I have a dedicated SQL 2008R2 box for our Solarwinds deployment. It is a very robust server with twin 4 core Xeon processors, 32GB ram and solid state hard drives. I decided to drop the new APPInsight template from SAM 6.0RC3 on the solarwinds box to show our SQL guys the type of info that was available to them on a very active database. As soon as I dropped the template on the box, the application went "red" and showed a number of critical performance errors. when I opened the solarwinds database it showed a lot of indexes were over 85% fragmented, and a high number of full page reads/writes. So, I logged onto the solarwinds server and launched the Database Maintenance routine. That routine shows that it is de-fragmenting the indexes. THe log file shows a "successful" de-fragmentation routine has been run, even when only marginal gains in index fragmentation are realized (i.e. 14.62 to 13.35%).
However, when I re-opened the AppInsight page on the Solarwinds web console, I am still showing a lot of the same indexes fragmented at over 85% all the way up to 97%.
My question is, how do I remedy this? Do I need to remedy this? is there a routine that I can run on the SQL box itself (instead of through Solarwinds tools) to de-frag the tables?
Also, I tried to clear out old data by changing the database settings to discard old data after 180 days, down from a year. I saw 0 decrease in the size of the database after running maintenance. Is that normal?
