I want to analyse if certain type of requests do have an important impact on the server.
My backend service is behind nginx which logs all requests including the time needed to get the response.
Now the problem is how do I extract useful information from these logs which are ~150mg.
I am looking for something what would allow me to build a graph and correlate the requests with the backend CPU load.
The response time of a single request is not relevant because when then backend server is under high load even simple requests that would take 1 second, would return in 10-20 seconds.