On Mon, 23 May 2022 11:03:32 +0000, "Morin, Michael" <michael.mo...@maine.gov> wrote:
>>On Mon, May 23, 2022 at 6:16 AM Morin, Michael <michael.mo...@maine.gov> >>wrote: >>> >>> In my case, I have no choice in the matter of what server and OS to move >>> to. We dont use VisualSVN. >>>We just use Tortoise with Subversion. When we last >did a migration from one >>>Windows server to another, >>>it was a tedious process of contacting the users of the repository, asking >>>them if the repo should be >>>migrated or archived, setting a date for the migration, telling the users to >>>make any outstanding commits, >>>locking down the repository, migrating the >repository, informing users of >>>the new URL of the repository, >>>and asking the users to confirm their ability to access the repo and >>>verifying everything >works as expected. >>>This was done for each repository. This process took us about 4 ½ months. >>>Id like to avoid taking that >>>long this time around. >> >>*Ouch*. Been there, done that. I'm assuming that each repo is, indeed, an >>entirely distinct Subversion >>repo rather than all being distributed under one >master repo? That makes >>migration much, much safer. >> >Yes, each repository is a distinct Subversion repository. We have about 300 >Subversion users located in >different departments and 171 repositories. > Just to clarify for my understanding: Are you saying that the users are accessing the repositories via file sharing in Windows using Tortoise-Svn? I.e. do they point the client (Tortoise-SVN) to a drive letter location where they operate on the repository? In that case you do not even have a *server* running so migrating the *server* is a non-issue. And the whole setup is pretty vulnerable to rogue users with write permissions on the repository side... PS: Please do not top-post, it makes it so much harder to follow the discussion. DS -- Bo Berglund Developer in Sweden
