Understanding hdbsetup configuration and odbc.ini, tnsnames.ora, service names
search cancel

Understanding hdbsetup configuration and odbc.ini, tnsnames.ora, service names


Article ID: 101635


Updated On:


CA Harvest Software Change Manager - OpenMake Meister


We are trying to work through service name / odbc.ini / tnsnames.ora / server name issues. With this new installation of Harvest to a production server we are getting errors starting the server related to the 'harvest' service name not found.


Release: CA Harvest SCM - All versions and platforms
Component: HINFST


To configure all the settings necessary to allow your HServer to connect to the Harvest database follow these steps:

  1. If your Harvest server is Linux or Unix, you need to be logged in as the user that owns the SCM Software folder and processes.

  2. Check your environment variables.  You should have:
    • ORACLE_HOME set to the home folder for your Oracle client.  If SCM is 32-bit, then 32-bit client.  If it’s 64-bit, then 64-bit Oracle client
    • ORACLE_SID set to the SID for your Harvest database
    • PATH should contain %ORACLE_HOME\bin for Windows, $ORACLE_HOME/bin for Linux/Unix
    • For Linux/Unix, LD_LIBRARY_PATH should contain $ORACLE_HOME/lib

  3. Check the $ORACLE_HOME/network/admin folder for a “tnsnames.ora” file.  Make sure it’s world-readable and contains a paragraph with the information that points to the hostname, port number, and service name of your Harvest database in Oracle.
Here is an example:

Remember the TNS Service name that we will use to configure Harvest is at the beginning of the paragraph.  If you don’t have this file or don’t have a paragraph that points to your Harvest database, the easiest thing to do is locate the tnsnames file on the Oracle server machine and copy it from there.
  1. Try to log in to the Harvest database with SQL Plus.  If SQL Plus can connect, then Harvest will be able to connect as well.  You’ll need to know the Oracle login userid and password to the Harvest database schema owner’s account and the TNS Service name from your tnsnames.ora file.  Use this command syntax:
sqlplus userid@servicename/password
-    userid = Oracle login userid for the Harvest database schema owner
-    servicename = the TNS Service name you found in step 3 above
-    password = Oracle login password for the Harvest database schema owner
A successful connection will look similar to this:

If Linux/Unix you should try this sqlplus command while logged in as the “cascm” user.  If you do not get a successful connection, work to resolve any connection problems before continuing to the next step.
  1. Execute HDBSetup and use the CO option to configure the “ODBC Data Source” definition for your Harvest database.  You will be asked for a Data Source Name (or to select an existing one if you want to re-use a name configured before) and the Service Name (your TNS service name).  The settings should look similar to this:

To confirm the settings, hit “Enter”.

Next, you will be asked if you want to test the connection.  Type “y” and hit “Enter” to agree.  You will be asked for your DBMS Administrator username [system] and the password for the DBMS Administrator.  After typing the password and hitting “Enter,” the connection test will happen, and you will either see “Connection test succeeded” or any error message that was encountered.  Any problems found at this step must be resolved, and a successful connection test must be completed before the HServer can use this information to connect to the Harvest database.
  1. Before exiting HDBSetup, it’s a good idea to execute the “EP” option to encrypt the Oracle login userid and password for the Harvest database schema owner.
  1. After completing HDBSetup, you should check the HServer.arg file in your $CA_SCM_HOME folder to ensure that the correct Data Source Name is listed there:

  1. After all steps have been completed, you can try to start the broker and check the HServer log files to confirm whether the HServer can connect to your Harvest database.