Back to Datahub

README

metadata-ingestion/docs/sources/quicksight/README.md

1.7.02.8 KB
Original Source

Overview

Amazon QuickSight is AWS's serverless, cloud-scale business intelligence service for building interactive dashboards and paginated reports. Learn more in the official QuickSight documentation.

The DataHub integration for QuickSight covers BI entities such as dashboards, analyses, and datasets, along with the folder hierarchy that organizes them. It also stitches cross-platform table-level and column-level lineage from QuickSight datasets back to their upstream warehouse/database tables (Athena, Redshift, Snowflake, S3, and more), and optionally captures ownership, AWS resource tags, users/groups, and stateful deletion detection.

Concept Mapping

QuickSightDataHubNotes
FolderContainerSubType "Folder"; nests via folder membership (Enterprise edition)
NamespaceContainerSubType "Namespace"; opt-in via add_namespace_container
DatasetDatasetSubType "Dataset"; schema from OutputColumns, upstream lineage
AnalysisDashboardSubType "Analysis"
DashboardDashboardSubType "Dashboard"; linked to its source Analysis
VisualChartOne Chart per visual; emitted when extract_dashboard_definitions is enabled
UserUser (a.k.a CorpUser)Optionally extracted via extract_users_and_groups
GroupGroup (a.k.a CorpGroup)Optionally extracted via extract_users_and_groups

QuickSight data sources (the raw warehouse/database connections — Athena, Redshift, Snowflake, S3, etc.) are not modeled as their own entities. As with Tableau/Looker/PowerBI, the connection is used purely to resolve the upstream platform of each Dataset's tables for lineage; the lineage points directly at the warehouse table.

Account separation is handled via platform_instance (the Glue / Redshift / PowerBI convention) rather than an account-level container.