[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: mkisofs and reiserfs vs ext3?

I am not on the list, always keep me CCd

Sam Watkins wrote:
> On Sun, Jan 16, 2005 at 11:47:34AM -0500, Wakko Warner wrote:
> > I noticed something strange today with mkisofs.  I was in the process of
> > moving a tree from ext3 to reiserfs.
> > 
> > When I build the directory that I'm going to burn, I always am able to use
> > every possible sector on the media.  In this case, it's a DVD-R (2297888
> > sectors)
> Do you mean if you run mkisofs from the ext3 filesystem, it uses more
> sectors than if you run mkisofs from the reiserfs filesystem with the
> same data?

I now can't reproduce it.  I'm not sure, but I think some hardlinks were
causing it to go nuts.  There should not have been any in the tree but there
may have been and it only shows up on one of the 2 FSs (not sure which at
this point)

> Perhaps it's got something to do with "holes" in files.  Maybe try
> running "cmp" on the two isos to find out where they differ, and "tail -c
> +1234 | xxd" or something to take a look at that part of the file (you
> have to add one to the offset used with tail).

I've already figured it out, but actually I already did this.  I just made
my DVD off the ext3 FS and did a compare on both trees.  (and since I do
archival, I delete the files as I check them, no files were left).  I
thought it might have been a stray directory, but I rm -rfed one tree and cp
-ax the original, same result.

> did you run diff -r to check if the two trees are in fact identical?
> perhaps you were editing a file within the tree when you made the first
> image, and there was a .swp file present or something.

Nope, did rm -rf followed by cp -ax and also did rsync -vaH --delete
<source> <dest>

Either way, I can't reproduce it now.  I'm working on another tree and both
are the exact same.  Wait, actually, maybe I can.  Oddly this time it was on
ISO9660 FS.  I think it has something to do with one of the file names.

[Some time passes as I verify the above]

Ok, it's true.  I duplicated it again.  See below
[root@vegeta:/mnt3] mkisofs -r -no-pad -print-size -quiet /cdrom
[root@vegeta:/mnt3] cp -ax /cdrom/. .
[root@vegeta:/mnt3] mkisofs -r -no-pad -print-size -quiet /mnt3 
[root@vegeta:/mnt3] df -T /cdrom /mnt3
Filesystem    Type   1K-blocks      Used Available Use% Mounted on
/dev/scd0  iso9660     4595776   4595776         0 100% /cdrom
/mnt2/ext3    ext3     6785260   4655736   1784844  73% /mnt3
[root@vegeta:/mnt3] find -type f | wc -l
[root@vegeta:/mnt3] find /cdrom -type f | wc -l
[root@vegeta:/mnt3] cd /mnt2
[root@vegeta:/mnt2] mkdir test
[root@vegeta:/mnt2] cd test
[root@vegeta:/mnt2/test] cp -ax /cdrom/. .
[root@vegeta:/mnt2/test] mkisofs -r -no-pad -print-size -quiet .    
[root@vegeta:/mnt2/test] df -T .
Filesystem    Type   1K-blocks      Used Available Use% Mounted on
	  reiserfs   114558544  54590140  59968404  48% /mnt2

Directory contents available upon request.  All files/directories contain
only A-Z a-z 0-9 _ + . and -

 Lab tests show that use of micro$oft causes cancer in lab animals

Reply to: