Does sys.dm_exec_procedure_stats.total_worker_time include signal waits or only time spent by procedure on scheduler?

The reason of question - when number of cores was increased on the server signal waits decreased as expected, but total_worker_time/execution_count has dropped too.


Neal Hambly in his post at SQL Server SOS_SCHEDULER_YIELD Wait Type includes this comment (highlights are mine):

If you have long running queries, especially when in conjunction with high signal wait times, it may be an indication of CPU pressure.

Reviewing the system performance will show if the CPUs are showing signs of CPU pressure, and reducing workload through performance tuning or other steps can help reduce the workloads being performed by SQL Server.

Certainly adding more cores will change the amount of pressure each CPU must face. So, it seems quite reasonable that more processors will reduce the pressure on each CPU, which in turn provides better CPU throughput.

  • Not answering my question. It was not a question on throughput, the question was - why average cpu consumption of procedures has decreased? – Alexander Tarasul Apr 12 '16 at 17:16
  • @AlexanderTarasul - Last paragraph: I italicized the note on reduction of CPU pressure. – RLF May 10 '16 at 18:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.