Through the virtual looking glass at SQLBits 9

At SQLBits 9 in September I’ll be presenting a session about how to monitor a virtual SQL Server’s performance and why we need a different approach to traditional physical server monitoring.

We often hear that once a server is virtualised tools like Task Manager and Perfmon become unreliable and there’s often a lot of truth behind those statements. How can taskman report 50% CPU utilisation when it doesn’t know what value 100% really presents? The same can be said for memory utilisation, yet both are key resources for SQL Server. So, how as SQL Server professionals do we tell if SQL Server is suffering from performance problems which maybe resource related?

My session on the Saturday community day will answer those questions for you by not only discussing some of the concepts behind performance monitoring SQL Server once virtualised, but we’ll also use live demos to simulate and monitor resource contention in a virtual environment.

The session’s agenda will be:

  • Common virtualisation terminology and components
  • Why virtualisation changes the approach to monitoring
  • Detecting hypervisor memory reclamation
  • Sources of performance and system information and misinformation
  • Using Perfmon counter at the host and guest server level to monitor:
    Memory
    CPU
    and SQL Server performance
  • Relating SQL Server’s wait stats to the resource pressures seen in Perfmon

If this subject is relevant to you I hope to see you there, but if you have any questions about this in advance of SQLBits please feel free to email me.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s