What you will be able to do
- Explain what Unity Catalog lineage captures and which analysis tasks it supports
- State the conditions for lineage to be captured and the privilege needed to view it
- Open a table's Lineage tab, expand the lineage graph, inspect edges and columns, and find consuming jobs and dashboards
- Predict what a user with limited privileges sees in a lineage graph
1.What lineage records and why you look at it
Data lineage shows where a table's data came from and where it goes. That covers the queries and files that populate the table, the jobs and notebooks that transform it, and the dashboards that consume the results. You do not configure anything for this. Unity Catalog records lineage automatically for queries run on Databricks, down to individual columns, and combines it across every workspace attached to the same metastore. Nodes in the lineage graph can be tables and views, ML model versions, external assets and file paths. If an external source such as Salesforce or a BI tool such as Tableau has been registered as an external asset, it appears in the same graph.
| Task | Question it answers | Direction you look |
|---|---|---|
| Impact analysis | What breaks if I change or delete this table or column? | Downstream |
| Root-cause investigation | Where did the numbers in this report go wrong? | Upstream |
| Sensitive data flow | Where does regulated data originate and who consumes it? | Both |
| Cross-team dependencies | Which teams own my sources or consume my tables? | Both |
Checkpoint 1 of 7· Check yourself
An analyst plans to rename a column in a shared table and wants to know which dashboards and jobs will be affected. What use of lineage is this?
Before changing a table or column, impact analysis identifies the downstream tables, jobs and dashboards that depend on it.
“Perform impact analysis: Before changing or deleting a table or column, identify the downstream tables, jobs, and dashboards that depend on it.”Source: docs.databricks.com
2.When lineage exists, and who can see it
Automatic capture has preconditions. The tables must be registered in a Unity Catalog metastore. The queries must run through the Spark DataFrame API, such as Spark SQL functions that return a DataFrame, or through Databricks SQL interfaces such as notebooks or the SQL query editor. Assets outside the metastore show up only if someone adds them as external metadata objects with relationships to Unity Catalog objects. An empty Lineage tab usually means one of these conditions was not met, not that the table has no history.
Viewing lineage has its own requirement. You need at least the BROWSE privilege on the table's parent catalog, and that catalog must be accessible from your workspace. Lineage nodes that are not tables have their own rules. Notebooks, jobs and dashboards follow the workspace access control settings, and a Unity Catalog-enabled pipeline requires CAN VIEW on the pipeline.
Checkpoint 2 of 7· Check yourself
Which is the minimum Unity Catalog privilege needed to view a table's lineage?
Viewing lineage requires at least BROWSE on the parent catalog. Write privileges and ownership are not required.
“You must have at least the BROWSE privilege on the parent catalog of the table or view.”Source: docs.databricks.com
3.Using the Lineage tab and the lineage graph
In your workspace, click Catalog, then search or browse for the table. Select its Lineage tab, and a panel lists the related tables. To see an interactive diagram, click See Lineage Graph.
The graph opens showing only one level of connections. To reveal more, click the icon on a node, if more connections exist. Click the icon on an edge between two nodes to open the Lineage details panel. It shows the source and target tables and the assets involved in that connection, and you can filter them by notebooks, jobs, pipelines and queries. Selecting an asset in that panel opens it. For column-level lineage, click a column in the graph and its links to related columns appear. This shows, for example, which upstream columns a revenue column was derived from.
Two consumer views are a filter away on the Lineage tab itself. For jobs, select Jobs and then Downstream, and any job that reads the table appears under Job Name. For dashboards, click Dashboards, and consumers appear under Dashboard Name. The same Lineage tab exists for Unity Catalog functions, where it lists connected assets with their Direction, Type and Last activity.
Checkpoint 3 of 7· Put it in order
Put the steps for opening a table's interactive lineage graph in order
- 1.Select the Lineage tab
- 2.Search or browse for the table
- 3.Click See Lineage Graph
- 4.Click Catalog in your workspace
The graph button is on the Lineage tab, which you reach by finding the table in Catalog.
“Select the Lineage tab. The lineage panel appears and displays related tables.”Source: docs.databricks.com
Checkpoint 4 of 7· Check yourself
You want to list the jobs that read from a table. Where do you look?
A job that consumes the table is downstream of it, and it appears under Job Name once you filter the Lineage tab to Jobs and Downstream.
“To view job lineage, go to a table's Lineage tab, select Jobs, and select Downstream.”Source: docs.databricks.com
Checkpoint 5 of 7· Exam question
An admin applies the governed tag `env = production` directly to the `analytics` schema. A table named `daily_metrics` lives inside that schema, and one of its columns is `customer_id`. Without any further tagging action, where does the `env = production` tag now appear?
Correct answer: B — On the `analytics` schema itself and on the `daily_metrics` table, since tables automatically inherit tags from their parent schema, but individual columns like `customer_id` do not.
- A. This is incorrect because tags do inherit from a schema down to the tables it contains automatically, so manual reapplication at the table level is not required.
- B. This is correct because objects within a catalog or schema automatically inherit governed tags applied at that level, with the documented exception that individual columns do not inherit them.
- C. This is incorrect because the inheritance explicitly stops short of columns, so `customer_id` would not receive the tag automatically even though the table does.
- D. This is incorrect because the schema itself retains the tag that was directly applied to it; inheritance only adds the tag to child objects, it does not remove it from the source.
Sources1
4.What a partially privileged user sees
Lineage graphs follow the same permission model as Unity Catalog. Having lineage for a table does not mean you can see all of it. The documentation's example grants userA access to a single table:
GRANT USE SCHEMA on lineage_data.lineagedemo to `userA@company.com`;
GRANT SELECT on lineage_data.lineagedemo.menu to `userA@company.com`;The fix is BROWSE, not SELECT on every table. One grant at schema level lets userB view the lineage graph for any table in the lineage_data schema. Workspace objects work differently. Because lineage is aggregated across workspaces, a table's graph can include notebooks and dashboards created elsewhere, but their details are masked outside the workspace where they were created.
GRANT BROWSE on lineage_data to `userB@company.com`;Checkpoint 6 of 7· Check yourself
A user has neither BROWSE nor SELECT on a table. What can they do with its lineage?
Lineage visibility requires at least BROWSE or SELECT on the table.
“If a user does not have the BROWSE or SELECT privilege on a table, they cannot explore its lineage.”Source: docs.databricks.com
Checkpoint 7 of 7· Exam question
A BI analyst wants to tag the `ssn` column of the `profiles` table with `pii = true` by running `ALTER TABLE ... ALTER COLUMN ... SET TAGS`. They already have `SELECT` on the table. Which additional privileges do they need before the statement will succeed?
Correct answer: D — `APPLY TAG` on the `profiles` table, plus `USE SCHEMA` on its parent schema and `USE CATALOG` on its parent catalog, since tagging requires traversing the full namespace.
- A. This is incorrect because tagging is governed by its own `APPLY TAG` privilege rather than being folded into `MODIFY`, and it still requires the surrounding namespace privileges.
- B. This is incorrect because tagging does not require catalog ownership; a narrower combination of object- and namespace-level privileges is sufficient for a non-owner to apply a tag.
- C. This is incorrect because tagging an existing column is not a table-creation action, so `CREATE TABLE` is unrelated to whether the tag statement succeeds.
- D. This is correct because assigning a tag requires `APPLY TAG` on the object being tagged along with `USE SCHEMA` and `USE CATALOG` to traverse the namespace down to that object.
Exam traps
Each one states something that sounds right. Open it to see what is actually true.
1.The lineage graph opens with the full end-to-end chain of upstream and downstream tables.Why is that wrong?
Only one level is shown by default. You expand further by clicking the icon on a node.
Covered in Using the Lineage tab and the lineage graph
2.Once you can see a table's lineage, you can drill into every connected table.Why is that wrong?
Tables you lack privileges on show up as masked nodes, and you cannot expand past them.
Covered in What a partially privileged user sees
3.Lineage is captured for any table that Databricks queries, whether or not it is registered in Unity Catalog.Why is that wrong?
Automatic capture requires the tables to be registered in a Unity Catalog metastore.
Covered in When lineage exists, and who can see it
Sources
Every claim above is drawn from one of these pages, quoted as it was written on the date shown.
- 1.
“Data lineage shows where data in Databricks came from and where it goes”
↩︎ What lineage records and why you look at it“Nodes can represent tables and views, ML model versions, external assets, and file paths.”
↩︎ What lineage records and why you look at it“Investigate root causes: When a downstream report shows unexpected results, trace upstream sources to find where the data diverged.”
↩︎ What lineage records and why you look at it“Tables must be registered in a Unity Catalog metastore.”
↩︎ When lineage exists, and who can see it“For a Unity Catalog-enabled pipeline, you must have CAN VIEW permission on the pipeline.”
↩︎ When lineage exists, and who can see it“To view an interactive graph of the data lineage, click See Lineage Graph.”
↩︎ Using the Lineage tab and the lineage graph“The Lineage details panel shows details about the connection, including source and target tables.”
↩︎ Using the Lineage tab and the lineage graph“To view column-level lineage, click a column in the graph to show links to related columns.”
↩︎ Using the Lineage tab and the lineage graph“To view dashboard lineage, go to a table's Lineage tab and click Dashboards.”
↩︎ Using the Lineage tab and the lineage graph“Lineage graphs share the same permission model as Unity Catalog.”
↩︎ What a partially privileged user sees“Detailed information about workspace-level objects like notebooks and dashboards in other workspaces is masked.”
↩︎ What a partially privileged user sees“You can use the lineage system tables to programmatically query lineage data.”
↩︎ What a partially privileged user sees“By default, one level is displayed in the graph.”
↩︎ Exam trap 1“userA cannot expand the graph to reveal downstream tables from tables they do not have permission to access”
↩︎ Exam trap 2“Tables must be registered in a Unity Catalog metastore.”
↩︎ Exam trap 3“captures lineage automatically for queries run on Databricks, down to the column level”
↩︎ Prediction“Perform impact analysis: Before changing or deleting a table or column, identify the downstream tables, jobs, and dashboards that depend on it.”
↩︎ Checkpoint“You must have at least the BROWSE privilege on the parent catalog of the table or view.”
↩︎ Checkpoint“Select the Lineage tab. The lineage panel appears and displays related tables.”
↩︎ Checkpoint“To view job lineage, go to a table's Lineage tab, select Jobs, and select Downstream.”
↩︎ Checkpoint“The dinner table appears as a masked node to userA”
↩︎ Prediction“If a user does not have the BROWSE or SELECT privilege on a table, they cannot explore its lineage.”
↩︎ Checkpoint - 2.
“Unity Catalog captures lineage automatically for queries run on Databricks and displays it in Catalog Explorer.”
↩︎ What lineage records and why you look at it - 3.
“Unity Catalog lets you add external lineage metadata to augment the Databricks data lineage it captures automatically”
↩︎ When lineage exists, and who can see it - 4.https://docs.databricks.com/aws/en/catalog-explorerOfficial docs
“Genie Code for prompt-driven information about table lineage and insights.”
↩︎ What a partially privileged user sees