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

Re: How to do a failover from primary to secondary system

From: Andy Levy <andy.levy_at_gmail.com>
Date: Mon, 28 Jul 2008 11:43:27 -0400

On Mon, Jul 28, 2008 at 11:39, Baeriswyl Kuno - Extern (IT-BA-MV)
<kuno.baeriswyl_at_sbb.ch> wrote:
> Hello!
> What is the recommend way to do a failover to a secondary system?
> We do a nightly backups of our repositories. In an emergency case, e.g.
> hardware failure, we would like to set up a second server using the backup
> from the last night. We tried to use the relocate functionality, that is
> implemented in TortoiseSVN and Subclipse. Though, we stated that the local
> repository and the server must be in sync in order to make it work with a
> relocation. Typically, this situation isn't guaranteed. Particularly,
> revisions in the local copy that weren't available on the server anymore let
> to errors while updating to the new repository instance.
> How can the developers reconnect to this new subversion instance without a
> fresh checkout and loosing data?

The only data they should lose is whatever they don't back up before
performing a fresh checkout.

> In order to achieve this objective, do I
> need to set up a slave instance that is continuously synchronised with its
> master?

Yes, otherwise there will always be a discrepancy between your master
& slave after someone has performed a commit following your last

Have you looked at using the svnsync (new with SVN 1.4) feature for
keeping things in sync with each commit?

To unsubscribe, e-mail: users-unsubscribe_at_subversion.tigris.org
For additional commands, e-mail: users-help_at_subversion.tigris.org
Received on 2008-07-28 17:43:53 CEST

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

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