[8422] in Info-AFS_Redistribution
Re: afs/dfs/coda
daemon@ATHENA.MIT.EDU (Thomas Vincent - 409-974-3678)
Mon Jan 22 22:50:09 2001
Date: Mon, 22 Jan 2001 19:18:33 -0800 (PST)
From: Thomas Vincent - 409-974-3678 <thomasv@apple.com>
To: iddwb <David.Bear@asu.edu>
Cc: <info-afs@transarc.com>
In-Reply-To: <Pine.LNX.4.21.0101230011100.2558-100000@moroni.pp.asu.edu>
Message-Id: <Pine.GSO.4.31.0101221908540.25281-100000@gala.apple.com>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Hi David,
I checked out Coda, and talked to one of the engineers at Caldera where
they did a more thorough evalution of Coda then I did. Coda is not ready
for prime-time.
Anything you run over high latency/low bandwidth connections will run
poorly.
All my research shows DFS to be a very poor solution. Most sites have
given up on implimenting it. UPenn is one of the few that I have heard is
successfull. There is some new file systems like Web FS (runs as an
extension of http) that seem immature at best. Plus of you have a high
latency, and low bandwidth connection your probably have a high rate of
packet loss. A TCP based protocol (NFS, http based file systems) would
probably slow you down even more.
I think you are stuck, unless you upgrade your connection.
What kind of connection are coming over?
Thomas Vincent
Apple Information Systems & Technology
My opinions are my own, not my employers.
On Tue, 23 Jan 2001, iddwb wrote:
>
> Afs is a fine idea -- but it really stinks over low bandwidth high latency
> networks. ARe there any good comparisons between afs/dfs and coda? Is
> coda ready for prime time? Are there any coda to afs translator?
>
> David Bear
> College of Public Programs/ASU
>
>
