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

RE: Re: Alternatives for remote access?

From: Bill Tutt <rassilon_at_lyra.org>
Date: 2002-08-30 22:52:07 CEST

> From: Scott Lenser [mailto:slenser@cs.cmu.edu]
> > > The first and biggest obstacle is that the members of my team
> > have
> > > expertise in setting up a web server. Network administration and
> > network
> > > coding are far from our area of expertise. CVS does not require
> > to be
> > > an expert in anything to install and use it. Subversion obviously
> > > requires
> > > more work and greater expertise to get your repository up and
> > Do
> > > you not consider this to be a significant drawback?
> > >
> >
> > That's so not true. CVS does require you to be an expert in lots of
> > things to install it and use it well. Including making sure regular
> > backups happen, and all of the odd/fun situations that remote CVS
> > repositories get into. (wedged locks, etc...)
> >
> I disagree. I use CVS at work to support me and about 8 other
> of the software I work on. I've done all of the setup work. The
> already had to have ssh setup to allow secure remote access (we don't
> any firewalls). Compared to creating a local CVS repository, the only
> extra
> work was adding a few additional user accounts to the machine hosting
> repository. I've never had any problems with wedged locks or anything
> else
> on the server that required my interaction to resolve. Total setup
> a few commands to create the repository, creation of some user
> installation of SSH key-pairs for these users, and addition of a
> environment variables for CVS. In comparison, I spent 4 hours once
> to set up a SVN repository for the code over encrypted connections.
> was before I gave up. I'm certainly no expert on web servers which
> probably
> added to the setup time for SVN, but I've never had to be to do
> control before. As for backups, with CVS I can do a backup using scp
> which
> is really easy to do and easily automated with a cron job. With SVN,
> would
> have to look up the backup procedures for BDB and find out the proper
> to backup since I don't know if simply copying the files would work.

Then you've been lucky and haven't been using CVS at a high enough
scalability threshold to run into these issues.

You can't just copy files out of CVS, and you can't just copy a SVN
datastore. The CVS data store is very non-atomic, and very non-ACID
like. We provide scripts and (hopefully instructions now) on how to do
BDB backups as well as how to dump/load your repository using svnadmin.
i.e. our Python hot backup script that does the BDB backup discipline
after every checkin.

Therefore, if we reduced the setup burden for you, your experience would
be very similar.


To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri Aug 30 22:52:39 2002

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

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.