0

We have a change moving into production which has increased response time and a increase in CPU utilization when tested in lower environment with a single POD.

Old response time= 1600ms, new response time=1800ms. Old CPU utilization 0.15 of 1 core POD vs new CPU utilization of 0.2 in 1 core POD. Old TPS 3.1 vs new TPS of 2.8.

How do I calculate or relate this CPU utilization increase in lower environment towards production increase which has 60 PODs at 28 cores peak utilization?

Rajesh Mappu
  • 483
  • 9
  • 23
  • Just plain 5% multiply it to whatever peak CPU usage seen in production when peak traffic happened? – Rajesh Mappu Oct 13 '22 at 17:24
  • 5 is the percentage points difference between old and new cpu utilization. – Rajesh Mappu Oct 13 '22 at 17:54
  • Differences in memory bandwidth and cache sizes might mean scaling is different as well. There's no way to be sure, or even to make a good estimate without knowing more about the different CPUs and the workload. – Peter Cordes Oct 15 '22 at 15:09

0 Answers0