Good Morning, We have a suite of scripts (11) doing end to end customer journeys. Total TPM generated is approx 220K
When we run these tests on Windows EC2 instances we have noticed CPU spikes every two minutes. Normal CPU about 10% & Spikes around 50%. I’ve tried running with different Executors & changing the thinktimes but it makes no difference. I’ve also disabled some of the more complex scripts but still see spiking, but on a smaller scale
The tests run fine but we find we are running larger instances to cope with the spikes, be interesting to know if anyone else has observed this and likely causes ?
Thanks
Hi @daveknowles,
Welcome to our community, thanks for sharing your question here 
I’m curious, and willing to help you.
So, could you please share the minimum example you’ve found so far that’s able to reproduce that behavior? Also, I’d love to see the shape of how those CPU spikes look like (e.g. in a chart?).
Additionally, you may want to replace private services or sensitive urls with any of the publicly available example sites, like https://httpbin.test.k6.io/ or the WebSocket equivalent you can find here.
Btw, what kind of tests are you running? Protocol (HTTP, etc) or browser-based? If protocol, what modules do you use? http? grpc? ws?
Thanks! 
