BackupPC-users

Re: [BackupPC-users] Backing up a BackupPC server

2009-06-03 14:51:46
Subject: Re: [BackupPC-users] Backing up a BackupPC server
From: Holger Parplies <wbppc AT parplies DOT de>
To: Les Mikesell <lesmikesell AT gmail DOT com>
Date: Wed, 3 Jun 2009 20:44:26 +0200
Hi,

Les Mikesell wrote on 2009-06-03 10:11:20 -0500 [Re: [BackupPC-users] Backing 
up a BackupPC server]:
> Craig Barratt wrote:
> > [...]
> > I recently heard about lessfs, which runs on top of FUSE to provide
> > a file system that does block-level de-duplication. [...]
> > 
> > Yes, taking this approach would require a very substantial rewrite.
> > BackupPC would become a lot simpler.  But it also creates a significant
> > issue of backward compatibility.  The only solution would be to provide
> > tools that import the old BackupPC store into a new one.  That is
> > possible, but would likely be very slow.
> 
> [...]
> How hard would it be to simply make the links 
> for pooling an option that you could disable if the filesystem handles 
> it better - as you can already do with compression?

I believe you would simply have to turn off BackupPC_link. No files end up in
the pool directories, each file is created anew (because the relevant pool
file doesn't exist).

The point is, you *can* simplify BackupPC a lot. You can get rid of the code
to determine if and which pool file matches an incoming file. You can probably
handle a lot of things differently (and simpler) than they are handled right
now - file attributes for one thing. In the long run, you would *want* to do
that. Craig, when you write of an import tool, you must be thinking of
something along this line. If you just wanted to drop pooling and keep
everything else as it is, your import tool would be named 'cp'.

But it does seem rather simple to make pooling optional for the start, if you
just want to get rid of hardlink problems and happen to have a file system
where you don't gain anything from pooling anyway.

> I'm not sure why you would need any other format change - and if you had
> a tool to reconstruct the pool links you could switch back if you wanted -
> at some cost in time and CPU.

Here we are back at the original problem. If you have a tool to do that, you
can also copy a pool without handling hardlinks and then re-establish them
(yes, I know, supposing you have enough space). And that *is* feasible right
now, it just comes at a significant cost, because you need to hash and link
every single file in every pc/ directory. For small amounts of data, you won't
mind doing that, but then, for small amounts of data, you don't need to,
because 'rsync -H'/'cp -d' will simply work.

Regards,
Holger

------------------------------------------------------------------------------
OpenSolaris 2009.06 is a cutting edge operating system for enterprises 
looking to deploy the next generation of Solaris that includes the latest 
innovations from Sun and the OpenSource community. Download a copy and 
enjoy capabilities such as Networking, Storage and Virtualization. 
Go to: http://p.sf.net/sfu/opensolaris-get
_______________________________________________
BackupPC-users mailing list
BackupPC-users AT lists.sourceforge DOT net
List:    https://lists.sourceforge.net/lists/listinfo/backuppc-users
Wiki:    http://backuppc.wiki.sourceforge.net
Project: http://backuppc.sourceforge.net/

<Prev in Thread] Current Thread [Next in Thread>