From owner-freebsd-stable Wed May 1 17:43:42 2002 Delivered-To: freebsd-stable@freebsd.org Received: from vinyl2.sentex.ca (vinyl2.sentex.ca [199.212.134.13]) by hub.freebsd.org (Postfix) with ESMTP id AFFCF37B405 for ; Wed, 1 May 2002 17:43:34 -0700 (PDT) Received: from house.sentex.net (cage.simianscience.com [64.7.134.1]) (authenticated bits=0) by vinyl2.sentex.ca (8.12.3/8.12.2) with ESMTP id g420hMdd003457 (version=TLSv1/SSLv3 cipher=DES-CBC3-SHA bits=168 verify=NO); Wed, 1 May 2002 20:43:24 -0400 (EDT) (envelope-from mike@sentex.net) Message-Id: <5.1.0.14.0.20020501202847.065fdb80@192.168.0.12> X-Sender: mdtancsa@192.168.0.12 X-Mailer: QUALCOMM Windows Eudora Version 5.1 Date: Wed, 01 May 2002 20:43:31 -0400 To: Peter Jeremy From: Mike Tancsa Subject: Re: MBUF allocation failures Cc: freebsd-stable@FreeBSD.ORG In-Reply-To: <20020502080023.X25071@gsmx07.alcatel.com.au> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG Interesting, I have suddenly started seeing this on a couple of stable boxes from March, and one from FreeBSD 4.5-STABLE #1: Sun Apr 21 14:40:29 EDT 2002 m_clalloc failed, consider increase NMBCLUSTERS value m_clalloc failed, consider increase NMBCLUSTERS value m_clalloc failed, consider increase NMBCLUSTERS value One seems to recover after a period of time, but the other froze up twice. netstat -m 68/4304/14336 mbufs in use (current/peak/max): 68 mbufs allocated to data 65/3584/3584 mbuf clusters in use (current/peak/max) 8244 Kbytes allocated to network (76% of mb_map in use) 16095 requests for memory denied 39 requests for memory delayed 0 calls to protocol drain routines Whats strange, is that this machine has functioned happily with the same load and users behind it for some time. The other machine, the one that froze auth% sysctl -a | grep clust kern.ipc.nmbclusters: 4576 auth% netstat -m 161/816/18304 mbufs in use (current/peak/max): 158 mbufs allocated to data 3 mbufs allocated to packet headers 130/354/4576 mbuf clusters in use (current/peak/max) 912 Kbytes allocated to network (6% of mb_map in use) 0 requests for memory denied 0 requests for memory delayed 0 calls to protocol drain routines auth% Is essentially big ass DNS server for my dialups and for my in bound SMTP servers. Some remote DoS ? Or local bug ? The only thing strange about the one hitting the MBUF limits is that there is another box that dials into it for a PPP session. What is the best way to try and track this down ? For now I am going to echo 'kern.ipc.nmbclusters=16384' >> /boot/loader.conf and see if the same machine pushes up to the new max. ---Mike At 08:00 AM 5/2/2002 +1000, Peter Jeremy wrote: >Last nights daily report on one of my machines included: > > m_clalloc failed, consider increase NMBCLUSTERS value > > m_retry failed, consider increase mbuf value > >But: >aalp03# netstat -m >71/1264/16384 mbufs in use (current/peak/max): > 71 mbufs allocated to data >69/1242/4096 mbuf clusters in use (current/peak/max) >2800 Kbytes allocated to network (22% of mb_map in use) >34 requests for memory denied >0 requests for memory delayed >0 calls to protocol drain routines >aalp03# uname -a > >Why am I getting m_clalloc() failures when the reported peak usage is >well below the maximum reported by netstat? This is an old -STABLE >(just before 4.5-RC1). > >Peter > >To Unsubscribe: send mail to majordomo@FreeBSD.org >with "unsubscribe freebsd-stable" in the body of the message -------------------------------------------------------------------- Mike Tancsa, tel +1 519 651 3400 Sentex Communications, mike@sentex.net Providing Internet since 1994 www.sentex.net Cambridge, Ontario Canada www.sentex.net/mike To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message