In the context of Data Vault architecture, a Link is a core modeling construct that represents a relationship between two or more Hubs. Unlike traditional relational modeling, where relationships are typically embedded using foreign keys, a Link is a standalone physical table that contains only the surrogate keys from related Hub tables—along with metadata such as load date, record source, and a unique identifier.
Links are used to capture business events or transactions that connect entities, such as “Customer places Order” or “Employee assigned to Project.” This design separates relationships from descriptive attributes (which belong in Satellites), enabling better scalability, auditability, and historical tracking.
In ER/Studio, Links can be created visually as part of a Data Vault 2.0 model. The tool supports proper linkage to Hubs, generation of supporting metadata, and alignment with Data Vault naming and structural standards. Links play a vital role in helping organizations design agile, traceable, and scalable enterprise data warehouses.
Key features of Links in a Data Vault model:
Outside of Data Vault, the term “link” is sometimes used informally to describe a relationship or connection between two data entities or tables. In this context, a “link” typically refers to a foreign key relationship—for example, a one-to-many or many-to-many association that defines how rows in one table relate to rows in another.
While “link” is not a formal modeling term in traditional relational design, it may be used conversationally to describe how data is connected. In ER/Studio, these connections are modeled as relationships, which are visual lines drawn between entities, mapping logical associations and enforcing physical constraints such as foreign keys.
To avoid confusion, it’s best to use “Link” when referring to Data Vault models, and “relationship” when describing general entity connections in other modeling contexts.
