[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: svn filesystem without Berkeley DBM?

From: Ben Collins-Sussman <sussman_at_collab.net>
Date: 2002-03-07 23:30:04 CET

Jens Askengren <jensus@linux.nu> writes:

> I've several times heard the claim that the svn filesystem is an
> interface layer that could be implemented using for example a relational
> database or any other storage system.

That's true. svn 2.0 will probably have optional SQL back-ends for
the filesystem. People are planning on it.

> Why was Berkeley DBM choosen? Why not use regular files? The later seems
> much easier to maintain using regular tools, and possibly more
> robust(?). Or does Berkeley DBM provides some features that simplifies
> the implementation?

Yes, we get more features with Berkeley DB than with the regular
filesystem. Atomic commits are the most important feature. We get
recoverability through logging (the ability to replay transactions.)
We also get the ability to make 'hot backups' while the server is
running. And it's a relatively lightweight dependency, compared to,
say, installing a full SQL server. :-)

To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Mar 7 23:28:17 2002

This is an archived mail posted to the Subversion Dev mailing list.