This page last changed on Feb 20, 2007 by david.soul@atlassian.com.
To run Confluence using the SQL Server, you must create a database table and user, download the database drivers and insert the connection details.
- Review Known Issues For SQL Server
- Identify Character Encoding
To identify which character encoding to use, check the encoding used by your application server and Confluence now. All three must use compatible encoding. For example, the default SQL Server encoding of USC-2 is compatible with UTF-8.
- Create Database Table
As an SQL administrator, create a new table. If you set your application server and Confluence to use an encoding incompatible with USC-2, specify that character encoding for the table.
- Create SQL User Account
As an SQL administrator, create a new user account for Confluence. Provide full create, read and write permissions for the table. Note that Confluence must be able to create its own schema.
- Install Database Drivers
SQL Server users are strongly recommended to install the jTDS JDBC drivers. Copy the driver file jtds-<version>.jar into your standalone's common/lib directory. If you are configuring a datasource to connect to your MS SQL server database, you may find this page helpful: http://jtds.sourceforge.net/faq.html
- Start Confluence and visit the home URL (eg http://localhost:8080) to start the Confluence Setup Wizard and select a Custom Install, insert the relevant connection information.
- When prompted for a driver class name in the database setup step enter:
net.sourceforge.jtds.jdbc.Driver
- When prompted for the jdbc url, the format to use is:
Confluence also reports a hibernate database error (connection could not be established) if your data folder in the first step is not configured properly.
Ensure that your data folder specified in [Confluence\confluence\WEB-INF\classes\confluence-init.properties] uses forward slashes (windows), and does NOT point to the /Confluence/confluence/data directory, instead create a data directory in the root:
confluence.home=C:/IRC/Java/Confluence/data
This was not easy to track down.
From our testing, Confluence works using SQL Server 2005, there should be no changes to SQL server 2005 necessary (but do ensure that tcp/ip is enabled, and if communicating web->db with the same box use port 1433, else if communicating web->db to another box, use port 1434)

Posted by dev@irc-australia.com at Dec 19, 2005 23:50
|
We have added a note/warning about the slashes to confluence-init.properties to make this clearer.

Posted by dave@atlassian.com at Dec 21, 2005 23:43
|
Using confluence with sql server 2000/2005 with named instances (eg: INSTANCE01):
http://jtds.sourceforge.net/faq.html#instanceName
http://www.eggheadcafe.com/ng/microsoft.public.sqlserver.jdbcdriver/post21039344.asp
In summary, use the instance port number in place of 1433/1434.
eg.
You can find out which port your named instance is using by going to
Enterprise Manager > Select <database>\<instance name> > Properties > General > Network Configuration > TCP/IP > Properties > (should be 1100+)
(If you're trying to bind to an existing confluence installation, delete your [\Confluence\data\confluence.cfg.xml] file, and restart setup.)

Posted by dev@irc-australia.com at Jan 24, 2006 00:04
|
I was only able to make this work with a SQL Server user; I could never get a Windows user to work. Perhaps this is implied by requiring mixed-mode, but maybe my observation will save someone else some time.

Posted by scottnelsonsmith at Nov 21, 2006 13:11
|
|