Hello!

On Mon, Oct 10, 2011 at 03:09:07PM -0400, codetr wrote:

> nginx closes every night; I could not find the cause ..
>
> my nginx.conf
>
> user nobody;
> # no need for more workers in the proxy mode
> worker_processes 4;
>
> error_log /var/log/nginx/error.log;

Try using

error_log /var/log/nginx/error.log notice;

here, it will instruct nginx to log at "notice" level. This
includes system events like controlling signal. I.e. on getting
SIGINT nginx will log something like this:

2011/10/10 23:19:32 [notice] 24178#0: signal 2 (SIGINT) received, exiting
2011/10/10 23:19:32 [notice] 24178#0: exit

I suspect the cause is some broken log-rotation script which kills
nginx instead of instructing it to reopen logs.

(This is completely unrelated to "upstream sent too many data"
messages though.)

Maxim Dounin

_______________________________________________
nginx mailing list
nginx@nginx.org
http://mailman.nginx.org/mailman/listinfo/nginx