Ingest Glossary from dbt | Official Documentation

Import glossary definitions from dbt projects to align semantic metadata with business context and governance.

Ingest Glossary from dbt

Ingest the table and column level glossary terms from manifest.json file

Requirements

For dbt Glossary, Glossary terms must be created or present in OpenMetadata beforehand for data ingestion to work.

Steps for ingesting dbt Glossary

1. Create a Glossary at OpenMetadata or Select a previously added glossary

A Glossary Term is a preferred terminology for a concept. In a Glossary term, you can add tags, synonyms, related terms to build a conceptual semantic graph, and also add reference links.

For details on creating glossary terms, refer to the OpenMetadata documentation

To view created Glossary Terms, navigate to the Glossary section within OpenMetadata govern->glossary->glossary_name->glossary_term_name

OpenMetadata also supports creating nested Glossary Terms, allowing you to organize them hierarchically and seamlessly ingest them into dbt.

2. Add Table-Level Glossary term information in schema.yml file

To associate glossary terms with specific tables in your dbt model, you’ll need their Fully Qualified Names (FQNs) within OpenMetadata.

Steps to Get Glossary Term FQNs:

1. Navigate to the desired glossary term in OpenMetadata’s glossary section.
2. The glossary term’s details page will display its FQN e.g. Glossary_name.glossary_term in the url like your-uri/glossary/Glossary_name.glossary_term.

Example

Suppose you want to add the glossary terms term_one (FQN: Test_Glossary.term_one) and more_nested_term (FQN: Test_Glossary.term_two.nested_term.more_nested_term) to the customers table in your dbt model.

To get FQN for term_one (Test_Glossary.term_one), navigate to govern->glossary->Test_Glossary->term_one.

And for more_nested_term (Test_Glossary.term_two.nested_term.more_nested_term), navigate to govern->glossary->Test_Glossary->term_two->nested_term->more_nested_term.

you can see the current url containing the glossary term FQNs as https://localhost:8585/glossary/`Test_Glossary.term_two.nested_term.more_nested_term`

In your dbt schema.yml file for the customers table model, add the Glossary Term FQNs under model->name->meta->openmetadata->glossary The format should be a list of strings, like this: [ 'Test_Glossary.term_one', 'Test_Glossary.term_two.nested_term.more_nested_term' ].

For details on dbt meta follow the link here

models:
  - name: customers
    meta:
      openmetadata:
        glossary: [
          'Test_Glossary.term_one',
          'Test_Glossary.term_two.nested_term.more_nested_term',
        ]
    description: This table has basic information about a customer, as well as some derived facts based on a customer's orders

    columns:
      - name: customer_id
        description: This is a unique identifier for a customer
        tests:
          - unique
          - not_null

After adding the Glossary term information to your schema.yml file, run your dbt workflow. The generated manifest.json file will then include the FQNs under node_name->meta->openmetadata->glossary as [ 'Test_Glossary.term_one', 'Test_Glossary.term_two.nested_term.more_nested_term' ]

"model.jaffle_shop.customers": {
  "raw_sql": "sample_raw_sql",
  "compiled": true,
  "resource_type": "model",
  "depends_on": {},
  "database": "dev",
  "schema": "dbt_jaffle",
  "config": {
      "enabled": true,
      "alias": null,
      "meta": {
          "openmetadata": {
              "glossary": [
                "Test_Glossary.term_one",
                "Test_Glossary.term_two.nested_term.more_nested_term"
              ]
          }
      }
  }
}

3. Add Column-Level Glossary term information in schema.yml file

To associate a glossary term with a specific column in your dbt model, follow these steps:

1. Locate the customer_id column within the customers table model in your schema.yml file.
2. Under the customer_id column definition, add the glossary term FQNs under model->name->columns->column_name->meta->openmetadata->glossary as [ 'Test_Glossary.term_two.nested_term' ].

models:
  - name: customers
    meta:
      openmetadata:
        glossary: [
          'Test_Glossary.term_one',
          'Test_Glossary.term_two.nested_term.more_nested_term',
        ]
    description: This table has basic information about a customer, as well as some derived facts based on a customer's orders

    columns:
      - name: customer_id
        description: This is a unique identifier for a customer
        meta:
          openmetadata:
            glossary: [
              'Test_Glossary.term_two.nested_term'
            ]
        tests:
          - unique
          - not_null

After adding the Glossary term information to your schema.yml file, run your dbt workflow. The generated manifest.json file will then include the FQNs under node_name->columns->column_name->meta->openmetadata->glossary as [ 'Test_Glossary.term_two.nested_term' ]

"model.jaffle_shop.customers": {
  "raw_sql": "sample_raw_sql",
  "compiled": true,
  "resource_type": "model",
  "depends_on": {},
  "database": "dev",
  "schema": "dbt_jaffle",
  "columns": {
    "customer_id": {
      "name": "customer_id",
      "description": "This is a unique identifier for a customer",
      "meta": {
        "openmetadata": {
          "glossary": [
            "Test_Glossary.term_two.nested_term"
          ]
        }
      },
      "data_type": null,
      "constraints": [],
      "quote": null,
      "tags": []
    },
  }
}

4. Viewing the Glossary term on tables and columns

Table and Column level Glossary term ingested from dbt can be viewed on the node in OpenMetadata