Thank you for clarifying the use of WSGIProcessGroup and
WSGIDaemonProcess.
Before reading your answer, I think I really misunderstood how they
relate to each other.
It helped me with fixing our server config.

Here is a bit of context why our server config is a bit different from
your recommendation (one processgroup per website):

Our config is now: one process group, n virtual hosts using this
single process group, using the same python code called with the
specific environment for the virtual host.
This makes our current server config a bit fragile, since one global
python variable can mess up our websites.
On the other hand, if we use separate processes per website we
actually need to estimate correctly, how many processes, each website
will need.
Even worse, if we want, let's say, 25 websites on a single server, we
would need 25 WSGI processes.
If our server can only take 10 WSGI processes (because of memory
requirements) it wouldn't be possible.
Our current setup allows any number of websites as long as the total
traffic can be handled by the 10 WSGI processes, with the caveat that
we can't have any globals by accident...

Jan-Eric


-- 
You received this message because you are subscribed to the Google Groups 
"modwsgi" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/modwsgi?hl=en.

Reply via email to