Good numbers. I believe that it would be very beneficial to the community if you wouldn't mind sharing the kernel tweaks and squid tweaks that you used to achieve these numbers. Thanks, Josh -----Original Message----- From: GarethC [mailto:gareth@xxxxxxxxxxxxxxxx] Sent: Tuesday, March 20, 2012 12:26 PM To: squid-users@xxxxxxxxxxxxxxx Subject: Re: RPS Hi there, As an example, I set up Squid 2.7 on a HP BL460c (4x Quad-core CPU, 24GB RAM) with Redhat 5 running bonded NICs over a 2x 2G port channel to a Cisco 6509. It took several days of testing to get the Kernel tuned to be able to handle a high rate of connections (things like tcp_max_syn_backlog, tcp_tw_recycle, tcp_rmem, tcp_fin_timeout etc). Squid was also tuned to maximise use of memory, as opposed to disk cache. The maximum sustained connections achieved was in the region of ~2,000 conns per second, and equated to ~980Mbps for a single server. The content that was being requested was purely static html and images. Hope that gives you some sort of view as to what is achievable. Gareth ----- Follow me on... My Blog Twitter LinkedIn Facebook -- View this message in context: http://squid-web-proxy-cache.1019090.n4.nabble.com/RPS-tp4480226p4489420.html Sent from the Squid - Users mailing list archive at Nabble.com.
