2 Replies Latest reply on Dec 14, 2015 10:42 AM by rottengeek

    Does removing waits to track alter the list of worst performing SQL?

    rottengeek

      At some point this week, I removed some of the extraneous wait types (I had removed memory/cpu and added everything but the obvious into the metrics).  Our biggest problem has been OLEDB and preemptive_com_getdata, which makes sense based on what the server is doing.  However, a few of the preemptive_com waittypes seemed to track (perecentage-wise) almost exactly with preemptive_com_getdata, and were just really cluttering things up.  So, I removed those from collection.

       

      I'm trying to remember the date/time I did that, because at 1:20 PM December 9th, the "top SQLs" suddenly changed - DRASTICALLY.  My question is, if I removed those wait types did i somehow change the metrics on calculating "top SQLs" ?