All of lore.kernel.org
 help / color / mirror / Atom feed
From: Heinz-Josef Claes <hjclaes@web.de>
To: Joe Cooper <joe@swelltech.com>
Cc: Hans Reiser <reiser@namesys.com>,
	Debian User <mtoseland@blueyonder.co.uk>,
	reiserfs-list@namesys.com
Subject: Re: duplicate files and recent changes
Date: 09 Jun 2002 14:17:14 +0200	[thread overview]
Message-ID: <1023625034.1041.10.camel@kakapo> (raw)
In-Reply-To: <3D0200CF.1050604@swelltech.com>

Am Sam, 2002-06-08 um 15.04 schrieb Joe Cooper:
> Hans Reiser wrote:
>  > Debian User wrote:
>  >
>  >> On Fri, Jun 07, 2002 at 10:53:57AM +0200, Heinz-Josef Claes wrote:
>  >>
>  >>
>  >>>
>  >>> But I've written a well tested tool for backing up directories on
>  >>>  another disk which uses md5 sums to recognize same files (even
>  >>> in one backup) and hard links them. By this way, it combines the
>  >>> advantages of full and incremental backup. It also uses
>  >>> compression and has a lot of other options and optimisations. It
>  >>> runs best with reiserfs, because reiserfs has no preconfigured
>  >>> number of inodes and is fast with hard linking. So, if you need
>  >>> to backup, try http://sourceforge.net/projects/storebackup
>  >>>
>  >>
>  > Would anyone care to check this tool out, and post a review?
>  >
>  > Hans
> 
> Not to discount the work Heinz-Josef has done, but rdiff-backup may be a 
> better way to achieve a similar goal.  It does not have the benefit of 
> being "best with reiserfs" but it works great on ReiserFS partitions (as 
> well as ext2/3 and probably anything else).
> 
> http://www.stanford.edu/~bescoto/rdiff-backup/
> 
> Binary diffs are fun.  I'd be surprised if an MD5/hardlink system could 
> match it for compression of increments over time, or beat the 
> flexibility of being able to keep potentially weeks or months worth of 
> daily diffs in a very small space.  I may be wrong, of course.  But 
> rdiff-backup is working great for me.
> -- 
> Joe Cooper <joe@swelltech.com>
> Web caching appliances and support.
> http://www.swelltech.com
> 
It depends on what you want to do. If you want backup big files
with only some changes in it (eg. vmware), you can simply forget
my tool.
But if you have "normal" user files, you have the possiblity to backup
them with the following advantages:
- reduced space (because of compression) [from my understanding rdiff
cannot do this, but I don't really know]
- every backup is a full backup (because of hard links). That means,
you can delete whatever backup you want (A future version will be able
to delete parts of an old backup for different livetime). So the tool
has features to delete special week days of old backups (and others).
[with rdiff, you have deltas, which should make the world much more
complicated]
- users can access there backups directly via the filesystem, eg. using
NFS or SAMBA or whatever. This helps to reduce administrational overhead
a lot.

(The principal idea came from the snapshot of netapp files, but it's not
implemented on the filesystem level and has a different behaviour. Only
for the user, it's seems to be a little similar.)

So, rdiff is another solution for the same problems with different
advantages and disadvantages ;-)

Regards,
Heinz-Josef





  reply	other threads:[~2002-06-09 12:17 UTC|newest]

Thread overview: 99+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2002-06-06  5:30 duplicate files and recent changes S. Alexander Jacobson
2002-06-06  5:45 ` Oleg Drokin
2002-06-06  7:20   ` Robert Brockway
2002-06-06  7:40     ` S. Alexander Jacobson
2002-06-06  7:50       ` Hans Reiser
2002-06-06  7:33   ` S. Alexander Jacobson
2002-06-06  9:25     ` Oleg Drokin
2002-06-06 13:58       ` Valdis.Kletnieks
2002-06-06 17:51         ` Hubert Chan
2002-06-07  8:11           ` The Amazing Dragon
2002-06-07 15:55             ` Hubert Chan
2002-06-06  7:47   ` Hans Reiser
2002-06-06 13:54 ` Valdis.Kletnieks
2002-06-07  8:53 ` Heinz-Josef Claes
2002-06-08 12:50   ` Debian User
2002-06-08 12:47     ` Hans Reiser
2002-06-08 13:04       ` Joe Cooper
2002-06-09 12:17         ` Heinz-Josef Claes [this message]
2002-06-09 14:30           ` Heinz-Josef Claes
  -- strict thread matches above, loose matches on Subject: below --
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:54 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 18:14 ` S. Alexander Jacobson
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 18:28 ` Richard Thornton
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 13:58 Valdis.Kletnieks
2002-06-06 17:51 Hubert Chan
2002-06-06 17:51 Hubert Chan
2002-06-06 17:51 Hubert Chan
2002-06-06 17:51 Hubert Chan
2002-06-06 17:51 Hubert Chan
2002-06-06 17:51 Hubert Chan
2002-06-06 17:51 Hubert Chan
2002-06-06 17:51 Hubert Chan
2002-06-06 17:51 Hubert Chan
2002-06-06 18:14 S. Alexander Jacobson
2002-06-06 18:14 S. Alexander Jacobson
2002-06-06 18:14 S. Alexander Jacobson
2002-06-06 18:14 S. Alexander Jacobson
2002-06-06 18:14 S. Alexander Jacobson
2002-06-06 18:14 S. Alexander Jacobson
2002-06-06 18:14 S. Alexander Jacobson
2002-06-06 18:14 S. Alexander Jacobson
2002-06-06 18:28 Richard Thornton
2002-06-06 18:28 Richard Thornton
2002-06-06 18:28 Richard Thornton
2002-06-07  4:57 ` Oleg Drokin
2002-06-06 18:28 Richard Thornton

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1023625034.1041.10.camel@kakapo \
    --to=hjclaes@web.de \
    --cc=joe@swelltech.com \
    --cc=mtoseland@blueyonder.co.uk \
    --cc=reiser@namesys.com \
    --cc=reiserfs-list@namesys.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.