This page last changed on Feb 17, 2006 by mryall.
Old documentation

This documentation is for old versions prior to Confluence 2.0. For the latest documentation see the Confluence Installation Guide.

You will need

  1. A Java Application Server, as described in the Requirements
  2. To know how to install and configure your application server. If you do not, we suggest you download Confluence Standalone instead.
  3. (Optional) A database, as described in the Requirements

Installing the WAR version is a piece of cake:

  1. Unzip the Confluence WAR zip that you have downloaded
  2. Find the confluence folder in this unpacked distribution and copy this over to the webapps directory of your application server (assuming you are using tomcat and resin).
  3. Edit the confluence-init.properties file in the webapps\confluence\WEB-INF\classes directory and set the confluence.home property to a directory of your choice. This directory is where Confluence will store it's configuration information, indexes and attachments.
  4. Restart your webserver if necessary.
  5. Point your browser at www.mydomain.com/confluence where mydomain is your website domain. You should see the Confluence Setup Wizard!

Unix without X11
On some Java installations (i.e. Unix without X11), you need to pass an additional parameter to Java when you are starting up to tell it how to use its image-manipulation libraries.

To get thumbnail support working in this configuration, you need to start java with the argument '-Djava.awt.headless=true'. If you are using tomcat (or the Confluence standalone edition), try setting the environment variable JAVA_OPTS before starting the server. Assuming you're using the bourne or bash shell:

export JAVA_OPTS=-Djava.awt.headless=true

Alternatives to confluence-init.properties

This is an advanced note for experienced Java application server administrators. If you do not know how to set system properties or edit web.xml in your application server, we recommend you stick with editing confluence-init.properties as described above.

If you wish to deploy the WAR directly into an application server (or several application servers), you may want to skip the step of exploding the WAR file.

Confluence needs one piece of information in order to get running: the location of the confluence.home directory in which it should save its configuration files and other data. It obtains this information by looking at the following three places (in order):

  1. System properties (i.e. start the application with -Dconfluence.home=/path/to/your/confluence/home)
  2. confluence-init.properties (i.e. follow the procedure above)
  3. Servlet context (i.e. customise the web.xml deployment descriptor to include a <context-param> called confluence.home)

The first not-null value Confluence finds will be used as the confluence home.

So, if your application server supports the easy editing of system properties or servlet context parameters when deploying an application, you may want to use one of these methods instead of editing confluence-init.properties

Document generated by Confluence on Oct 10, 2007 18:35