Poul-Henning Kamp wrote: > In message <[email protected]>, Nick Loman writes: > >> Precisely, we only have perhaps 50 PHP children serving requests, so if >> these are kept open to serve idle keep-alive connections, that severely >> limits the numbers of dynamic page requests we can serve. > > The difference between letting them say open 10msec and 0msec is very > significant when it comes to performance.
Hi Poul-Henning, Which way round do you mean? Apache specifies Keep-Alive in seconds, and my sites will certainly die if I set it to even 1 second. I've no problem with Varnish performance right now, it works really well for me. And solves the problem nicely of how to give Keep-Alive connections to my users whilst maintaining only a small pool of PHP FastCGI processes. But I would like to eliminate my thousands of TIME_WAIT sockets if at all possible, as this represents a scaleability ceiling right now. Of course I could just throw more webservers at the problem - if necessary. Cheers! Nick. _______________________________________________ varnish-misc mailing list [email protected] http://projects.linpro.no/mailman/listinfo/varnish-misc
