Migration & Upgrade of Confluence

[Marcel Stock]

Hello everyone,

i received the task to migrate our current Confluence Infrastructure to a different server with a new OS (Windows Server 2008 to Windows Server 2016) and do an upgrade from 5.10.7 to the newest version of Confluence as well.

Since i am not the person that initially installed Confluence i need some tips what would be the best way to solve this task. 

My initial plan was to clone our current Confluence database (Microsoft SQL) and install the same Confluence version we are currently using in production (5.10.7) on the new Server/OS. From then on i wanted to edit the server.xml so it points to the new cloned database and install the newest version of Confluence. Are there any additional steps in between that i should do?

Would i have to re-download all Confluence Apps or is there anything else that wouldn't get migrated?

I will be using this guide for this task but i wanted to get some feedback by experienced community members so nothing goes wrong :

https://confluence.atlassian.com/doc/upgrading-confluence-manually-255363437.html

Best Regards

[Igor M.]

Hello,

You can follow Migrating Confluence Between Servers guide to migrate instances between servers.

Your action plan seems to be on track, but you can simply copy over install and home directories. Most important from home being Attachments directory. To change which DB new instance is pointing, you need to modify <confluence.home>/confluence.cfg.xml file. At the same time, if the path of new home is different, you need adjust parameters for that too (see above guide). Your plug-ins data can be preserved if you copy whole home directory, the only folders that do not have to be copied are the following, as these will be recreated upon first boot:
<confluence-home>/index/
<confluence-home>/journal/
<confluence-home>/bundled-plugins
<confluence-home>/plugins-cache
<confluence-home>/plugins-osgi-cache
<confluence-home>/plugins-temp
<confluence-home>/bundled-plugins_language

Always make sure you have the right permissions set on your target server for the account that will be running Confluence.

[Marcel Stock]

Hello Igor,

thanks for the quick response!

I have a specific question regarding the confluence.cfg.xml file. In our current productive confluence.cfg.xml file the path to the instance and DB is the following :

name="hibernate.connection.url">jdbc:jtds:sqlserver://xxx.domain.domain/Confluence;instance=Confluence</property>

The path is obviously pointing to the SQL Server Instance where the Confluence database is located in (xxx.domain.domain/confluence). What im trying to understand is if the "instance=Confluence" after the path to the SQL Server Instance is the name of the actual database INSIDE the SQL Server Instance?

I'm kind of confused since the person that installed Confluence named the SQL Server Instance "Confluence" and the actual database inside the Instance "Confluence" as well. I was planning to create a clone of the confluence database inside the already existing SQL Server Instance and use the clone for the upgrade.

To sum up my question : 

sqlserver://xxx.domain.domain/Confluence   = Path to the SQL Server Instance where the DB is located in?

instance=Confluence   = Name of the DB? Or name of the SQL Server Instance where the DB is located in?

Hope my question is not too confusing!

Best Regards

[Marcel Stock]

..duplicate

[Marcel Stock]

Hello Igor,

did you receive my last reply? For some reason it keeps getting deleted after i refresh the page.

[Igor M.]

@Marcel Stock sorry for the late reply

You can always @ someone to get their attention :)

name stands of DB name, there is no need to specify the instance as it is assumed there is only one instance running on the server and you are specifying connection port to it. Have a read in How to find Confluence's database connection parameters guide for the exact definition of params.

Regards,

Igor

[Marcel Stock]

.. duplicate