gunicorn中max_requests和max_requests_jitter之间的机制是什么

Question

According to the official guide

https://docs.gunicorn.org/en/latest/settings.html#settings

a worker will restart when it has handled max_requests requests.

But when max_requests_jitter is set, a worker will restart when it has handled randint(0, max_requests_jitter) request, to stagger worker restarts to avoid all workers restarting at the same time.

Is that means max_requests_jitter setting will override max_requests and make it be invalid？

Answer 1

From the docs -

The jitter causes the restart per worker to be randomized by randint(0, max_requests_jitter) . This is intended to stagger worker restarts to avoid all workers restarting at the same time.

What I understand is that the jitter is a random addition to each worker and the term max_requests_jitter should be (though not necessary) smaller than max_requests .

In other words, worker_1 will restart after max_requests + j1 requests, worker_2 will restart after max_requests + j2 requests etc. where the values of j1, j2, j3... are determined by the max_requests_jitter argument.