Confluence - distributed via SQL Server Replication

[Tom Morgan]

We are considering using Confluence as a documentation system for our distributed web application. We have 10 remote sites, and use SQL server replication for our web app to keep it in sync.

Has anyone used SQL Server replication to replicate changes from one Confluence installation to another?

[JamieA · Accepted solution]

It certainly won't work. For one thing, the Lucene index will not be notice updates in the database beneath it.

Various caches will not be filled etc.

It's probably suitable for use as the database for a cold standby, but you won't be able to do replication through it.

Why can't the remote sites use the central confluence?

[Tom Morgan]

The instances would be at remote sites in countries with very low bandwidth from our central site.

[AlysonA]

Hi Tom,

Honestly I've never used SQL Server Replication to replicate data between Confluence servers, so I'm not sure how it would work. Assuming it only clone the data and sync it into another server, it should work if using the same Confluence configuration, but that's only a guess.

Alternatively, the Confluence Cluster Edition may be an option for you as it allows you to run multiple copies of Confluence in a cluster using the same database, so that users may access any of the nodes and see the same information. Not sure if that may be of your interest, but thought it was worth to share.

Cheers!

[Tom Morgan]

Sorry, clustering does not apply in my situation. The Confluence nodes in my situation would be widely separated in a WAN. The clustering document is intended for nodes that are on the same switch behind a load balancer for HA.

I guess I'll have to try it out in a test environment using SQL Server replication.

TPM

[Totara DevOps]

did you ever get this working ? I've got a similar replication need over long distance (UK <-> NZ).