ETL1 High Level Architecture for FSR Framework
1
ETL1 High Level Architecture for FSR Framework The purpose of this document is to portray the architecture of the FSR framework in a high level view.
GFIS Tables (ETL DB)
Reference Data Engine
Pre-processing Engine
Validation Engine
T_GW_STP_ STARTUP (GII DB)
Derivation Engine
Calculation Engine
ETL1 FSR Framework Balance Control and Reconciliation Engine
Remapping and Reclassification Engine
Error and Exception Processing Engine
Target Load Ready Engine
Target System
ETL1 High Level Architecture for FSR Framework
2
Dig: ETL1 High Level Architecture Diagram for FSR Framework
The FSR framework process begins once the run for the individual feed is completed and the data is available in the corresponding GFIS table in the ETL environment. The value ‘Y’ in DATA_PRESENCE_FLAG column in the T_GW_STP_STARTUP table indicates that the data is available in the corresponding GFIS table and it is ready for further processing. This source data from GFIS table is taken and it is directly loaded in the T_GW_TMP_PROCESSED_DATA table in the GII environment with a STAGE_NO value ‘0’. Every row that is populated in this T_GW_TMP_PROCESSED_DATA table will acquire a unique RECORD_ID. The records with STAGE_NO ‘0’ are taken for a particular PRC_PROCESS_ID and FEED_ID and are allowed to traverse through the various engines present in the framework. There are 9 (from the diagram above) processing engines in which each engine comprises of many components. The input record is propelled through each component and the respective output record is also stored in the same T_GW_STP_STARTUP table, with the STAGE_NO increased by 1 for every stage. In this process the PRC_PROCESS_ID, RECORD_ID and FEED_ID remains the same for a particular record. On completion of the first 8 (from the diagram above) processing engines the ‘Target Load Entry’ engine, which is the final engine in FSR starts processing. This Target Load Entry engine selects the records with the maximum STAGE_NO for a particular PRC_PROCESS_ID and FEED_ID and makes it available for generating the target file. The target file is then generated for the respective target systems depending on the requisite file format.