Nov 062014
 

PASS Summit 2014 Logo_930x260Today I presented my SQL Server virtual machine “right-sizing” session at the SQL PASS 2014 Summit to a fantastic audience. I really appreciate each and every one of you for attending. This session was also broadcast on the Summit PASStv live video streaming session, and thanks to all of you who tuned in to watch!

As promised, I have attached the schema and scripts from the session, available to download here. These scripts will help you take Perfmon data that you have loaded into the database (use the Midnight DBAs code available here to do so, demo how-to video here) and run a statistical analysis on the data. The end goal is to help you understand the resource consumption of your SQL Servers, both physical or virtual, and take this data to determine the appropriate amount of CPU and memory needed to keep this SQL Server performing well while minimizing the size of the VM.

Also announced in this session is the target date of the end of the year for the beta release of a FREE standalone utility to run on your SQL Servers that will profile the performance trends of the server and present an estimation of the “right” amount of CPU and memory needed to balance the VM’s performance with the VM’s size. If you’d like to be added to the beta tester list, just contact me here and request to be added. Keep an eye on this blog for more details as the project nears a release date!

  4 Responses to “VM Right-Sizing PASS 2014 Session Wrap Up”

  1. […] VM Right-Sizing PASS 2014 Session Wrap Up by David Klee […]

  2. Hi David,

    Sign me up for the beta “right-sizing vm tool”
    We are about to virtualize multiple SQL servers… biggest being 2x 12cores, 200GB with about 5-6k trans/sec

  3. I’ve got you on the list! We have had some delays this year with development of the tool but are actively putting the finishing touches on it.

  4. […] last sample interval, which is usually 20 seconds. Therefore, SQL Servers that have an appropriate (or too high, but that’s a different story) amount of memory assigned to the VM will have most of the commonly accessed data in memory, but […]