Blog
By Pete Zerger on 3/24/2010 9:25:38 AM • Rank (2373) • Views 2434
0

0

Arrows If you’ll please bear with me, I want to share what I think is useful and important information…but it’s going to take a few paragraphs to do it. If you find this information useful, great – maybe you can buy me a pint at MMS. If not, maybe I will be the one buying a pint. ;)  Anyway, let’s get down to business.

From the beginning, the Database Hygiene series has been about how to tackle what can be a difficult problem. Trying to identify what type of data is filling our Operational database AND which workflow(s) are at the source….to find what the product team gurus call the “top generators”. Those heinous workflows that get out of control when we’re not looking and turn our healthy environment into one with sluggish console performance…and into a “steaming pile” if we’re not careful. Well, that’s what my grandmother would have called it anyway ;)

A few weeks ago I had thought I might publish some of the reports I’ve used in my own environments, and I still might. But first, I want to show you how to utilize what may be the single most useful operational reports released for Operations Manager in the past year – the new “Data Volume by MP” and “Data Volume by Workflow” reports. These reports came in the latest update to the Operations Manager management pack, and they greatly simplify the implementation of a process to more closely monitor data volumes and data sources in your Operations Manager environment alluded to previous installments.

Didn’t read the first five installments? Doesn’t matter really, but here they are anyway.

The First Five Installments

 

The Process

So let’s start by spelling out our end goal and put in context with a sample scenario. Our goal is to answer the following questions:

1) What kind of data is filling my database?

2) Is this normal behavior or an anomaly?

3) In either case, which workflows are responsible?

Sample Scenario

Let’s begin with the hypothetical situation in which my Operational database is full, or nearly full. maintenance is failing, the Operations console is slow, operators are complaining. Okay, operators are complaining more than normal. We need to identify where the database is coming from and take corrective action.

Let us visualize the process. I think you’ll find it’s pretty self explanatory, but realize this. Step 0, the “Review data volume reports” task, is recurring. You need to be doing this on a recurring basis to understand over time what “normal” is. Schedule reports to be delivered to your mailbox so you’re more likely to take time to look at them.

image

 

Reports

Here is a look the OpsMgr Data Volume reports and how they answer the questions in the process flow. The latest Operations Manager MP contains these reports. We want to start at the top and drill down, so begin by launching the “Data Volume by Management Pack” report.

image

image

By default, all data types and all management packs are selected, which is what we want.

image 

figure 1. smart parameter header for data volume by management pack report

At the top of the report, you’ll see the top generators of data in your database – not only the top MPs, but the data types. this answers question 1 - What kind of data is filling my database?

image

figure 2. data volume by management pack and data type

Here we can see also a visualization of this data in charts, which actually have a specific and useful purpose. It answers question 2 - Is this normal behavior or an anomaly?

image

figure 3. charts in the data volume by management pack report

 

When you click on any of the numbers in figure 1, it will take you into more detailed information to answer question 3 - which workflows are responsible? In this case, I click on the top performance data generating MP (Perf Count column).

image 

The click through report looks like this. Here you can see which workflows are generating the data so I know which workflows require some fine tuning.

image

If you repeat this process for the top generators, you should be on your way to striking a balance of data collection and performance in your Operations Manager 2007 deployment.

Conclusion

I hope you’ve found this series informative and would love to hear your thoughts. Feedback on my ideas and information on how you tackle these challenges would be great. What other operational reports would you like to see. I have 8 or 10 operational reports to draw from if there is a need. There may be a part 7…questions and feedback will determine the need for an addendum.

Comments (3) - Comment RSS
KMV wrote: on Mar 25, 2010 09:18 AM
Hi Pete. I tried to run the report you have indicated but it only display data about "Not Sealed" Management Packs.

Thanks for this interesting article,.. I´m waiting Part 7... :-) Best Regards
Pete Zerger wrote: on Mar 25, 2010 11:09 AM
Any thoughts on other information you'd like to see in a report?
KMV wrote: on Mar 25, 2010 12:43 PM
Hi again... I resolved the issue. It´s a strange behaviour of the scom console. We have consoles in spanish and english. Well, if you run the report from the spanish console, only display MPs not sealed. Otherwise if you run the report from the english console you get all MPs, sealed and not sealed.

About your question, let me think and if i have some ideas i´ll send you. Thank you again.



Who Viewed
Who Reviewed
Categories
Related Pages
Shortened URL
http://tinyurl.com/yymx33p

Top Contributors
Featured Members
Pete Zerger
Points: 65622
Level: System Center Expert
Tommy Gunn
Points: 42748
Level: System Center Expert
Simon Skinner
Points: 40804
Level: System Center Expert
Stefan Koell
Points: 28999
Level: System Center Expert
Andreas Zuckerhut
Points: 27734
Level: System Center Expert