From: "howard chen" Subject: High loading with NFS (Performance issue?) Date: Tue, 15 May 2007 00:55:03 +0800 Message-ID: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: nfs@lists.sourceforge.net Return-path: Received: from sc8-sf-mx2-b.sourceforge.net ([10.3.1.92] helo=mail.sourceforge.net) by sc8-sf-list2-new.sourceforge.net with esmtp (Exim 4.43) id 1Hndp8-0003eC-QX for nfs@lists.sourceforge.net; Mon, 14 May 2007 09:55:02 -0700 Received: from nz-out-0506.google.com ([64.233.162.224]) by mail.sourceforge.net with esmtp (Exim 4.44) id 1HndpA-00018E-Eo for nfs@lists.sourceforge.net; Mon, 14 May 2007 09:55:05 -0700 Received: by nz-out-0506.google.com with SMTP id i11so625493nzi for ; Mon, 14 May 2007 09:55:03 -0700 (PDT) List-Id: "Discussion of NFS under Linux development, interoperability, and testing." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: nfs-bounces@lists.sourceforge.net Errors-To: nfs-bounces@lists.sourceforge.net Hello! My NFS server has the following stats: top -d 5 ============================================ 00:38:21 up 449 days, 31 min, 2 users, load average: 3.89, 3.94, 4.04 143 processes: 142 sleeping, 1 running, 0 zombie, 0 stopped CPU0 states: 19.2% user 4.3% system 0.0% nice 0.0% iowait 75.4% idle CPU1 states: 15.0% user 2.3% system 0.0% nice 0.0% iowait 82.0% idle CPU2 states: 23.2% user 4.3% system 0.0% nice 0.0% iowait 71.4% idle CPU3 states: 12.0% user 2.2% system 0.0% nice 0.0% iowait 85.2% idle Mem: 3355136k av, 3210212k used, 144924k free, 0k shrd, 141920k buff 1011360k active, 926060k inactive Swap: 8184188k av, 230756k used, 7953432k free 1652008k cached iostat ============================================ Linux 2.4.31 avg-cpu: %user %nice %sys %idle 66.75 0.03 20.18 13.05 Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn dev8-0 607.43 95.16 114.58 626487552 754282354 dev8-1 28.92 230.62 182.00 1518207266 1198153062 cat /proc/net/rpc/nfsd ============================================ rc 2126161 488150588 1889246958 fh 288957 2463330121 0 135034 364483 io 2185592286 1281658699 th 8 152389353 957530.010 223376.040 843529.850 0.000 671914.750 520625.610 399462.450 304515.480 0.000 830962.390 ra 16 132542108 407104 72151 88010 34713 54685 46379 22455 47001 24140 89627147 net -1915443474 852923647 1526600175 27092 rpc -1915443589 115 115 0 0 proc2 18 139 866802296 74334343 0 799082289 1 222966486 0 243460468 58134635 56119993 53173221 0 0 119968 3 4934118 395747 [ cat /etc/fstab (SERVER) ============================================ /dev/sda2 /home ext3 defaults 1 2 cat /etc/fstab (CLIENT) ============================================ xxxx nfs rsize=32768,wsize=32768,bg,intr 0 0 The box is a dual Intel Xeon (HT enabled) 3.0Ghz, 4GB RAM server, traffic is considered quite heavy at busy hours (in general, around 7-10GB traffic per day), but...is the above stats look normal? Finally thought: Is it important to mount the partition using noatime? If yes, how many % gain? Thanks.... ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ NFS maillist - NFS@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs