not represent a specific problem. (Forensic monitors do not alert.)

Also the Reporting feature has been changed significantly. Some high lights:

  • The reporting about Mail flow statistics is based on message tracking logs.
  • Uptime measures application/feature (Service oriented reporting), not the server itself.
  • Availability reporting is Exchange aware.

DPM 2010
Jason Buffington
, Senior Tech. Product Manager took care of this part of the Meeting.

Very interesting it was as well since some new high lights of DPM 2010 were being shared with the audience as well. For instance, DPM 2010 will be capable of backing up servers residing in workgroups and/or DMZs! In the beta version this option isn’t available. But in the RC version it is very much likely to be present! Nice!

The most important message of this part of the Meeting was about the way DPM 2010 integrates tightly with Exchange (2010) and the way it backups the data. Only the native VSS Writers for Exchange is being used here, so the data is correctly backed up in a way that Exchange fully supports, so there is always a way back. And not just that, but one can go back 15 minutes up to 30 days (for instance) with only a few mouse clicks. Again, DPM 2010 takes care of the rest!

Also the fully transparent interface and wizards do take away many questions people do have to answer when using other backup products, like creating backup schema’s with full backups and the lot. Also where the data resides is taken care of. So no more hassle. Just implement a valid backup schedule/schema and be done with it. Also the way DPM 2010 checks its own health is great. Really Set & Forget! Wish I had a backup tool like that in my days being a Systems Engineer. Would have saved me many many days! :) And the company I worked for many euro’s!

Also good solid advise was given about virtualization and Exchange 2010:

A very interesting Live Meeting it was. An hour very well spent. The only ‘bad’ thing was that it was only one hour :). Time really went fast and both speakers had way much more to tell. Future sessions I will certainly attend.

No comments: