How do I monitor request to be served or scheduled across uvicorn workers?

54 views Asked by At

You can assume, there are 4 uvicorn workers running on 4 vCPU. I am getting higher response time with more concurrent request. To debug that issue, I believe monitoring pending requests to be served by uvicorn would help. How can I monitor request in queue to be served in total by those workers ? Just for your information, I just receive JSON file around 500 KB through API call and immediately response something dummy.

0

There are 0 answers