CPU Co-Stop

David Klee discusses an important hypervisor-level metric:

VMware’s CPU Co-Stop metric shows you the amount of time that a parallelized request spends trying to line up the vCPU schedulers for the simultaneous execution of a task on multiple vCPUs. It’s measured in milliseconds spent in the queue per vCPU per polling interval. Higher is bad. Very bad. The operating system is constantly reviewing the running processes, and checking their runtime states. It can detect that a CPU isn’t keeping up with the others, and might actually flag a CPU is actually BAD if it can’t keep up and the difference is too great.

This is extremely useful information for DBAs in virtualized environments.  My crude and overly simplistic answer is, don’t over-book vCPUs on hosts running important VMs like your SQL Server instances.

Related Posts

Get Windows Failover Cluster Errors

John Morehouse walks us through the Get-ClusterLog cmdlet in Powershell: Sometimes you know that a problem occurred, but the tools are not giving you the right information.  If you ever look at the Cluster Failover Manager for a Windows Cluster, sometimes that can happen.  The user interface won’t show you any errors, but you KNOW […]

Read More

SQL Undercover Inspector V1.3

Adrian Buckman announces a new version of the SQL Undercover team’s Inspector: We know some of you really hate linked servers so we have been working on a powershell collection which will allow you to install the inspector without using linked servers to centrally log the information and instead the powershell function Invoke-SQLUndercoverInspector will do […]

Read More

Categories

March 2016
MTWTFSS
« Feb Apr »
 123456
78910111213
14151617181920
21222324252627
28293031