From owner-freebsd-isp Thu Aug 12 11:43:23 1999 Delivered-To: freebsd-isp@freebsd.org Received: from dt011n65.san.rr.com (dt011n65.san.rr.com [204.210.13.101]) by hub.freebsd.org (Postfix) with ESMTP id D8E8B14DB9 for ; Thu, 12 Aug 1999 11:43:14 -0700 (PDT) (envelope-from Doug@gorean.org) Received: from localhost (doug@localhost) by dt011n65.san.rr.com (8.9.3/8.8.8) with ESMTP id LAA81168; Thu, 12 Aug 1999 11:43:13 -0700 (PDT) (envelope-from Doug@gorean.org) Date: Thu, 12 Aug 1999 11:43:13 -0700 (PDT) From: Doug X-Sender: doug@dt011n65.san.rr.com To: Dag-Erling Smorgrav Cc: ndear@areti.net, freebsd-isp@FreeBSD.ORG Subject: Re: mbuf clusters. In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-isp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 12 Aug 1999, Dag-Erling Smorgrav wrote: > Doug writes: > > I ran highly loaded IRC servers like that for years. In 2.2.8 increasing > > maxusers beyond 512 or NMBCLUSTERS above 15k is a pessimization, but those > > levels are safe as long as you have the physical ram to handle it. > > No, they're not. There are "sweet spots" (specific values of maxusers > which work), but increasing maxusers beyond 128 on quarter-gig-kva > systems basically makes the system unstable. If you believe that you should document it with accompanying evidence to support your claims. I'm talking real world experience on several machines (c. 20) over the course of two years, so I'm not going to back down from saying that those settings do work. Doug -- On account of being a democracy and run by the people, we are the only nation in the world that has to keep a government four years, no matter what it does. -- Will Rogers To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-isp" in the body of the message