Can I run multiple instances of Confluence and connect them to a central database?
Platform notice: Server and Data Center only. This article only applies to Atlassian products on the Server and Data Center platforms.
Support for Server* products ended on February 15th 2024. If you are running a Server product, you can visit the Atlassian Server end of support announcement to review your migration options.
*Except Fisheye and Crucible
Confluence can be clustered.
If running Confluence on a single node, you can only have one instance of Confluence connecting to a single database.
There are a couple of reasons for this, but it all comes down to the fact that the Confluence application maintains a lot of state (e.g., caches, search indexes) outside of the database, and multiple front-ends will quickly see that state get out of sync, with disastrous effect.
Because of this, Confluence periodically checks to make sure that it's the only application accessing its database. If it finds a conflict (called a "cluster panic"), it will shut down rather than risk corrupting your data.