21 jan 2007 kl. 10.52 skrev bill lam:
> A similar question had be posted but I cannot locate an answer.
>
> What is the difference between FCGI_MAX_CONNS and FCGI_MAX_REQS, if
> 1) FCGI_MPXS_CONNS=0
> 2) FCGI_MPXS_CONNS=1
>
> I cannot understand the wording in the spec and will appreciate any
> insight.
Initial observation: FCGI_MAX_REQS is per application, not per
connection.
If it had been per connection, the FCGI_MPXS_CONNS value would have been
redundant; multiplexing would have been allowed if and only if
FCGI_MAX_REQS > 1.
If FCGI_MPXS_CONNS = 0, the number of simultaneous requests can't exceed
the number of simultaneous connections, so the number of simultaneous
requests is limited to the minimum of FCGI_MAX_CONNS and FCGI_MAX_REQS.
If FCGI_MPXS_CONNS = 1, the number of simultaneous requests is limited
only by FCGI_MAX_REQS; FCGI_MAX_CONNS doesn't matter.
Final observation: having FCGI_MAX_CONNS > FCGI_MAX_REQS is not as
useless
as it might seem. It allows the web server to pre-establish a
connection
without immediately using it for a request. This could (theoretically)
improve response times and throughput.
/Bo Lindbergh
___________________________________
fastcgi-developers mailing list
http://fastcgi.com/fastcgi-developers/
