From mboxrd@z Thu Jan 1 00:00:00 1970 From: "J.Rinas" Subject: invalidate cache / NFS performance on a diskless system Date: Thu, 9 Dec 2004 17:18:48 +0100 Message-ID: <200412091718.48973.jrinas@gmx.de> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-15" Return-path: Received: from sc8-sf-mx1-b.sourceforge.net ([10.3.1.11] helo=sc8-sf-mx1.sourceforge.net) by sc8-sf-list2.sourceforge.net with esmtp (Exim 4.30) id 1CcdJX-0004uQ-T0 for nfs@lists.sourceforge.net; Thu, 09 Dec 2004 21:27:35 -0800 Received: from mail0.ewetel.de ([212.6.122.10]) by sc8-sf-mx1.sourceforge.net with esmtp (Exim 4.41) id 1CcdJV-00041O-RT for nfs@lists.sourceforge.net; Thu, 09 Dec 2004 21:27:35 -0800 Received: from jrdesk.example.com (dialin-80-228-48-015.ewe-ip-backbone.de [80.228.48.15]) by mail0.ewetel.de (8.12.1/8.12.9) with ESMTP id iB9GIme7028116 for ; Thu, 9 Dec 2004 17:18:48 +0100 (MET) Received: from [127.0.0.1] (helo=localhost) by jrdesk.example.com with esmtp (Exim 4.34) id 1CcR0D-0002ma-CS for nfs@lists.sourceforge.net; Thu, 09 Dec 2004 17:18:49 +0100 To: nfs@lists.sourceforge.net Sender: nfs-admin@lists.sourceforge.net Errors-To: nfs-admin@lists.sourceforge.net List-Unsubscribe: , List-Id: Discussion of NFS under Linux development, interoperability, and testing. List-Post: List-Help: List-Subscribe: , List-Archive: Hello! We've set up our network using a linux 2.6.9 server and abt. 25-30 linux clients using NFS. The clients are booting through PXE and work in a diskless fashion. i.e.=20 they mount their root directory via NFS only /var and /tmp are on a tmpfs device. To improve performance we have currently set the root directory=20 (including the /usr tree) to use the mount options ro, noatime, v3, rsize=3D32768,wsize=3D32768, actimeo=3D180, hard,tcp,nocto,nolock To further improve the performance of the clients,=20 I'd like to increase actimeo to abt. 9 hours. But this leads to severe problems, if some small configurations are done=20 during the day (e.g. installing new soft or configuration work...): The changes do not apprear on the client nodes in reasonable time. So I'd like to know if there is a switch to invalidate the whole NFS / VFS?= =20 cache at once! (e.g. an echo something > /proc/somewhere )? An there any other hints/experiences to improve performance on such a syste= m? (Our new server with 10k rpm Disks is already ordered!) tnx fr ur help J=FCrgen ------------------------------------------------------- SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://productguide.itmanagersjournal.com/ _______________________________________________ NFS maillist - NFS@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs