> [alert] 11371#0: worker process 24870 exited on signal 9

This is almost certainly the cause of your problems - you need to figure
out why the nginx processes are crashing and resolve that. Most likely a
3rd party module is responsible.

On Fri, Feb 16, 2018 at 10:39 AM, Andrzej Walas <nginx-forum@forum.nginx.org
> wrote:

> For me 40-50 GB cache is ok, because I have multiple files like 2-5GB.
> Problem in my mind is this that I have settings:
> proxy_cache_path /ephemeral/nginx/cache levels=1:2
> keys_zone=proxy-cache:4000m max_size=40g inactive=1d;
> but I have over 40GB on disc and files older than 1 day inactive.
>
> Can you tell me what happend with downloaded part of files when I have:
> [error] 16082#0: *1264804 upstream prematurely closed connection while
> reading upstream
> [crit] 16082#0: *1264770 pwritev() <path> has written only 49152 of 151552
> while reading upstream
> This part of file is still on disc and don't deleted after error?
>
> On most of my proxy rate is 90% HIT to 9% MISS and 1% ERROR. But couple of
> them have stats like 10% HIT, 60% MISS, 30% ERROR.
>
> Sometimes I have problem with MISS on existing file. In logs I see 1 MISS,
> after that 10-20 HIT and after that only multiple MISS on this same file.
>
> Posted at Nginx Forum: https://forum.nginx.org/read.
> php?2,278589,278613#msg-278613
>
> _______________________________________________
> nginx mailing list
> nginx@nginx.org
> http://mailman.nginx.org/mailman/listinfo/nginx
>
_______________________________________________
nginx mailing list
nginx@nginx.org
http://mailman.nginx.org/mailman/listinfo/nginx