Bottleneck difference Linux vs Windows Core

[B.F.]

So we had a few Linux proxies setup where the bottleneck would never be on the proxies. Since we are primarily a Windows shop, I thought I'd replace the Linux proxies with Windows 2022 Core Server's with the same VM configs. The overall performance seems to be about the same but I've noticed that the Load numbers have shifted. With the Linux proxies, the load would be very low at between 0% - 8%. However, now that the Windows Core's are in place, the numbers can range between 71%-94% and the primary bottle neck is the proxy.

Is this expected?

Thanks

[veremin]

If the performance stays the same, I would not be that concerned with what exact component is identified as bottleneck - remember there always will be a "bottleneck" in every pipeline or job processing. Thanks!