Data Express – How to restart only the failed data stores in the Distributed Extension

Problem

 When the data extraction/masking process is interrupted for some reasons  (errors, abend …) the retrieval of data will resume from the beginning at a subsequent restart.  This leads to a waste of time due to re-running successfully processed data stores.

Resolution

 In order to process only the failed data stores during the restart after extraction interruption the following configuration is needed before running the command dxestart:      

1. A new parameter must be added to the dxeconfig.cfg file called:

             Execute Aborted Tables:N

        The parameter "N" is the default value. 

           -     When "N", the process works as before, but logs the successfully passed data stores.

           -     When "Y", the program processes only the failed data stores from the previous run.

2. In order to work, a database update must be done. A new column named “PASSED” must be added  to the runtime table HSENVELB before using the Distributed Extension. The following query must be executed towards the runtime knowledge base (DEKB):

 -- For XDB, DB2

SET LOCATION TO TUTORIAL;

ALTER TABLE DEKB.HSENVELB ADD COLUMN PASSED  SMALLINT NOT NULL WITH DEFAULT;

COMMIT;

-- For MSSQL Server

ALTER TABLE DEKB.HSENVELB ADD PASSED SMALLINT NOT NULL DEFAULT(0); 

GO

-- For Oracle DB:

ALTER TABLE DEKB.HSENVELB ADD PASSED NUMBER DEFAULT 0 NOT NULL;

COMMIT; 

Note: The solution is valid for ODBC and Oracle Extension.