One of the setup options for Confluence Data Center is to run a Synchrony cluster separately from the Confluence cluster, rather than allowing the Confluence nodes to spawn and manage their own Synchrony processes.

This is documented here. It's pretty clunky, and requires the user to copy the synchrony JAR file out of the Confluence distro and run it themselves.

In a Docker deployment, this entails specifyng a custom entrypoint when the container is created. Again, this is clunky.

It woud be great if Atlassian could provide Docker images that just contained Synchrony, allowing much easier deployment of dedicated Synchrony clusters. These would need to versioned independently from Conflence, but some sort of version compatibility matrix wuld be needed.

As well as making Synchrony deployment easier, it would remove the need to download the full monster docker image just to run Synchrony (which is relatively tiny)

Alternatively, perhaps an another entrypoint script could be included in the Confluence images, to be invoked when you just wanted to run a Synchrony container.