Re: fsync() fails under NFS, right?

New Message Reply About this list Date view Thread view Subject view Author view Attachment view

From: braver@pobox.com
Date: 09/15/99-06:01:01 PM Z


From: braver@pobox.com
Date: Wed, 15 Sep 1999 19:01:01 -0400
Message-Id: <199909152301.TAA04765@roll.setup.org>
Subject: Re: fsync() fails under NFS, right?


I didn't describe the fault in detail, but clearly
indicated that TCP returns were not corresponding
to eventual safe write at all points through.  We
thoroughly checked against the TCP specs and sure
there was a problem.  Your news are no news, and
there's no misinformation.

I am an avid Linux supporter and point out ways to
improve NFS implementation, etc., to make it
better.  There's no Linux bashing; I have it
running on all my machines, and my home and office
networks use Linux and Linux NFS.

Your original material described a technical
agenda of kernel optimization, which we believe
should be done more carefully when implementing
complex semantics such as NFS.  Moreover, such
protocols must be designed with the appreciation
and apprehension of such optimizations.  Thus it
warranted a response based on our experience with
problems of such optimizations, and as being fully
on-topic.

As per your suggestion about other forums, I
received specific pointers from other members,
better phrased and more helpful.  (Of course, I'll
keep rebutting any ad hominem stuff ad infinitum.)
Thanks.

-- 
Cheers,
Alexy V. Khrabrov -- www.suffix.com -- braver@pobox.com


New Message Reply About this list Date view Thread view Subject view Author view Attachment view

This archive was generated by hypermail 2.1.2 : 03/04/05-01:47:35 AM Z CST