Date: Thu, 12 Feb 1998 12:03:36 -0500 (EST) From: Charles Owens <owensc@enc.edu> To: hackers list FreeBSD <freebsd-hackers@FreeBSD.ORG> Subject: Coda FS: FBSD port done!, but development favors Linux Message-ID: <Pine.BSF.3.95q.980212112548.17150D-100000@itsdsv2.enc.edu>
next in thread | raw e-mail | index | archive | help
Hi all, I've been researching CMU's Coda filesystem (see http://www.coda.cs.cmu.edu/index.html) with hopes of doing an implementation. This distributed filesystem has some very interesting properties that should eventually allow the creation of large, fault-tolerant filespace capable of supporting a vast number of clients. (ie. AFS, but better in many ways) I just had a nice chat with Peter Braam, CMU faculty and Coda development leader, in which he stated the following: * The FreeBSD port is done (I'm not sure if it's actually downloadable yet) * Development, particularly in the area of scalability, is focused on Linux. Why? His stated reasons: * Linux's ext2fs filesystem is much faster than *BSD's ffs (How good is FreeBSD's ext2fs support these days? Is it in 2.2.6 or must we wait for 3.0?) * Several other Linux-based research groups are hacking on Coda, and RedHat is showing interest * Current work is being done to develop Linux kernel extensions that will allow access to files via raw inodes. This development is seen as key to allowing Coda to support large filespaces with reasonable performance. See this URL for Peter's notes on these extensions: http://telemann.coda.cs.cmu.edu/maillists/linux-coda/0225.html Peter was very interested in seeing FreeBSD development continue, but regretted that he had no programmers to spare on creating similar FreeBSD kernel extensions. Any reactions to this? I personally think that Coda could be the greatest thing since sliced bread... we certainly don't want FreeBSD to miss out. Does the idea of these kernel extensions making their way into the FreeBSD kernel rub anyone the wrong way? Is there a better way to go? I know we have a penchant for doing things the "right" way. :-) The Coda implementation that I'd like to do would serve a 60 Gig filespace (eventually two or three times larger) to upwards to 1500 clients (actually SMB, AFS, and HTTP clients which would connect to N Coda "clients"). This is far larger than anything anyone has done with Coda to date, and certainly got Peter's attention. Unfortunately, he thinks that he's about a year away from being able to reliably work with filespaces of this magnitude. So, I'm stuck with a NFS/rdist mess for now (my implementation needs to be live in summer '98). I'll stop rambling now... ;-) --- ------------------------------------------------------------------------- Charles N. Owens Email: owensc@enc.edu http://www.enc.edu/~owensc Network & Systems Administrator Information Technology Services "Outside of a dog, a book is a man's Eastern Nazarene College best friend. Inside of a dog it's too dark to read." - Groucho Marx ------------------------------------------------------------------------- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.3.95q.980212112548.17150D-100000>