From owner-freebsd-stable Mon Apr 17 21:44: 0 2000 Delivered-To: freebsd-stable@freebsd.org Received: from alcanet.com.au (mail.alcanet.com.au [203.62.196.10]) by hub.freebsd.org (Postfix) with ESMTP id ECF0737BB85 for ; Mon, 17 Apr 2000 21:43:53 -0700 (PDT) (envelope-from jeremyp@gsmx07.alcatel.com.au) Received: by border.alcanet.com.au id <115226>; Tue, 18 Apr 2000 14:44:18 +1000 Content-return: prohibited From: Peter Jeremy Subject: Re: 3-STABLE hangs In-reply-to: <20000417112535.A26098@gsmx07.alcatel.com.au>; from peter.jeremy@alcatel.com.au on Mon, Apr 17, 2000 at 11:25:36AM +1000 To: stable@FreeBSD.ORG Message-Id: <00Apr18.144418est.115226@border.alcanet.com.au> MIME-version: 1.0 X-Mailer: Mutt 1.0i Content-type: text/plain; charset=us-ascii References: <20000417112535.A26098@gsmx07.alcatel.com.au> Date: Tue, 18 Apr 2000 14:44:18 +1000 Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On 2000-Apr-17 11:25:36 +1000, Peter Jeremy wrote: >I have a Dell OptiPlex GXi (Pentium-133 with 96MB RAM) running >3-STABLE (from last night) which is used to simulate a WAN (using >dummynet). Under normal operation, it hangs about every 2 weeks, >but I can make it happen in about 15 minutes by loading up the >box. My testing is to saturate the networks with a mixture of FTP and scp traffic requiring a mixture of NAT, delaying or neither. At the same time, I perform a 'make buildworld' and 'dd if=/dev/rwd0c of=/dev/null bs=64k'. This fairly reliably killed the original configuration within 5 minutes. So far, I've tried: - swapping all the DIMMs. This has had no effect. - Changing the amount of RAM. The lockups take much longer to occur when I only have 32MB or 64MB RAM. - Increasing UPAGES to 3 (on the offchance that nested interrupts were overflowing the kernel stack). This had no effect. The only oddity I noticed was that with smaller amounts of RAM, the transmit side of fxp2 would sometimes stop (leading to `out of buffer space' errors from routed). "ifconfig fxp2 down; ifconfig fxp2 ip" fixed it. I haven't seen this at 96MB, or on any other interface. There's no evidence of mbuf (or any other memory) shortage using "vmstat -m" or "netstat -m". I've just about run out of ideas. Does anyone else have any? Peter To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message