This page last changed on Sep 15, 2011 by edawson.
Please be aware that Confluence clustered is not available for version 4.0 yet. It will be forthcoming in a minor release of Confluence following the release of version 4.0.

 

Atlassian recommends a network topology similar to the one shown below, to get the best results from a Confluence Clustered deployment.

The number of Confluence nodes in the deployment is adjustable — select the number which suits your own requirements.

The most important aspect is that cluster, database and HTTP (client) traffic are all carried on separate subnets. It is possible, on a sufficiently fast network, to carry cluster and database traffic on the same subnet but we do strongly recommend that HTTP traffic be always confined to a separate subnet on production deployments.

Confluence Clustered does not support clustered communication over WAN, VLAN or VPN. All Confluence Clustered nodes must be on the same local subnet, ideally networked via an ethernet hub or simple switch. The cluster communication network must also support multicast IP networking.

Use this example as a basis for your own network diagram

When you are considering a Confluence Clustered deployment, you should prepare a network diagram like the one on this page. This will facilitate discussion with Atlassian Support and help with your own planning. Please refer to the cluster checklist for more guidance on planning your clustered deployment.


Network topology (text/xml)
Network topology (text/xml)
Network topology (text/xml)
Network topology (text/xml)
Network topology84961.jpg (image/jpeg)
Network topology (text/xml)
Network topology.png (image/png)
Document generated by Confluence on Sep 19, 2011 02:40