High request processing latency in apigeex

[raghunathapalle]

team, we are facing issue with high request processing latency in production. Not sure if there is anything in the policies that we could fine tune.

We see that "Proxy request flow started" step have high latency during load. Is there a way to drill down on this issue to figure out which component of the flow is having bottle neck?

[dchiesa1]

Hmmm, well if it takes a long time to start the request, then .... maybe the system is overloaded and has not responded to scale-up signals the way it should. I advise you to contact Apigee support immediately.  

If it is the API Proxy and the policies within it that is exhibiting slowness, then.... you'd want to look at your own configuration.  A couple  ways I can think of that would help: 

• Distributed tracing with sampling will give you some good data. 
• the Debugging session.... that's sort of a manual "sampling" you can do. 

It would be good to understand the P95, P99 latency on the proxies for which you are seeing anomalous latency.