Add dashboard for tracking migration progress

src/databricks/labs/ucx/install.py

@@ -634,9 +634,16 @@ def _handle_existing_dashboard(self, dashboard_id: str, display_name: str, paren
     def _create_dashboard(self, folder: Path, *, parent_path: str) -> None:
         """Create a lakeview dashboard from the SQL queries in the folder"""
         logger.info(f"Creating dashboard in {folder}...")
-        metadata = DashboardMetadata.from_path(folder).replace_database(
-            database=f"hive_metastore.{self._config.inventory_database}",
-            database_to_replace="inventory",
+        metadata = (
+            DashboardMetadata.from_path(folder)
+            .replace_database(  # Assessment and migration dashboards
+                database=f"hive_metastore.{self._config.inventory_database}",
+                database_to_replace="inventory",
+            )
+            .replace_database(  # Migration progress dashboard
+                catalog=self._config.ucx_catalog,
+                catalog_to_replace="ucx_catalog",
+            )
         )
         metadata.display_name = f"{self._name('UCX ')} {folder.parent.stem.title()} ({folder.stem.title()})"
         reference = f"{folder.parent.stem}_{folder.stem}".lower()

src/databricks/labs/ucx/progress/install.py

@@ -47,13 +47,15 @@ class ProgressTrackingInstallation:
     _SCHEMA = "multiworkspace"
 
     def __init__(self, sql_backend: SqlBackend, ucx_catalog: str) -> None:
-        # `mod` is a required parameter, though, it's not used in this context without views.
-        self._schema_deployer = SchemaDeployer(sql_backend, self._SCHEMA, mod=None, catalog=ucx_catalog)
+        from databricks.labs import ucx  # pylint: disable=import-outside-toplevel
+
+        self._schema_deployer = SchemaDeployer(sql_backend, self._SCHEMA, mod=ucx, catalog=ucx_catalog)
 
     def run(self) -> None:
         self._schema_deployer.deploy_schema()
         self._schema_deployer.deploy_table("workflow_runs", WorkflowRun)
         self._schema_deployer.deploy_table("historical", Historical)
+        self._schema_deployer.deploy_view("objects_snapshot", "queries/views/objects_snapshot.sql")
         logger.info("Installation completed successfully!")
 
 

src/databricks/labs/ucx/queries/progress/main/00_0_migration_progress.md

@@ -0,0 +1,16 @@
+---
+height: 4
+---
+
+# Migration Progress
+
+> If widgets show `Unable to render visualization.` verify if
+> the [UCX catalog exists](https://github.com/databrickslabs/ucx?tab=readme-ov-file#create-ucx-catalog-command).
+
+This dashboard displays the migration progress, with data visualized from the `migration-progress-experimental`
+workflow. This workflow is designed to run regularly — either daily or weekly — to provide an up-to-date overview of the
+migration progress.
+
+In addition to offering real-time insights into migration progress, the dashboard also facilitates planning and task
+division. For instance, you can choose to migrate one workspace or schema at a time. By assigning a migration owner to
+each workspace and/or schema, the dashboard shows how the resources allocated to that owner are progressing.

src/databricks/labs/ucx/queries/progress/main/01_0_percentage_migration_progress.sql

@@ -0,0 +1,5 @@
+/* --title 'Overall progress (%)' --width 2 */
+SELECT
+    ROUND(100 * try_divide(COUNT_IF(SIZE(failures) = 0), COUNT(*)), 2) AS percentage
+FROM ucx_catalog.multiworkspace.objects_snapshot
+WHERE object_type IN ('ClusterInfo', 'Grant', 'JobInfo', 'PipelineInfo', 'PolicyInfo', 'Table', 'Udf')

src/databricks/labs/ucx/queries/progress/main/01_1_percentage_udf_migration_progress.sql

@@ -0,0 +1,5 @@
+/* --title 'UDF migration progress (%)' */
+SELECT
+    ROUND(100 * TRY_DIVIDE(COUNT_IF(SIZE(failures) = 0), COUNT(*)), 2) AS percentage
+FROM ucx_catalog.multiworkspace.objects_snapshot
+WHERE object_type = "Udf"

src/databricks/labs/ucx/queries/progress/main/01_2_percentage_grant_migration_progress.sql

@@ -0,0 +1,5 @@
+/* --title 'Grant migration progress (%)' */
+SELECT
+    ROUND(100 * TRY_DIVIDE(COUNT_IF(SIZE(failures) = 0), COUNT(*)), 2) AS percentage
+FROM ucx_catalog.multiworkspace.objects_snapshot
+WHERE object_type = "Grant"

src/databricks/labs/ucx/queries/progress/main/01_3_percentage_job_migration_progress.sql

@@ -0,0 +1,5 @@
+/* --title 'Job migration progress (%)' */
+SELECT
+    ROUND(100 * TRY_DIVIDE(COUNT_IF(SIZE(failures) = 0), COUNT(*)), 2) AS percentage
+FROM ucx_catalog.multiworkspace.objects_snapshot
+WHERE object_type = "JobInfo"

src/databricks/labs/ucx/queries/progress/main/01_4_percentage_cluster_migration_progress.sql

@@ -0,0 +1,5 @@
+/* --title 'Cluster migration progress (%)' */
+SELECT
+    ROUND(100 * TRY_DIVIDE(COUNT_IF(SIZE(failures) = 0), COUNT(*)), 2) AS percentage
+FROM ucx_catalog.multiworkspace.objects_snapshot
+WHERE object_type = "ClusterInfo"

src/databricks/labs/ucx/queries/progress/main/01_5_percentage_table_migration_progress.sql

@@ -0,0 +1,5 @@
+/* --title 'Table migration progress (%)' --width 2 */
+SELECT
+    ROUND(100 * TRY_DIVIDE(COUNT_IF(SIZE(failures) = 0), COUNT(*)), 2) AS percentage
+FROM ucx_catalog.multiworkspace.objects_snapshot
+WHERE object_type = "Table"

src/databricks/labs/ucx/queries/progress/main/01_6_percentage_pipeline_migration_progress.sql

@@ -0,0 +1,5 @@
+/* --title 'Pipeline migration progress (%)' */
+SELECT
+    ROUND(100 * TRY_DIVIDE(COUNT_IF(SIZE(failures) = 0), COUNT(*)), 2) AS percentage
+FROM ucx_catalog.multiworkspace.objects_snapshot
+WHERE object_type = "PipelineInfo"

src/databricks/labs/ucx/queries/progress/main/01_7_percentage_policy_migration_progress.sql

@@ -0,0 +1,5 @@
+/* --title 'Policy migration progress (%)' */
+SELECT
+    ROUND(100 * TRY_DIVIDE(COUNT_IF(SIZE(failures) = 0), COUNT(*)), 2) AS percentage
+FROM ucx_catalog.multiworkspace.objects_snapshot
+WHERE object_type = "PolicyInfo"

src/databricks/labs/ucx/queries/progress/main/01_8_distinct_failures_per_object_type.sql

@@ -0,0 +1,14 @@
+/* --title 'Distinct failures per object type' --width 6 */
+with failures AS (
+  SELECT object_type, explode(failures) AS failure
+  FROM ucx_catalog.multiworkspace.objects_snapshot
+  WHERE object_type IN ('ClusterInfo', 'Grant', 'JobInfo', 'PipelineInfo', 'PolicyInfo', 'Table', 'Udf')
+)
+
+SELECT
+    object_type,
+    COUNT(*) AS count,
+    failure
+FROM failures
+GROUP BY object_type, failure
+ORDER BY object_type, failure

src/databricks/labs/ucx/queries/progress/main/02_0_migration_status_of_tables_and_views.md

@@ -0,0 +1,4 @@
+# Tables and Views
+
+This section presents the migration progress of tables and views, detailing which data objects are migrated and which
+are pending migration.

src/databricks/labs/ucx/queries/progress/main/02_0_owner.filter.yml

@@ -0,0 +1,4 @@
+title: Filter for owner(s)
+column: owner
+type: MULTI_SELECT
+width: 6

src/databricks/labs/ucx/queries/progress/main/02_1_pending_migration_data_objects.sql

@@ -0,0 +1,4 @@
+/* --title 'Pending migration' --description 'Total number of tables and views' --height 6 */
+SELECT COUNT(*) AS count
+FROM ucx_catalog.multiworkspace.objects_snapshot
+WHERE object_type = 'Table' AND array_contains(failures, 'Pending migration')

src/databricks/labs/ucx/queries/progress/main/02_2_migration_status_by_owner_bar_graph.sql

@@ -0,0 +1,24 @@
+/*
+--title 'Pending migration'
+--description 'Tables and views per owner'
+--width 5
+--overrides '{"spec": {
+    "version": 3,
+    "widgetType": "bar",
+    "encodings": {
+        "x":{"fieldName": "owner", "scale": {"type": "categorical"}, "displayName": "owner"},
+        "y":{"fieldName": "count", "scale": {"type": "quantitative"}, "displayName": "count"}
+    }
+}}'
+*/
+WITH owners_with_failures AS (
+    SELECT owner
+    FROM ucx_catalog.multiworkspace.objects_snapshot
+    WHERE object_type = 'Table' AND array_contains(failures, 'Pending migration')
+)
+
+SELECT
+    owner,
+    COUNT(1) AS count
+FROM owners_with_failures
+GROUP BY owner

src/databricks/labs/ucx/queries/progress/main/02_3_migrated_data_objects.sql

@@ -0,0 +1,4 @@
+/* --title 'Migrated' --description 'Total number of tables and views' --height 6 */
+SELECT COUNT(*) AS count
+FROM ucx_catalog.multiworkspace.objects_snapshot
+WHERE object_type = 'Table' AND SIZE(failures) == 0

src/databricks/labs/ucx/queries/progress/main/02_4_migration_status_by_owner_overview.sql

@@ -0,0 +1,15 @@
+/* --title 'Overview' --description 'Tables and views migration' --width 5 */
+WITH migration_statuses AS (
+    SELECT *
+    FROM ucx_catalog.multiworkspace.objects_snapshot
+    WHERE object_type = 'Table'
+)
+
+SELECT
+    owner,
+    DOUBLE(CEIL(100 * COUNT_IF(SIZE(failures) = 0) / SUM(COUNT(*)) OVER (PARTITION BY owner), 2)) AS percentage,
+    COUNT(*) AS total,
+    COUNT_IF(SIZE(failures) = 0) AS total_migrated,
+    COUNT_IF(SIZE(failures) > 0) AS total_not_migrated
+FROM migration_statuses
+GROUP BY owner

src/databricks/labs/ucx/queries/views/objects_snapshot.sql

@@ -0,0 +1,25 @@
+WITH last_workflow_run AS (
+    SELECT
+      workspace_id,
+      MAX(STRUCT(finished_at, workflow_run_attempt, started_at, workflow_run_id)) AS max_struct
+    FROM $inventory.workflow_runs  -- $inventory is a hardcoded name for replacing target schema in a view definition
+    WHERE workflow_name = 'migration-progress-experimental'
+    GROUP BY workspace_id
+)
+
+SELECT
+    historical.workspace_id,
+    historical.job_run_id,
+    historical.object_type,
+    historical.object_id,
+    historical.data,
+    historical.failures,
+    historical.owner,
+    historical.ucx_version
+FROM
+    $inventory.historical AS historical  -- $inventory is a hardcoded name for replacing target schema in a view definition
+  JOIN
+    last_workflow_run
+  ON
+    historical.workspace_id = last_workflow_run.workspace_id
+    AND historical.job_run_id = last_workflow_run.max_struct.workflow_run_id