ETL Subsystem 5 - Error Event Schema
2 posters
Page 1 of 1
ETL Subsystem 5 - Error Event Schema
First, a quick introduction. I've been playing around with data in various forms for years (please don't ask how many, suffice it to say that I remember when it was cool to camp out at the campus computer center waiting for your turn on the mainframe). I am now afforded the opportunity to create a data warehouse from scratch, albeit somewhat on my own (I know I'm a data geek because I'm completely excited about the project even though there's so many things wrong with the setup!!).
I'm in the process of reading the Kimball library of books, and have a few questions regarding the Error Event Schema discussed as Subsystem 5 of the 34 ETL Subsystems. In the Error Event Detail table, the Table (FK) column implies the creation of a table to hold table names; likewise with the Field Identifier (FK) column.
1. Is this Field Identifier table generic (i.e., one record in the table for a specific field that occurs frequently throughout the source database) or does the Field Identifier table contain one row for each column in each table?
2. Is it implied that the Record Identifier (FK) column ties to the Master Dimension Reference table (when the error is thrown by a dimension row)? If not, is it best practice to have a Master Record Identifier table? Or is it better to have a Master Dimension Reference table and a Master Fact Reference table?
3. This is simply a point of clarification: I assume that the FK references mentioned above are to ETL structures. (I feel that this is obvious, but it never hurts to ask).
I appreciate any input that the members of this forum are willing to provide. Thanks.
Kris
I'm in the process of reading the Kimball library of books, and have a few questions regarding the Error Event Schema discussed as Subsystem 5 of the 34 ETL Subsystems. In the Error Event Detail table, the Table (FK) column implies the creation of a table to hold table names; likewise with the Field Identifier (FK) column.
1. Is this Field Identifier table generic (i.e., one record in the table for a specific field that occurs frequently throughout the source database) or does the Field Identifier table contain one row for each column in each table?
2. Is it implied that the Record Identifier (FK) column ties to the Master Dimension Reference table (when the error is thrown by a dimension row)? If not, is it best practice to have a Master Record Identifier table? Or is it better to have a Master Dimension Reference table and a Master Fact Reference table?
3. This is simply a point of clarification: I assume that the FK references mentioned above are to ETL structures. (I feel that this is obvious, but it never hurts to ask).
I appreciate any input that the members of this forum are willing to provide. Thanks.
Kris
Surf6705- Posts : 1
Join date : 2012-10-04
Error Event Detail Help
Hi,
I'm having the same issues trying to understand how to put the error event schema together. Anyone have thoughts on this?
-Mike
I'm having the same issues trying to understand how to put the error event schema together. Anyone have thoughts on this?
-Mike
happymike- Posts : 1
Join date : 2014-05-02
Similar topics
» Fact event error and Audit Dimension (Fact)
» Start Schema vs Snow flake schema
» odbc error
» Implementing audit dimension error
» ETL Audit and Data Error log Table
» Start Schema vs Snow flake schema
» odbc error
» Implementing audit dimension error
» ETL Audit and Data Error log Table
Page 1 of 1
Permissions in this forum:
You cannot reply to topics in this forum