Disaster Recovery for Confluence (2017 Edition)

Question

This post is inspired by a 2013 post regarding the same topic.  I also needed a little more information.  So our configuration is:

1. Two datacenters.  Don't need hot-hot configuration, this is purely for if the primary site goes down.

2. We already backup our DB and Attachments every night to the DR site.

3. Failover is understood to be a manual process (script-assisted is desirable!)

4. We do have the option of replicating our PostgreSQL DB up to the DR site (async, read-only at DR), but it's not required, especially if it adds significant complexity.

Questions:

1. How do I setup the "DR" installation?  I've got the installer running and trying to sort out of I just setup the "Demo" DB and then reconfigure after that?  It's a chicken/egg thing.

2. Can anyone describe the steps that I'd need to perform to get the "backed up" stuff running in "production" at our DR site?

Thanks for reading!

Nic Brough -Adaptavist- · Answer

Igor points you to migration docs that point out all the other steps beyond the database (attachments, homer directory, re-indexing etc).  You need to go through all of those.

unknown · Answer

Thanks for your reply!  I'm really interested in scripting this.  So I can easily script the DB with the following:

* Drop "old" DB
* Create "new" DB
* Import latest DB backup into "new" DB

But I'm sort of lost on how I can "reload" Confluence to recognize the new data, or does it need any particular tweaks?  Maybe just a stop/start?

Igor M_ · Answer

Hello,

You can follow a combination of Migrating to Another Database and Site Backup and Restore instructions, to understand what exactly is required to backup and restore your Confluence instance on your DR site.

In particular, Step 6 of Migrating to Another Database covers your specific questions

Hope this is what you were looking for.

Regards,

Igor