Tables
A table is the entity that actually holds your master records: one per concrete subject, such as Customer, Product, or Branch. Its shape is defined by attributes; its contents are records.
What is a Table?
A Table is the concrete entity that holds records. It belongs to exactly one data model (and therefore one enterprise), and its schema is defined by its attributes. Typical examples: Customer Master, Product SKU, Branch, Cost Center.
The Tables screen
While each data model shows its own tables, the Tables screen is a global directory of every table you can access, useful when you want to see all your entities side by side, regardless of which model owns them.
Two cascading filters at the top narrow the view: Enterprise restricts to one tenant, and Data Model restricts to one domain within it. The + Create Table button appears only when you can edit the selected scope.
| Column | What it shows |
|---|---|
| Table | The table's name and short description. |
| Data Model | The model the table belongs to. |
| Attributes | The number of attributes configured. |
| Golden Records | Approved masters, shown as a green pill. |
| Pending | Pending records as an amber pill, or "—" when the queue is clear. |
| Properties | Counters for schema characteristics such as required fields and business keys. |
| Accesses | How many users can reach the table. |
| Responsibles | Stewards and approvers assigned to it. |
| Actions | The Records > button and the row menu. |
The row menu
The three-dot menu on each row is the fastest route to everything table-related:
- Manage Attributes — jump straight to the attribute editor for the table.
- Edit Table — rename the table or update its description.
- Duplicate Table — copy the table and all its attributes into a target enterprise and data model.
- Create Integration — open the inbound sync wizard pre-filled for this table.
- Delete Table — soft-deactivate the table.
Creating a table
Click + Create Table and choose the parent Data Model (pre-selected if the filter is already set), a Name, and an optional Description that will appear in the catalog.
A new table starts with no attributes: defining the schema is the next step. Use Manage Attributes from the row menu to begin, starting with the business key.
Editing and renaming
The Edit dialog shows a summary of the current name and attribute count, and lets you change the name or description. Renaming is safe: existing attributes and records remain intact, and any Snowflake views built on the table are rebuilt automatically to match the new name.
Duplicating a table
The duplicate dialog asks for a target enterprise, a target data model, and a new name. One constraint to know: if the original table has Reference attributes pointing at other tables, duplication succeeds only if equivalent referenced tables exist in the target data model.
Deleting a table
Deletion is a soft deactivation: the table and its records become invisible to all users but remain in the database and can be restored. No master data is destroyed.
Common questions
What is the difference between a table and a data model?
A data model is a folder for a domain; a table is an entity within it that actually holds records. Customers is a data model; Customer Master and Customer Address are tables inside it.
Is it safe to rename a table that downstream systems query?
Yes, with one caveat. Records and attributes are untouched, and the outbound Snowflake view is rebuilt automatically, but the view's name follows the table name, so downstream queries referencing the old view name will need updating. Plan renames with your data team.
Why did my table duplication fail?
The most common cause is a Reference attribute pointing at a table that has no equivalent in the target data model. Duplicate (or create) the referenced tables in the target model first, then retry.
Can a table exist without attributes?
It can exist, but it cannot hold meaningful records until you define its schema. Every table needs at least one business-key attribute before real governance begins: that key is what drives versioning and duplicate detection.