From owner-freebsd-questions@FreeBSD.ORG Tue Aug 28 12:07:10 2007 Return-Path: Delivered-To: freebsd-questions@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 9B2DA16A420 for ; Tue, 28 Aug 2007 12:07:10 +0000 (UTC) (envelope-from frank@altpeter.de) Received: from canismajor.corpex-net.de (canismajor.corpex-net.de [194.6.192.31]) by mx1.freebsd.org (Postfix) with ESMTP id 6323213C46A for ; Tue, 28 Aug 2007 12:07:10 +0000 (UTC) (envelope-from frank@altpeter.de) Received: from pegasus.corpex.de ([194.6.192.248]) by canismajor.corpex-net.de with esmtpsa (TLSv1:AES256-SHA:256) (Exim 4.66 (FreeBSD)) (envelope-from ) id 1IPyPo-0008HY-Di; Tue, 28 Aug 2007 12:35:20 +0200 Received: by pegasus.corpex.de (sSMTP sendmail emulation); Tue, 28 Aug 2007 12:35:20 +0200 Date: Tue, 28 Aug 2007 12:35:20 +0200 From: Frank Altpeter To: freebsd-questions@freebsd.org Message-ID: <20070828103520.GB37646@corpex.de> Mail-Followup-To: Frank Altpeter , freebsd-questions@freebsd.org MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit User-Agent: Mutt/1.5.15 (2007-04-06) X-CPX-Track: canismajor.corpex-net.de;1IPyPo-0008HY-Di;194.6.192.248 Subject: lost DNS queries X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 28 Aug 2007 12:07:10 -0000 Hi list, I'm trying to resolve a strange problem on my DNS resolver system for some time now, and still have no clue where to catch the reason. Perhaps someone here has some hints for me... The prerequisites: DNS Server with FreeBSD 6.2-RELEASE-p4 running bind as resolver (hardware sufficient, e.g. dual Xeon 3.0 GHz with 2GB RAM, OS on SCSI RAID1). DNS client machine connected via HP ProCurve switch and in the same VLAN as the DNS server machine. Test software: dnsperf Command executed: dnsperf -d /tmp/input-10000.txt -s ${DNSSERVER} -c -A Result output of the command: ---------------------------------------------------------------------------- Parse input file: once Ended due to: reaching end of file Queries sent: 10000 queries Queries completed: 9770 queries Queries lost: 230 queries Returned NOERROR: 7235 queries Returned SERVFAIL: 107 queries Returned NXDOMAIN: 2428 queries Avg request size: 41 bytes Avg response size: 184 bytes Percentage completed: 97.70% Percentage lost: 2.30% Started at: Tue Aug 28 12:22:51 2007 Finished at: Tue Aug 28 12:25:30 2007 Ran for: 159.355209 seconds Queries per second: 61.309574 qps ---------------------------------------------------------------------------- I tweaked several settings on the dns server to reduce the percentage lost: sysctl net.inet.icmp.icmplim=1024 sysctl net.inet.raw.recvspace=65536 sysctl net.inet.udp.recvspace=256000 sysctl net.inet.tcp.sendspace=65536 sysctl net.inet.tcp.recvspace=65536 sysctl kern.ipc.somaxconn=512 sysctl kern.ipc.maxsockbuf=8388608 but still i'm not able to find a reason why so much queries are lost and i don't see any reason for this. It even happens when dnsperf is started directly on the dns server. Does anyone has an idea where to search for this problem? Le deagh dhùraghd, Frank Altpeter -- *** FA-RIPE *** http://blog.foxalpha.de/ *** SIG at 11 *** Join XING - the Open Business Club: http://www.xing.com/go/invite/27666.e9249b