I am running mod_wsgi (v4.6.7) + django in an OpenShift environment on a 
RHEL7 base image.

I have noticed a semaphore leak due to restarts of mod_wsgi because of 
container restarts. This is k8s so the good old pause container with shared 
IPC is in play for pods.

When my application containers restart they are allocating new semaphores 
each time mod_wsgi starts and eventually this results in IPC semaphore 
exhaustion.

This originally manifested because of a cluster platform issue which caused 
a high rate of container restarts in a short period of time.

This is my mod_wsgi invocation:

python manage.py runmodwsgi --https-port=$FOO_SERVICE_PORT 
--server-name=foo  --server-alias=*.example.com --server-alias=localhost 
--allow-localhost --https-only --log-to-terminal --log-level=info 
--ssl-certificate=/etc/tls-certs/tls --trust-proxy-header=X-Forwarded-For 
--trust-proxy-header=X-Forwarded-Host --trust-proxy-header=X-Forwarded-Port 
--trust-proxy-header=X-Forwarded-Server 
--trust-proxy-header=X-Forwarded-Proto

Is there anything glaring I could try to mitigate this? Perhaps with apache 
mutex settings?

Thanks,
John

-- 
You received this message because you are subscribed to the Google Groups 
"modwsgi" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/modwsgi/ac5a77ca-9ace-4f34-acf0-6a6444ab3928%40googlegroups.com.

Reply via email to