Db2 for z/OS operational considerations

Limitations

Each RDRS pipeline requires a separate Striim application. Each Striim application can read from only a single RDRS pipeline.
Tables without primary keys cannot be read.
Tables with names beginning with DB2_GENERATED_ cannot be read.
Due to a limitation in RDRS, Inserts and updates larger than 1 MB cannot be captured.
Primary key updates are not supported if the table contains an LOB or XML column. The application may halt or terminate, and data may be corrupted. However, this limitation does not apply when using a PartialRecordPolicy, which skips LOB/XML fields during processing.
Dropping a Striim application containing a Kafka-persisted stream being written to by RDRS will delete the associated Kafka topic. Before dropping such an application, stop the RDRS pipeline and anything else that is writing to the associated Kafka topic.

Start a pipeline:

setupRDRS_DB2zOS.sh -a start -f <path to configuration file>

Stop a pipeline:

setupRDRS_DB2zOS.sh -a stop -f <path to configuration file>

Drop a pipeline (stop the pipeline first):

setupRDRS_DB2zOS.sh -a drop -r <RDRS repository password> -f <path to configuration file>

Get a list of all pipelines and their statuses:

setupRDRS_DB2zOS.sh -a get -r <RDRS repository password> -f <path to configuration file>

Schema evolution must be handled manually. See Handling planned DDL changes.

The log file is in the RDRS directory at <RDRS directory>/log/tcAgent.log.

To be able to restore RDRS in the event of a failure, back up the RDRS directory and the PostgreSQL or Oracle repository.

Scenario	Solution
RDRS pipeline and Striim application are running but no events are appearing in the output of the external source.	Check the RDRS dashboard and the RDRS Agent log.
RDRS is having connection failures.	Check the RDRS Agent log.
RDRS pipeline does not work after stop and restart.	Rename the RDRS process and delete the files in the RDRS/tmp directory for the old process.