Repoint Confluence to a different SQL server

[Kimberly McCarty]

Hello,

I need to repoint Confluence to a different SQL server. Are there instructions on how to do this? Anything I need to look out for?

Thank you,

Kimberly

[Kimberly McCarty]

Hi Nic,

Great question. I didn't install Confluence ordinally and the guy that did is no longer on our tema. I am looking for guidance on what I need to do to repoint Confluence from one SQL server to another. So I don't currently have a preferred route, I really don't even know where to start. Are there some instructions for doing so?

Thank you,

Kimberly

[Nic Brough -Adaptavist-]

Ok, makes sense.

Confluence is a system that uses a database (as a data store, not a real database, but that's not too important here).  Take the database away, you lose everything you've put into Confluence.

So.  If you want to change the database that Confluence is using for its datastore, you need to think about how to move the data from the old one to the new.

It's not as simple as a "repoint" - repointing just means telling a system to use a new database, but if you do that with any system, you're telling it to stop using the data you have in the old database.  Which means you are starting afresh and you lose all the old stuff.

There are broadly two ways to move (copy really) data from one place to another so that Confluence keeps everything:

• Take a backup of Confluence via its XML export routine, run up a new, empty Confluence against the new database, then import the backup.  
  • Pro: database independent - you can use it to jump between any database types (Oracle, Postgres, MySQL and MS-SQL are the supported ones, but it also works with other unsupported ones that you can make Confluence work with!)
  • Cons: slow, fails for larger databases, prone to errors and needing minor tweaks like odd character sets and unusual characters, and may need quite intricate reconfiguration of the database connections
• Take a backup/dump of the old database using standard database tools, restore it to the target database, repoint confluence and restart it.  Pretty much a like-for-like lift and shift
  • Pro: fast, reliable, minimal reconfiguration
  • Con: needs old and new databases to be 100% compatible

So, the next question is how different the old and new SQL servers are.  If they're similar enough, I would very strongly recommend you get the DBAs to do the work - export, copy and import the database wholesale.

[Kimberly McCarty]

This is fantastic info! I appreciate how simply you stated this. I will be reviewing this with my IT group and let you know if this brings up any questions.

-Kimberly

[Kimberly McCarty]

Hello,

Our current confluence db is on a Windows 2008 with 2012 SQL server SP3.

We need to move it.

We need to build an environment to get this moved too, in preparing for that I have a question around using this process to move our confluence db: (see below)

• Take a backup/dump of the old database using standard database tools, restore it to the target database, repoint confluence and restart it.  Pretty much a like-for-like lift and shift
  • Pro: fast, reliable, minimal reconfiguration
  • Con: needs old and new databases to be 100% compatible

My question: is it ok if we build a Windows 2019 with a 2012 SQL server SP4?

Will this allow us to be careful on the CON addressed above: needs old and new databases to be 100% compatible because the SP's would be different.

Should we install SP4 on the current hardware where the confluence db resides before we make this move?

Thank you,

Kimberly

[Nic Brough -Adaptavist-]

Will the database on the new SQL server be a clone of the existing Confluence database?

If so, then it's just a case of stop confluence, edit the config file in the home directory that has the connection in it, restart it.  I'd always recommend a re-index too, but it should not strictly be needed.

If it's not, then we'll need to know what you are trying to do in more detail!

[Kimberly McCarty]

Hi Nic,

The new SQL server will not be a clone of the current one. We currently have confluence pointed to an older SQL server that is going to be brought down, it houses our old electronic health record version and that environment will be spun down. 

Therefore the need to point Confluence to a new SQL server and not one where our electronic health record resides.

Thank you for asking,

Kimberly

[Nic Brough -Adaptavist-]

Ok, so have you got a plan for exporting and migrating the data from the old database to the new one?  Or even a preferred route for doing it?