Details about Informatica MDM metadata or infrastructure tables

You might have come across the term metadata tables in infrastructure tables during your Informatica MDM project implementation. What are these infrastructure tables? What is the significance of these tables? How can we access it and use it? Are you facing these questions and would like to know more about these? If so, then you reached the right place. In this article, we will explore the infrastructure tables get generated during the Base Object, Stage and Landing tables configuration. So let's start.

Introduction:

The MDM infrastructure tables are the core part of Informatica MDM. These tables are created, whenever we configure the basic tables such as Base Object (BO), Stage and Landing tables along with their properties such as Raw Retention, Delta detection on the Stage table or match and merge setting on the Base Object table.

What are the MDM infrastructure tables?

Assume that we create Landing table, Stage table, and Base Object table as C_L_PARTY, C_S_SALES_PARTY, and C_B_PARTY respectively. Also assume that we configure raw retention, delta detection, tokenization, match and merge rule as well. After doing all these configurations at table level the supporting tables are created.

1. Tables at Landing table level: There is no infrastructure table created at the landing table level
2. Tables at Staging table level: The tables created at the Staging table level are 

• C_S_SALES_PARTY_RAW
• C_S_SALES_PARTY_PRL
• C_S_SALES_PARTY_OPL
• C_S_SALES_PARTY_REJ

Each of these tables has its own importance and are used during MDM batch job execution.

           3. Tables at Base Object table level: There are 14 supporting infrastructure tables are created.

• C_B_PARTY_MTCH
• C_B_PARTY_HIST
• C_B_PARTY_XREF
• C_B_PARTY_HXRF
• C_B_PARTY_DRTY
• C_B_PARTY_CTL
• C_B_PARTY_HMRG
• C_B_PARTY_HCTL
• C_B_PARTY_EMI
• C_B_PARTY_EMO
• C_B_PARTY_VXR
• C_B_PARTY_HVXR
• C_B_PARTY_VCT
• C_B_PARTY_STRP  

What is the need of the MDM infrastructure tables?

The Informatica MDM implementation involves various process such as Stage, Load, Tokenization, Match, Merge, etc. During each process, the data is transferred from the source table to the target table. During this transfer data is manipulated with the help of supporting table. e.g. During the stage job, the data is transferred from the landing tables to Staging tables. During this transfer, the landing data is maintained in _PRL, _RAW tables. The _PRL table data is used to determine delta of the source record which is subprocess during stage job. 

Similar cases are involved during load job as well tokenization job. These infrastructure tables play a vital role in Informatica MDM implementation.

Relationship between Landing table and the Base Object table

• The load job loads data from the Stage table to a Base Object
• There is still the dependency on landing table data to handle the rejection. 
• The batch job will try to pull the source table record for inserting into the reject table.
• If the landing table is missing the corresponding records, then the reject table will have an entry to state that the source table entry not found. 
• If the landing table is huge and performance issues occur in the load job during the rejection handling, then assess the environment to add a custom index on the landing table.

Is it ok to modify the existing structure of the MDM infrastructure tables?

Informatica strongly recommends that do not modify the structure of these tables as these designed for internal processing purpose only. If you modify these tables, metadata validation may complain error.

The video below provides detailed information about the MDM infrastructure tables -