Community technical support mailing list was retired 2010 and replaced with a professional technical support team. For assistance please contact: Pre-sales Technical support via email to sales@march-hare.com.
On Tue, 19 Sep 2006 16:40:04 -0400, "Bryan Leber" <bryan.leber at fischerinternational.com> wrote: >So if for some reason the server hung up, and we did a reboot the >information would stay no matter what? Another question may be, if the >server crashes and we do a fresh install of cvs server, is there any way >to get the information from past branches and tags back in as if nothing >has ever happened? Retrieving the head branch wouldn't be a problem, but >I'm not sure what about all our previous versions? You seem to not at all have grasped the way CVS and CVSNT works.... On the server there is a service application that listens for client requests via the network. If such are received this application starts the proper cvs process to handle the request. The service application and the cvs executable are part of what you install, upgrade, uninstall and reinstall on the server. BUT, your data, which the cvs process handles, reside in the *repository*, which is a system of folders and files where the version info is stored. Each "real" file has a corresponding so called RCS file in the repository and this file contains *everything* known about the real file including *all* past revisions and all tags. Whatever you do about the CVSNT server application itself will *not* touch the repository and therefore everything there is not affected at all by such operations. The only thing that might be a (remote) risk is if a very buggy version is installed and it trashes the version files during operations. But for this the only real protection is a sensible backup system. You do have nightly backups, right? HTH /Bo (Bo Berglund, developer in Sweden)