Hello!

On Fri, Dec 31, 2021 at 02:24:13PM +0000, 幼麟 封 wrote:

> You are right, we can’t completely avoid this kind of problem,
> but we can reduce some. Maybe "keepalive_graceful_close" is not
> exact, it should be "wait_keepalive_on_exit". I mean, with this
> patch, I want to optimize the connection closure problem caused
> by the exit of the work process, is an optimization for a
> specific scene. By setting a shorter expiration time for the
> keepalived connection on the client side than the
> "keepalive_timeout" on the nginx side, we can really reduce the
> number of client retries.

Sure enough, by waiting longer before closing the keepalive
connection we can reduce retries. But this delays worker shutdown
and therefore comes at cost, and this cost might be bigger than
saved client retries. Further, this won't completely eliminate
retries anyway, even if clients are configured to use shorter
timeouts than nginx: there are cases when nginx closes keepalive
connections before keepalive_timeout expires, notably when there
are no free worker connections.

I've committed the patch to disable keepalive as long as we know
it cannot be used when sending response headers. This should
significantly reduce retries in practical cases, and comes at
(almost) no cost. Thanks for prodding this.

--
Maxim Dounin
http://mdounin.ru/
_______________________________________________
nginx-devel mailing list
nginx-devel@nginx.org
http://mailman.nginx.org/mailman/listinfo/nginx-devel