From owner-freebsd-questions Fri Feb 16 11:04:18 1996 Return-Path: owner-questions Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id LAA24340 for questions-outgoing; Fri, 16 Feb 1996 11:04:18 -0800 (PST) Received: from halloran-eldar.lcs.mit.edu (halloran-eldar.lcs.mit.edu [18.26.0.159]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id LAA24334 for ; Fri, 16 Feb 1996 11:04:15 -0800 (PST) Received: by halloran-eldar.lcs.mit.edu; (5.65/1.1.8.2/19Aug95-0530PM) id AA20823; Fri, 16 Feb 1996 14:03:41 -0500 Date: Fri, 16 Feb 1996 14:03:41 -0500 From: "Garrett A. Wollman" Message-Id: <9602161903.AA20823@halloran-eldar.lcs.mit.edu> To: Conny Larsson Cc: questions@freebsd.org Subject: TCP/IP performance - urgent! In-Reply-To: References: Sender: owner-questions@freebsd.org Precedence: bulk < said: > In short, I get substantial performance degradation using TCP/IP > stream sockets when the size of the data being sent/received is within > certain intervals. You are probably experiencing performance degradation due to interference between the mbuf and socket buffer layers. In /sys/kern/uipc_socket2.c, look for the line: sb->sb_mbmax = min(cc * 2, sb_max); ...and change the `* 2' factor to something more reasonable (like `* 8'). That should (if this is the source of your problem) cause transfers to run faster, at the expense of more memory usage. -GAWollman -- Garrett A. Wollman | Shashish is simple, it's discreet, it's brief. ... wollman@lcs.mit.edu | Shashish is the bonding of hearts in spite of distance. Opinions not those of| It is a bond more powerful than absence. We like people MIT, LCS, ANA, or NSA| who like Shashish. - Claude McKenzie + Florent Vollant