> Thank you for the quick response!
>
>  > Each app will live in its interpreter in the process, and when uWSGI
>  > receive the request it choose the interpreter to set.
>  > So if you set 10 processes for this instance and you host 5
>  > virtualhosts,
>  > you will ends with 50 interpreters (5 per process)
>  > by default uWSGI allocate all the resources you have configured,
>
> I'm not sure I understand this one because when I start uWSGI with
> virtualhosting, it does not know how many (and which) apps will be
> served so it cannot load the interpreters. Do you mean that as soon as
> one request is received for an app, uWSGI creates one interpreter in
> each process of the pool?


Yes, exactly

>
> I also have another question regarding the reload strategy with
> virtualhosting. Is it possible to only reload the interpreters for one
> application without having to reload all the processes in the pool
> (e.g., if the code changes in one app, but not in the others)?


no it is not possible, as even if we destroy an interpreter (there is a
function for this) this is a non-perfect clean. In the long run you will
leak tons of memory.



-- 
Roberto De Ioris
http://unbit.it
_______________________________________________
uWSGI mailing list
[email protected]
http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi

Reply via email to