[OpenAFS] Re: hard link behavior

Andrew Deason adeason@sinenomine.net
Wed, 7 Jul 2010 13:27:45 -0500


On Wed, 07 Jul 2010 18:13:50 +0200
Jaap Winius <jwinius@umrk.nl> wrote:

> > afs has a better scheme for this -- copy on write.  create a volume.
> > dump in your backup.  now vos clone that volume.  this would be a
> > snapshot of that backup.  future backups go into the original volume,
> > vos clone'ing as necessary for the snapshots in time.
> 
> That's interesting for sure, but it sounds as though you're not
> familiar with disk-based backups, because that's no substitute for
> what I have in mind.

Why? Is there any reason besides being limited to 4 snapshots?

Fixing the 4-copy limitation would require changes, but so would the
single-ACL approach.

> All I'm suggesting is that a specific switch be added so that one
> limitation can be traded for another and new things can become
> possible.

I'm getting the feeling that you think that if you have a single-ACL
volume, making cross-dir hardlinks work is as simple as removing the
"you can't have cross-dir hardlinks" check. It's not. The ACL problem is
_one_ reason they are not allowed, but not the only one (although
usually thought of as the "most unsolvable" one). Todd is correct in
that a file in multiple directories violates assumptions in the code,
and also in the current disk format and multiple places in the wire
protocol.

Whereas, if you want to use copy-on-write volume snapshots, I'm pretty
sure you just have to change the disk format limitation of 4 volume
clones.

-- 
Andrew Deason
adeason@sinenomine.net