This page describes the common scenarios and capabilities of AnalyticDB for PostgreSQL.
Use cases
Build a cloud data warehouse
Move data from Alibaba Cloud database services such as ApsaraDB RDS and PolarDB, or self-managed databases, into AnalyticDB for PostgreSQL using Data Transmission Service (DTS) or Data Integration. Schedule extract, transform, and load (ETL) pipelines on petabyte-scale datasets through DataWorks. Connect BI tools — Quick BI, DataV, Tableau, and FineReport — to query live data and publish reports.
Compatible data sources and tools:
Role | Compatible options |
Data sync | DTS, Data Integration |
Source databases | ApsaraDB RDS, PolarDB, self-managed databases |
Task scheduling | DataWorks |
BI and reporting | Quick BI, DataV, Tableau, FineReport |
Accelerate big data analytics
Import data from MaxCompute, Hadoop, and Spark via Data Integration or Object Storage Service (OSS). Run high-performance analysis, processing, and online data exploration directly on that data.
Query data in OSS directly
Use OSS foreign tables to run parallel queries against data stored in OSS. This lets you build an Alibaba Cloud data lake analytics platform on top of your existing OSS data estate.
Capabilities
AnalyticDB for PostgreSQL provides the following capabilities for online analytical processing (OLAP) services:
ETL and offline data processing
AnalyticDB for PostgreSQL handles complex SQL workloads at scale:
Capability | Details |
SQL compatibility | Standard SQL syntax, OLAP window functions, and stored procedures |
Query optimization | ORCA query optimizer handles complex queries without manual tuning |
Processing architecture | MPP architecture processes petabytes of data in seconds |
Storage efficiency | Column store with high compression for fast large-table scans |
Online high-performance queries
For workloads that mix real-time exploration, warehousing, and updating of data:
Capability | Details |
Write throughput | Supports high-throughput INSERT, UPDATE, and DELETE operations |
Point query latency | Millisecond results using row store with B-tree and bitmap indexes |
Transaction support | Distributed transactions, standard isolation levels, and hybrid transaction/analytical processing (HTAP) |
Multi-modal data analysis
For workloads that go beyond structured data:
Capability | Details |
Geospatial analysis | PostGIS extension for geographic data queries |
Machine learning | MADlib in-database ML library for AI-native analytics |
Unstructured data | Vector retrieval for images, audio, and text |
Semi-structured data | JSON support for log analysis and processing |