Hey folks! I found this slack through the wiremock site. I'm running perf tests at 300 TPS against my app. My wiremock instance is set to return responses at a fixed delay of 10ms. It has 2000 container, and async threads, with all the other recommended perf flags on. My wiremock CPU is at 30% and memory is at 40%. However, I get delays in the hundreds of milliseconds instead of the programmed 10ms. Is this a bug within your system or something I'm doing wrong on my side? Wiremock code seems to use a simple

Thread.sleep

but for the delays to be this high...it almost makes our perf tests unusable. Any help here?

Hard to say what’s going on here without more context. If you’re doing very complex matching, have large stub sets, or are sending large response bodies (particularly gzipped ones) then all these can push latencies up. 2000 is too many threads. I’d suggest more like 200. BTW WireMock only uses Thread.sleep() to implement delays when async is disabled.