Planning

Cloud-native architecture

AnalyticDB for PostgreSQL separates computing from storage, so you can:

• Scale compute independently of storage — no data movement required.

• Complete scaling within seconds without service interruption.

Cost efficiency

AnalyticDB for PostgreSQL continuously optimizes its execution engine, optimizers, and storage engine to deliver high-throughput, low-latency, high-concurrency reads and writes. For specific workloads, the following capabilities reduce query latency and storage costs:

• Real-time materialized views: Precompute and persist query results for fast repeated access.

• Parallel queries: Distribute query execution across compute nodes to accelerate large-scale analytics.

• Result set cache: Return cached results for repeated identical queries, reducing compute overhead.

• Hierarchical storage: Reduce storage costs by moving less-frequently accessed data to lower-cost storage tiers.

HTAP scenarios

AnalyticDB for PostgreSQL supports hybrid row-column storage and multi-replica query extension, enabling a single system to handle both online transaction processing (OLTP) and analytical workloads without resource contention:

• Row store: Optimized for transactional processing with low-latency point reads and writes.

• Column store: Optimized for analytical processing with high-throughput sequential scans.

Multi-replica query extension supports both online transaction processing and analytical processing within a single system.