From: Daniel Ellard Subject: NFS trace wish list? Date: Thu, 10 Apr 2003 18:59:11 -0400 (EDT) Sender: nfs-admin@lists.sourceforge.net Message-ID: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Return-path: Received: from bowser.eecs.harvard.edu ([140.247.60.24] helo=mail.eecs.harvard.edu ident=postfix) by sc8-sf-list1.sourceforge.net with esmtp (Exim 3.31-VA-mm2 #1 (Debian)) id 193l0j-0002ds-00 for ; Thu, 10 Apr 2003 15:59:13 -0700 Received: from localhost (localhost [127.0.0.1]) by mail.eecs.harvard.edu (Postfix) with ESMTP id CE67954C4B6 for ; Thu, 10 Apr 2003 18:59:11 -0400 (EDT) To: nfs@lists.sourceforge.net Errors-To: nfs-admin@lists.sourceforge.net List-Help: List-Post: List-Subscribe: , List-Id: Discussion of NFS under Linux development, interoperability, and testing. List-Unsubscribe: , List-Archive: Hello-- As part of my research in file systems, I have written an NFS tracing program (based largely on tcpdump, but with the ability to capture more information) and a bunch of PERL scripts for scanning through the traces and finding useful info. It has been very useful to my work, and from my conversations with other people, I think it might be a useful tool to other people. I'd like to document and bundle up the tools I have and/or add some new tools that make them useful in a general context. However, I don't really know what functionality other people are looking for. Here's where you can help. I'm trying to put together a wish list of NFS analysis tools -- what tools would you like that you haven't got? Some ideas people have already suggested: - Load pattern analysis. If your network is like ours, then 3am is not a good time to do backups. Everyone thinks the system is idle in the middle of the night, so they all schedule their big jobs to run in the small hours of the morning... - Hot file/directory spotting. Watch for files that get read a lot by lots of clients, but not written. (maybe they can be replicated, possibly closer to the clients) - Hot client spotting. For servers with multiple interfaces on different subnets, figure out which clients are soaking up all the bandwidth and spread them out between the subnets to balance the load. These are pretty mundane. Tell me what you really want... Thanks, -Dan ------------------------------------------------------- This SF.net email is sponsored by: Etnus, makers of TotalView, The debugger for complex code. Debugging C/C++ programs can leave you feeling lost and disoriented. TotalView can help you find your way. Available on major UNIX and Linux platforms. Try it free. www.etnus.com _______________________________________________ NFS maillist - NFS@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs