I'm wondering how I can determine the amount of time that may have passed since a request was first received by the time my handler starts running.

It seems like when we get a huge spike of requests there is an increase in latency, but when we measure the latency of our handler it is consistently low. So I'm suspecting that these requests are sitting in a buffer somewhere, and if that's happening on my machine then I'd like to be able to detect that I'm encountering overload as much as possible.

1 post - 1 participant

Read full topic