Jetstress Field Guide – April 2011 Release


When I released the first version of this guide back in November 2010 I had no idea how many people were out there having problems with Jetstress testing. Over the last 5 months I have had the opportunity to work through and discuss some interesting cases with a number of you.

So I thought it was about time that I updated the guide with the lessons learned from those cases and to include some stuff that I had forgotten to include in the first release. I have also included some feedback from the internal Microsoft Jetstress community. Hopefully this has produced some more useful information and clarified some sections.

So, here is version 1.0.0.11.

Change List

  • Updated for version 14.01.225.017 of Jetstress
  • Added Exchange Server 2003 instructions
  • Added more troubleshooting information
  • Added guidance for running Jetstress on a production Exchange Server
  • Included guidance for testing Raid Arrays
  • Included information and guidance on Background Database Maintenance
  • General corrections and clarity improvements

Lastly I want to thank everyone that contributed to this guide, from the Exchange product group, our partners and my Exchange colleagues out there in the real world, working at the sharp end every day – without the feedback from you all this guide would not be possible, so huge thanks to everyone and please keep the feedback coming!

You can download the updated whitepaper here.

Happy Jetstressing!

Neil Johnson
Senior Consultant, MCS UK

Comments (4)
  1. Jeremy Wood says:

    Very nice Neil!  This is definitely a tool that will go into my toolbox.  :)

  2. Jack says:

    What is your thought if the server is virtualized with pass through disks on a hypervisor?

  3. Neil Johnson says:

    Virtualisation doesn't change the spirit of the test, we are still trying to validate that the storage is able to meet the predicted working load.  From a Jetstress perspective we dont really care if we are virtualised or not… however ensuring that the test is a valid representation is more challenging, especially if you are sharing your virtual host with other workloads.   My advice is to configure Jetstress as if it was a physical deployment, but you also need to ensure that any other shared workloads are also simulated during the test duration to ensure a valid test.  More on this topic to come in the  next version of the guide… its a hot topic :)

  4. Shag says:

    We are migrating to Exchagne 2010 from Groupwise.  They already built the Exchange environment, but want to expand.  I am not sure if they ever ran JetStress before, but want to do it now.  We are probably going to build two more mailservers.  Before we install Excchange, I thought it would be a good idea to run JetStress on the two new servers.  I never ran this tool before.  I had one question, when we run jetstress on the two new servers, will this give us the information for the whole Exchange environment.  What I mean is will I get informaiton on the existing servers as well.  We only have a handful of users who are testing on Exchange right now.  Thanks!

Comments are closed.