enqueue/dequeue takes many seconds sometimes

If my server is down and gets a new request, which it offloads to a queue, I see very large delays between enqueue time and dequeue.

How can I avoid this?
My requirements is to return 200 immediately and do offline processing which is time sensitive.

I'm afraid that calling an async function without `await` could be aborted due to my server going down.

No repository files or tests are named. Start by reproducing the reported enqueue/dequeue delay around a request that calls an async function without await, then trace the server-down and queue lifecycle described in the report. Done would require a verified explanation of the delay and a clearly scoped change or documentation update.

Written by the indexing model from the issue text.