One of our databases had almost 20 million rows in localized text. Thanks for the post
Derek
16 Oct 2008 8:33 PM
xacolabril
31 Oct 2008 5:50 PM
Oh! Our OperationsManager database hd 18 milions of rows in LocalizedText and our performance is very poor.
Thank you for the post too.
Zied
28 Apr 2009 8:48 PM
we have 6 millions of rows in LocalizedText, the performance, alerts and events are poor...
zied
28 Apr 2009 8:48 PM
Thank you very very much for the informations, and for this wonderful Blog..
8 May 2009 8:35 PM
Kevin, great post. We really had a big table. Unfortunately the first time we ran into diskspace issues. That's how I figured out the errorprocessing part does not delete the first temporary table (#PublisherMessageReverseIndex).
Anonymous
23 May 2009 10:01 PM
I upgraded my test lab from SP1 to R2-RTM this weekend.   My current test lab consists of the following
Anonymous
28 May 2009 2:27 PM
I had troubles with the Opmsgr Console performance at serveral clients. One of the most important things
Anonymous
10 Jun 2009 6:24 PM
If you’ve been reading this blog for any length of time then you’re already familiar with the postings
Daniel Elliott
22 Jun 2009 11:46 AM
Thank you Kevin,
Our environment is large with 30 Exchange 2007 servers and 180,000 mailboxes.
Finding this post has save me many hours of frustration, our DB was 70gb with 50 million entries in the localizedtext table. After running your scripts our DB is now 4gb and operating with far less noise.
It would be good if this was resolved as I don’t really want to rebuild our SCOM environment with R2. Would also be nice if Microsoft release the native Exchange 2007 MP, two years after exchange 07 and we are still monitoring with a converted 03 exchange MP (poor form msoft).
Once again, thanks Kevin!!
22 Jun 2009 8:02 PM
1. This issue is resolved in R2, for the most part.
2. By moving to the native Exchange MP for R2, you also remove this issue. Microsoft HAS shipped the native Exchange MP. However - due to it needing some native powershell modules that were added in R2, it is R2-only MP.
3. R2 is not a "rebuild" it is an upgrade, and installs just as most hotfixes do.
Pascal Slijkerman
23 Jun 2009 4:50 PM
Thanks for the script!
Whe monitor about 130 agents with more then 25 MP's. Whe had over 20.000.000 rows of localizedtext.
The script took 27,5 hours!!! On a HP Proliant DL380 G5 with P500 raid controller and 512MB BBC.
Our database was 26GB and has dropped to a 2.4GB
Ít's now nice and fast :-)
23 Jun 2009 6:13 PM
So, typically, I expect 20 minutes per million rows.
I would have estimated your cleanup to be around 7 hours, 14 MAX.
This makes we wonder if you really have good disk I/O for the SCOM subsystem? You are using a DL380 - with only 130 agents, how are the disks set up on this server? Are you using RAID5? Do you have the RMS and DB on the same server?
Gerardo
25 Jun 2009 1:41 AM
Hi, may be you can help me. The result of the script was that WorkflowTimestamps table has 9 million of rows so i dont know how to reduce the space.FYI the grooming process finish ok. so what can i do?
25 Jun 2009 1:49 AM
I have no idea what "WorkFlowTimestamps" table is, or how it relates to this issue. Can you give more information on how you deduced that table name? Was that in an error output or something?