Using oracle cdc component in redo log mode to export data into CSV. Created a generic job instead of creating a separate job for each CDC table.
Get the table list (metadata) using Oracle Input component and extract the latest data (delta changes) for those table.
tOracleInput (Read config_table metadata) -> tFlowIterate (Iterate config_table) ==> tOracleCDC (Retrieve delta changes for those table records selected using tOracleInput component) -> tFileOutputDelimeted (Output the delta changes in CSV Output)
table1 schema: (CDC table) id number full_name varchar2(50) mobile number
table2 schema: (CDC table) id number first_name varchar2(30) middle_name varchar2(30) last_name varchar2(30) prev_sur_name varchar2(30) address_ref number
As you can see, my CDC tables schemas are different.
Configuration in tOracleCDC: CDC mode: Log Table using CDC: (String)globalMap.get("Table_name") (This Table_name attribute is passed from tOracleInputComponent) Events to catch: Insert, Update, Delete (Selected all 3) Schema using CDC: Built-In with attribute as "columnName" with "Dynamic" type (This is where the problem occurs. tOracleCDC doesn't support Dynamic type)
Please can someone help. is there any alternative solution to set the schema dynamically so that tOracleCDC component can use it.
There are 2 potential options to solve this 'industrialisation' requirement, namely:-
Add Dynamic schema support to existing tOracleCDC component This will help create the opportunity for a single, generic job design template to build multiple CDC monitoring implementations
Create a JobScript template to read Oracle metadata and automate Job design This will help create a repeatable process for automatically building the various individual Job designs to support CDC monitoring implementations
Both would typically require custom development / delivery work either from Talend Professional Services or Talend Engineering directly in my view.
