[MapProxy] MapProxy and Apache stops responding

[dirk.thalheim at bkg.bund.de]

Dear list,

once in a while we're experiencing some stability issues on our productive MapProxy server. The server stops responding and only a reboot helps to solve the problem. Currently we're using a script which checks periodically if the server is responding and if not performs a reboot.

We're running MapProxy as recommended through mod_wsgi in Apache. When the error occurs the last lines of the Apache error log shows something like this:
[Tue Jan 12 09:16:18.075153 2016] [mpm_prefork:error] [pid 1122] AH00161: server reached MaxRequestWorkers setting, consider raising the MaxRequestWorkers setting
[Tue Jan 12 09:18:48.765912 2016] [core:error] [pid 30268] [client ...] Script timed out before returning headers: mapproxy_server.py

The mapproxy log reveals no further information.

I already tried to increase the MaxWorkers setting without real success? This is our Apaches mpm_prefork.conf:
<IfModule mpm_prefork_module>
        StartServers            5
        MinSpareServers         5
        MaxSpareServers         10
        MaxRequestWorkers       600
        ServerLimit             600
        MaxConnectionsPerChild  0
</IfModule>

Has anyone experienced similar issues? Is it simply a scaling issue or are there other things to consider? Could the cache storage format be a bottleneck? For the service in question we're using a SQLite Cache which is located on a network share and accessed by three MapProxy-servers.

Kind regards,

Dirk

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/mapproxy/attachments/20160112/8133bbef/attachment.html>