From owner-freebsd-hackers Thu Mar 25 15:26:27 1999 Delivered-To: freebsd-hackers@freebsd.org Received: from leaf.lumiere.net (leaf.lumiere.net [207.218.152.15]) by hub.freebsd.org (Postfix) with ESMTP id A792715554; Thu, 25 Mar 1999 15:26:25 -0800 (PST) (envelope-from j@leaf.lumiere.net) Received: (from j@localhost) by leaf.lumiere.net (8.9.2/8.9.1) id PAA02048; Thu, 25 Mar 1999 15:26:04 -0800 (PST) Date: Thu, 25 Mar 1999 15:26:04 -0800 (PST) From: Jesse To: Matthew Dillon Cc: Luigi Rizzo , freebsd-net@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: Re: 3.1-STABLE dies on 40+ connects (resolved) In-Reply-To: <199903252231.OAA03635@apollo.backplane.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG > On the face of it it sounds to me like icecast is perhaps making the > socket buffers bigger --- maybe too big. The default socket buffer I talked to the icecast developers. A simple #undef SOCKET_OPERATIONS 1 stops it from changing the buffer size and they'll be incorporating that change (for *BSD systems). New testing with #undef SOCKET_OPERATIONS 1: I have 121 clients connected and mbuf usage hovers bewteen 500-1100 (as opposed to where it'd die at 4096 with 56 before). Additionally, if I cut it down to 61 clients the mbuf is in the 50-600 range (usually sitting around 90-150 range). I suspect the exponential increase at 91 and 121 clients is because of the massive amount of bandwidth (even for lo0), i.e., around 22Mbps. Thanks everybody for your help! FreeBSD is great and it's 80% about the community. Not that I didn't always know it. =) --- Jesse http://www.lumiere.net/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message