Data Vault Ensemble Modeling

Managing time in a Data Vault: How to design, insert and update (bi-)temporal data

Dirk Lerner, TEDAMOH

“Over time, things change –  things like customers, products, accounts, and so forth. But most of the data we keep about things describes what they are like currently, not what they used to be like. When things change, we update the data that describes them so that the description remains current. But all these things have a history, and many of them have a future as well, and often data about their past or about their future is also important.” – Tom Johnston

Today, most data warehouses already store the history of the data. With Data Vault, this is an integral part of the method. But what about events that took place at a different time than what the data warehouse represents to us? Or data that will be valid in the future? For example, generally planned prices for products and goods in the future or special prices for discount battles around the “Black Friday”.

The speaker will focus in this session on the method and techniques for storing bitemporal data into a Data Vault. He will show bitemporal basics for a better understanding of loading data as well as the concepts to develop SQL Queries to insert and update temporal data within Data Vault.

What attendees will learn in this session:

  • Basic bitemporal concepts
  • Examples of data modeling a bitemporal Data Vault
  • Load bitemporal data into Data Vault
  • Examples of SQL Queries to insert and update temporal data

Dirk Lerner is an experienced independent consultant and owner of TEDAMOH. For 15 years he has led BI projects and is considered a global expert on BI architectures and data modeling. As a full scale data architect, Dirk advocates flexible, lean and easily extendable data warehouse principles and practices.

As a pioneer for Data Vault and FCO-IM in Germany, he wrote various publications, is an international speaker at conferences and author of the blog https://DataVaultModeling.de.