[MapProxy] tms request via gunicorn

[Travis Kirstine]

Are you running gunicorn on a specific port and proxy request via nginx?

docs
https://mapproxy.org/docs/nightly/deployment.html


On 7 September 2016 at 11:57, Andrew Terry <andrew.terry at centremaps.co.uk>
wrote:

> Hi,
>
>
>
> I’ve been following the mapproxy, gunicorn, eventlet, nginx demo and have
> a sticking point.
>
>
>
> Direct calls to mapproxy tms demo work fine but when I do this via
> nginx/gunicorn I can get everything to work (e.g. WMS requests) but the tms
> is slow and regularly fails workers.
>
>
>
> At the moment, I’m using proxy through nginx with an additional:
>
>
>
>     location /mapproxy/demo/static {
>
>         try_files $uri $uri/;
>
>     }
>
>
>
> Which allows me to load the demo site including html and css. I presume
> this is necessary to combat gunicorn static file issues but perhaps this is
> my first error.
>
>
>
> From the gunicorn logging I see:
>
>
>
> [2016-09-07 15:20:42 +0000] [2325] [DEBUG] GET
> /mapproxy/tms/1.0.0/test_wms_cache/EPSG27700/0/1/1.jpeg
>
> [2016-09-07 15:21:11 +0000] [2319] [CRITICAL] WORKER TIMEOUT (pid:2325)
>
>
>
> Whereas just mapproxy dev running direct works fine for:
>
>
>
> [info] XXX.XXX.XXX.XXX - - [07/Sep/2016 14:52:19] "GET
> /tms/1.0.0/test_wms_cache/EPSG27700/0/1/1.jpeg HTTP/1.1" 200 –
>
>
>
> Any pointers or prompts to where I can read more gratefully appreciated.
>
>
>
> Andy
>
> _______________________________________________
> MapProxy mailing list
> MapProxy at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/mapproxy
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/mapproxy/attachments/20160907/e308fe6b/attachment.html>