From: Noveck, Dave (dave.noveck@netapp.com)
Date: 06/10/99-04:14:08 PM Z
Message-ID: <7F608EC0BDE6D111B53A00805FA7F7DA0330363A@TAHOE.netapp.com>
From: "Noveck, Dave" <dave.noveck@netapp.com>
Subject: fsid.{major,minor} not required attributes?
Date: Thu, 10 Jun 1999 14:14:08 -0700
I just discovered that the fsid twins are not required attrbutes.
This seems bizarre to me. We have this whole mechanism that
relies on noting a change in the fsid attributes going down the
directory tree. Right now, the spec (section 6.7) doesn't even
reflect the fact that there are two of these but the thought of
trying to properly document the rules in the face of all the
combinations of fsid's when some or all of the attributes are
not present is daunting. Why bother?
These should be required. A server is either going to have the
concept of multiple filesystems or not. If it does, it is going
to have to designate them in some way internally and I can't see
why it can't some up with some value for the fsid fields. In fact
it will have to or the client will have no way of knowing that
we have crossed into it. If a server doesn't have file systems
(everything is one) then it can return zero for fsid all the time.
The possiblity of fsid.major or fsid.minor or both not being
present just adds complexity to the spec and the code.
This archive was generated by hypermail 2.1.2 : 03/04/05-01:47:12 AM Z CST