How to Build Data Vault Satellites in Coalesce.io

Data Vault Satellite in Coalesce.io

Building Data Vault structures can feel complex when you first begin working with them, especially when implementing them inside a modern transformation platform such as Coalesce.io. In today’s article, we will walk through the full lifecycle of creating Satellites, from source data to stage layers, hash key generation, private and non-private satellites, and finally V1 satellites that support historization and Data Vault 2.0 best practices.

This guide is based on a hands-on demo scenario using supplier data. The goal is not only to show you how to technically create satellites in Coalesce.io but also to explain why these steps matter and how they fit into the larger Data Vault methodology.

Why Satellites Matter in Data Vault Modeling

In Data Vault, satellites store descriptive attributes about business entities. They sit alongside hubs and links, extending these structures with the contextual information that typically changes over time. Because descriptive data can evolve—names, addresses, account balances, and other attributes—satellites allow us to capture the full history of changes while keeping hubs and links stable and lean.

A typical Data Vault satellite includes:

• A hub hash key to tie satellite rows to a business key
• A hash diff to detect attribute changes
• Descriptive attributes such as names or phone numbers
• Load timestamps and record source metadata
• Optional historization fields such as load end date timestamps and an is_current flag

Coalesce.io makes it easier to generate these components through its Data Vault package. The platform handles much of the boilerplate SQL, letting you focus on modeling rather than syntax.

Understanding the Supplier Data Example

In the example used throughout the demo, our source system provides several fields:

• Supplier key (a numerical unique identifier)
• Name
• Address
• Nation key (used in links)
• Phone number
• Account balance
• Comments

To keep things realistic, imagine the name and address values contain personal identifiable information. Because Data Vault supports privacy-aware modeling, we split the satellite into:

• A private satellite for sensitive fields like name and address
• A non-private satellite containing the remaining descriptive data

Separating data in this way supports compliance, access control, and sensitive-data masking—common requirements in real-world deployments.

Step 1: Building the Stage Layer

Before building satellites, we must create a stage table in Coalesce. The stage prepares the data for Data Vault modeling by generating the hash keys and hash diffs we will need later.

Inside Coalesce, we right-click the supplier source node and select Add Node → Stage. The purpose of this stage is to normalize the structure and add the necessary technical metadata.

What we generate in the stage:

• Hub Hash Key (HK_Supplier): created from the Supplier Key
• Hash Diff for Private Satellite: created from Name and Address
• Hash Diff for Non-Private Satellite: created from Phone, Account Balance, and Comment
• Load Date Timestamp: represents when the data was loaded
• Record Source: tracks where the data came from

After saving and creating the stage node, Coalesce generates a view containing all the original source data plus the new hash fields and technical metadata. This prepares us to build the satellites cleanly and consistently.

Step 2: Creating the Private Satellite (V0)

Next, we right-click the stage and add a new node, selecting the V0 Satellite from the Data Vault package. This satellite will contain only the sensitive columns.

Private satellite includes:

• Hub hash key
• Private hash diff (Name + Address)
• Name
• Address
• Load Date Timestamp
• Record Source

We remove the non-private columns and keep only what belongs to the private satellite. After configuring the node properties and Data Vault options, we create and run the satellite.

The private satellite now holds the sensitive data, along with the hash diff that allows Coalesce to detect changes over time.

Step 3: Creating the Non-Private Satellite (V0)

We repeat the process for the second satellite, this time focusing on non-sensitive attributes such as phone number, account balance, and comments.

Non-private satellite includes:

• Hub hash key
• Non-private hash diff
• Phone
• Account Balance
• Comments
• Load Date Timestamp
• Record Source

Once configured and created, this satellite is also loaded and ready for historization.

Step 4: Creating V1 Satellites for Historization

V0 satellites store the raw history and all versions of descriptive data. To simplify querying, Coalesce supports generating a V1 Satellite, which is a view layered on top of the V0 Satellite.

The V1 Satellite adds:

• Load End Date Timestamp (LEDT): identifies until when a record was valid
• Is_Current Flag: marks the latest version of each record

These fields allow analysts to easily filter for the “current” state of descriptive attributes or build temporal reports when needed.

Creating a V1 Satellite is straightforward: right-click the V0 Satellite, add a node, and select the V1 Satellite template. Coalesce automatically generates all required SQL and fields.

After running the V1 satellites for both private and non-private data, you now have a complete Data Vault satellite layer: historized, query-friendly, and fully compliant with Data Vault 2.0 standards.

Why This Matters in Real Data Vault Implementations

This workflow demonstrates the core principles of Data Vault modeling:

• Separation of concerns: Private data stays protected.
• Change detection through hash diffs: Efficiently track what changed.
• Historization: V1 satellites provide easy access to current and historical states.
• Consistent metadata: Load timestamps and record sources support auditability.

Coalesce automates much of the repetitive work required in Data Vault, enabling teams to produce reliable, scalable models faster and with less manual SQL.

Final Thoughts

Creating Data Vault satellites in Coalesce.io becomes a smooth process once you understand the core concepts: preparing the stage, generating hash keys and hash diffs, organizing attributes into private and non-private structures, and finally adding V1 satellites for historization. With Coalesce’s Data Vault package, this modeling pattern becomes not only repeatable but highly efficient.

If you’re new to Data Vault or want to deepen your understanding of how its components work together, consider reviewing a Data Vault handbook or exploring more Coalesce transformation sessions. Each layer builds on the previous one, ultimately forming a flexible, auditable, and future-proof data warehouse architecture.

Watch the Video

Meet the Speaker