Data vault modeling

Datavault or data vault modeling is a <a href="/facts/Database/VYcpkevb">database</a> modeling method that is designed to provide long-term historical storage of <a href="/facts/Data/AUBFN97o">data</a> coming in from multiple operational systems. It is also a method of looking at historical data that deals with issues such as auditing, tracing of data, loading speed and <a href="/facts/Resilience_(organizational)/Qxz9zoEg">resilience</a> to change as well as emphasizing the need to <a href="/facts/Audit_trail/s3nQzrF4">trace</a> where all the data in the database <a href="/facts/Data_lineage/cKtPUrde">came from</a>. This means that every <a href="/facts/Row_(database)/PIpv7c9e">row</a> in a data vault must be accompanied by record source and load date attributes, enabling an auditor to trace values back to the source. The concept was published in 2000 by <a href="/facts/Dan_Linstedt/1wmDPCJj">Dan Linstedt</a>.
Data vault modeling makes no distinction between good and bad data ("bad" meaning not conforming to business rules). This is summarized in the statement that a data vault stores "<a href="/facts/Single_source_of_truth/uXqMNxcz">a single version of the facts</a>" (also expressed by <a href="/facts/Dan_Linstedt/1wmDPCJj">Dan Linstedt</a> as "all the data, all of the time") as opposed to the practice in other data warehouse methods of storing "a <a href="/facts/Single_version_of_the_truth/R54wIvS6">single version of the truth</a>" where data that does not conform to the definitions is removed or "cleansed". A data vault enterprise data warehouse provides both; a single version of facts and a single source of truth.
The modeling method is designed to be resilient to change in the business environment where the data being stored is coming from, by explicitly separating <a href="/facts/Data_structure/yNsNvL1I">structural information</a> from descriptive <a href="/facts/Attribute_(computing)/EpQSmn02">attributes</a>. Data vault is designed to enable <a href="/facts/Parallel_computing/nQzcpzQt">parallel</a> loading as much as possible, so that very large implementations can scale out without the need for major redesign.
Unlike the <a href="/facts/Star_schema/hXsGMSKf">star schema</a> (<a href="/facts/Dimensional_modelling/fNywevVS">dimensional modelling</a>) and the classical <a href="/facts/Relational_model/VFf4dDIU">relational model</a> (3NF), data vault and <a href="/facts/Anchor_modeling/kpUlSkbS">anchor modeling</a> are well-suited for capturing changes that occur when a source system is changed or added, but are considered advanced techniques which require experienced <a href="/facts/Data_architect/1LBnp61L">data architects</a>. Both data vaults and anchor models are <a href="/facts/Entity_(computer_science)/4upU2Ukg">entity-based</a> models, but anchor models have a more normalized approach.

Data vault modeling open-in-new

Data vault modeling