Both PS and PXC dev Jenkins pipelines use instances labeled as docker-32g; however, PXC is significantly slower.

What I see in logs is:

For PS:
+ grep -c '^processor' /proc/cpuinfo

16

For PXC:

+ grep -c '^processor' /proc/cpuinfo

8

So they are not the same.

I can’t see another reason for this execution speed difference. I understand that with switching instances it can be that they will be interrupted more frequently, but pipeline implementation is ready for this. For now my main goal is to confirm that this is the reason.