Re: Upstream least_conn behavior irregularity

On Thu, Mar 21, 2013 at 12:03:59PM -0700, John Watson wrote:
> Well doesn't make sense when theres >4 concurrent requests
> At any given time there's around 12 active_connections, but sock-3 is
> still never being used

Can you see a difference with only one worker process?

Currently, different workers have distinct counters of active connections.
It should be unnoticed under a high load.

> On Thu, Mar 21, 2013 at 9:34 AM, John Watson <[1]john@disqus.com> wrote:
>
> Ohhhh... that makes complete sense now.
> Had 4 workers.
> Thanks!
>
> On Thu, Mar 21, 2013 at 4:47 AM, Ruslan Ermilov <[2]ru@nginx.com> wrote:
>
> On Thu, Mar 21, 2013 at 01:45:14AM -0700, John Watson wrote:
> >    Was investigating some issues today when we noticed that
> least_conn wasn't
> >    behaving as expected.
> >    upstream backend {
> >      least_conn;
> >      server unix:/tmp/sock-1.sock;
> >      server unix:/tmp/sock-2.sock;
> >      server unix:/tmp/sock-3.sock;
> >    }
> >    The expected behavior for 4 simultaneous requests it should
> distribute
> >    them:
> >    sock-1: 2
> >    sock-2: 1
> >    sock-3: 1
> >    However, what we're seeing is:
> >    sock-1: 3
> >    sock-2: 1
> >    sock-3: 0
> >    Which coincidentally lines up with the number of requests a
> socket can
> >    service simultaneously.
> >    This is using 1.2.7
>
> And the number of configured worker processes is?

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