[cvsnt] Migrate repository from CVS (GNU) to CVSNT
Tim Chippington Derrick
tim at chippingtonderrick.co.uk
Mon Jan 22 16:43:38 GMT 2007
Just a thought... I haven't done a CVS migration from GNU CVS to
CVSNT, or from Unix/Linux to Windows, but even moving my repository
from an existing Windows+CVSNT to a new Windows+CVSNT box I would
always use CVSNT to create the repository first and then copy the
repository contents into the (newly created and initialised)
repository. I've done this many times and never had a problem; I
don't know what extra stuff CVSNT does or needs in the repository
beyond what GNU CVS does, but can imagine that not having a correctly
initialised CVSNT repository as a starting point could be a problem.
At 14:26 22/01/2007, you wrote:
>We have an existing GNU CVS v 1.11 server, running on Solaris. We are
>migrating to CVSNT 2.5.03, running on Windows 2003. We are having issues
>with the repository on the new server. (We are also changing the
>authentcaition protocol to sspi, but that is working OK). Our process is as
>1. On existing CVS server, have all of the developers check in their code,
>stop the CVS daemon, and tar up the repository.
>2. Copy the tarball to new server.
>3. Install CVSNT on new server as per best practices in the Wiki.
>4. Unpack the tarball (I used WinRAR).
>5. Set up the CVSNT control panel repository configuration tab to point to
>the extracted repository from the tarball. The Publish Repository, Default
>Repository, and Online items were selected.
>The repository "looks" OK, when running an ls against it for instance.
>However, when the developers check out code and try to compile it, they end
>up with missing file errors. Any thoughts? I know that the docs
>specifically state that the you should not use the built in tools to extract
>anything, but could WinRAR be the culprint?
>cvsnt mailing list
>cvsnt at cvsnt.org
More information about the cvsnt