Why does the DW Refresh Script Hang/Die?
There are several reasons why the DW refresh script (TeamConnect_Warehouse.sh or TeamConnect_Warehouse_initial.sh) may not work correctly. To identify the cause, you will need to check the DW script logs (in the TCWH_logs directory).
Please note these special cases --
1. If the process dies and simply stops logging without an error, please consider the following:
* Has the script actually done any work? If the logs contain only a few lines, it is probable that the script was unable to find Java, or that the version of Java installed on the machine is wrong (for TCDW 3.3 or 3.4, you must use Java 6, not Java 7).
* Has the server itself shutdown? If so, please run the script again.
* Are you using Linux on a remote server, and did you close your ssh session? If so, the script may have terminated after receiving SIGHUP. Either keep your ssh connection active, or start the script using the nohup utility.
2. If the process dies due to a database or JDBC error, a stack trace including the database error message will appear. Common error messages include those indicating networking problems, as well as deadlocks.
* If the message indicates networking problems, check the network connection to the database, as well as the status of the database listener (if using Oracle). Also, check to make sure that the database supports creating enough sessions for the DW scripts. Note that the DW scripts will use several hundred sessions, so increasing the maximum sessions parameter may be advisable.
* If the message indicates a deadlock, try reducing the commit size on the step that failed. Also, check to see if there are other steps that might be conflicting with this step. If you find one, report a product defect for escalation to the engineering team.