Deployment options

I'm wondering about the pros and cons of the different fastcgi
configurations.

Could someone please provide a bit of guidance on deciding which mode
to run an application under?

I need to deploy an application that will handle all requests for a
site.  It's a Perl-based application (Catalyst) which can be run with
its own process manager (FCGI::ProcManager) in "Server" mode.

I see advantages of running in Sever mode in that (IIRC) the
application can run under a different UID than the web server.  A
disadvantage is that there has to be a separate way to make sure the
application continues to run (some kind of watchdog script, perhaps).
Another disadvantage, if I understand correctly, is that the perl
process manager will not spawn new processes based on load.

Another option is to run in "Static" mode.  That uses a process
manager in mod_fastcgi, correct?  This does load more processes based
on demand, correct?  But the application runs as the uid of the
Apache, unless the FastCGIWrapper is used.

The host where the application will be deployed already has fastcgi
setup for Dynamic mode applications (via .fcgi ).

I'm not clear about running as dynamic application.  My concern is
startup time of the application.  Perl isn't known for its fast
startup time.  Can dynamic mode induce a bit of latency into the
site more than the other modes?


Thanks,


-- 
Bill Moseley
[email protected]

___________________________________
fastcgi-developers mailing list
http://fastcgi.com/fastcgi-developers/