[prev in list] [next in list] [prev in thread] [next in thread] 

List:       reiserfs-devel
Subject:    (reiserfs) Re: Vladimir gets to the bottom of [Fwd: (reiserfs) Is this bad?]
From:       "Stephen C. Tweedie" <sct () redhat ! com>
Date:       1999-02-10 14:23:36
[Download RAW message or body]

Hi,

On Mon, 08 Feb 1999 11:16:45 -0800, Hans Reiser <reiser@idiom.com> said:

> I am going to put this into our paper as a concrete example of the
> possible effects of changing our tuning defines.

Yes, I've just seen the same tar behaviour here on ext2.  It appears
that tar is doing 4k page-aligned reads from the original file, and
because the internal tar file's headers are only 512-byte aligned, we
can get those 4k boundaries appearing at arbitrary 512-byte offsets
within the output files.

For what it's worth, this is going to be a pretty nasty worst-case
scenario for BSD FFS with fragments too.  I'm quite surprised gnu tar
isn't trying to page-align (or fs-block-align) the output itself.

--Stephen

[prev in list] [next in list] [prev in thread] [next in thread]