[460] in Info-AFS_Redistribution
Re: AFS and NeXTs
daemon@ATHENA.MIT.EDU (Marc Horowitz)
Tue Nov 19 01:42:37 1991
To: Craig_Everhart@transarc.com, info-afs@transarc.com
Cc: Cal_Thixton@next.com
Cc: "Michael T. Stolarchuk" <mts@terminator.cc.umich.edu>
In-Reply-To: [445] in Info-AFS_Redistribution
Reply-To: Marc Horowitz <marc@MIT.EDU>
Date: Tue, 19 Nov 91 00:38:30 EST
From: Marc Horowitz <marc@athena.mit.edu>
>> That is, I don't think that there's a pervasive problem with lying
>> about the vcache state. And my proposal, at least, deals only with
>> cross-cell mount points, not same-cell mount points. (Most cross-cell
>> mount points reside under /afs itself.)
I think there's a problem with doing anything differently between
NeXT's and other platforms. The real problem here is that the vnode
layer provides a convenient abstraction in some ways, but like all
abstractions, it hides information. It would be useful to have a
supported, portable interface to determine "stuff" about a given file,
even something as simple as an isaremotemountpoint() call.
Unfortunately, we don't have that. So, I think this falls under the
classification of "evil, but necessary for performance reasons". The
Browser should know an AFS mountpoint when it sees one (this is not a
lot of code), and Transarc should look at POSIX (or whichever standard
for filesystems is likely to get stuffed down our throats) and make
sure this can be done more intelligently in the future.
Marc
