Recently I wrote about how Octopus Deploy originally used a Microsoft SQL Server database to store deployment information, and how I switched to RavenDB for the 1.0 release.
Many customers had data in those 0.9.* releases of Octopus, and weren't able to upgrade to 1.0 because there was no way to migrate that data over. I've been working on a tool to import this data from SQL Server into RavenDB, and it's now included in the latest Octopus Deploy release.
In this post I'll walk you through the migration process. Obviously this post is only of interest to customers who have a 0.9.* database.
Before you begin the migration from 0.9 to 1.0:
- Make a backup of the SQL Server database
- Download and install the latest Octopus Deploy server. Work through the setup wizard as per normal (see the installation guide for details).
- Make sure the Octopus Windows Service is running
By this point, you should have a working 1.0 Octopus installation, but without any of your old data. Now, from the start menu, look for the Octopus Beta Migration Tool shortcut:
The wizard will ask you to enter the connection string to your SQL Server database (it should detect it automatically from your previous installation):
Once you've tested that the connection string works, you can start the import:
Once the import completes you can browse to your Octopus web portal as per normal. You should find that all of the previous data has been imported.
There is however one extra step that you'll need to do. The security model for Octopus and Tentacle has changed (we now exchange thumbprints), so you'll have to run the new Tentacle installer on all of your machines, and edit each of the imported machines in the Octopus web portal to register the machine thumbprint.
With that out of the way, you'll now have a happy 1.0 installation!
If you have any problems during the migration, please let us know on our support forum.