Installation Notes ------------------ On RT 3.8.2 and later, RT deployment for Oracle databases is very straightforward. You don't need to configure Oracle beforehand. During installation a user is created and all RT's objects are created in his schema. The user is created with the following parameters: CREATE USER IDENTIFIED BY DEFAULT TABLESPACE USERS TEMPORARY TABLESPACE TEMP QUOTA UNLIMITED ON USERS And the user is also granted 'CONNECT' and 'RESOURCE'. It's up to you to do decide how to manage users, change quotas, table spaces, etc. RT has an option $DatabaseName which is used to define the SID of the Oracle database. You don't have to set up the TWO_TASK environment variable or any helper files for establishing connections. Example: ./configure \ --with-db-type=Oracle \ --with-db-database=XE \ --with-db-host=192.168.0.1 \ --with-db-dba=system \ --with-db-rt-user=rtdb1 \ --with-db-rt-pass=rtdb1secret \ ... other configure options ... That should be enough to get you started on Oracle, but to complete installation you must follow the general instructions in the README file. Client Drivers -------------- When installing Oracle client drivers, run the same version as the database you are connecting to. If you are running RT on the same server as the Oracle database, this should happen automatically as the client drivers will be installed with the Oracle database software. If the Oracle database runs in a separate server and you are installing client libraries from the Oracle Instance Client, find and install the version that corresponds with the database you are connecting to. You will also need to install the DBD::Oracle perl module to allow RT to connect to the Oracle database. Some recent versions of DBD::Oracle have had problems with RT and should be avoided. DBD::Oracle version 1.74 is known to work well with RT. The next known version to work well is version 1.90. Versions in between those had an issue on disconnect that would cause a segfault. Updating Statistics ------------------- As with all databases it is important to analyze the schema and get current statistics after every significant dataset change. Oracle's cost-based optimizer can provide particularly bad performance when the schema statistics are inaccurate. To analyze the schema of the user called RT, execute the following from within sqlplus: execute dbms_utility.analyze_schema( 'RT', 'estimate');