August expands the database foundation across document scale, lakehouse access, isolated development, and AI observability. PolarDB for MySQL scales MongoDB-compatible workloads with Limitless multi-master architecture, PolarDB for PostgreSQL unifies metadata across multiple catalogs, AnalyticDB for PostgreSQL brings branch-based workflows to data development, and ApsaraDB for SelectDB strengthens observability for LLM and agent applications. Below is an at-a-glance view of all 10 verified releases, followed by four highlights with the broadest promotional value.
| Product | Feature | Type |
|---|---|---|
| PolarDB for MySQL | Limitless Multi-Master Architecture for MongoDB Compatibility | New Feature |
| PolarDB for PostgreSQL | LakeBase Catalog | New Feature |
| PolarDB for PostgreSQL (Compatible with Oracle) | Structure and Privilege Verification Tool | New Feature |
| AnalyticDB for PostgreSQL | Data Branch Management | New Feature |
| ApsaraDB RDS for PostgreSQL | AliPG 20260830 | New Version |
| Tair | France (Paris) Region | Region Expansion |
| ApsaraDB for ClickHouse | ClickHouse 26.3 LTS | New Version |
| ApsaraDB for SelectDB | More Transparent Kernel Version Upgrades | Enhancement |
| ApsaraDB for SelectDB | DataLens AI Observability with Langfuse Compatibility | Enhancement |
| ApsaraDB for SelectDB | AI Assistant | New Feature |
What it is:
PolarDB for MySQL expands its MongoDB compatibility capabilities with Limitless multi-master architecture, allowing databases or collections to be sharded across up to 63 read/write nodes.
Description:
MongoDB-compatible applications can outgrow a replica-set architecture and require sharding to distribute data and throughput. PolarDB for MySQL now provides a Limitless architecture that combines horizontal sharding with MongoDB Wire Protocol compatibility, giving document workloads a path to scale on PolarDB while continuing to use standard MongoDB drivers. The architecture can also simplify migration from existing MongoDB replica sets or sharded clusters by reducing application-level protocol changes. Each database or collection has one write-owning node at any given time, so “multi-master” refers to ownership distributed across the cluster rather than concurrent writes from multiple nodes to the same collection.
Why developers and enterprises should pay attention:
Note: This capability is in canary release and requires submitting a ticket to apply for access. A cluster supports up to 63 read/write nodes. Each database or collection has only one write-owning node at any given time. Available regions and versions depend on the approved activation scope.
What it is:
PolarDB for PostgreSQL introduces LakeBase Catalog to centrally manage metadata from PostgreSQL, Apache Paimon, DuckDB, and Lance catalogs and query across them with federated SQL.
Description:
Lakehouse and AI teams often spend significant effort locating datasets, maintaining multiple metadata systems, and moving data before it can be queried together. LakeBase Catalog provides a unified entry point for discovering and accessing heterogeneous data while allowing each catalog to remain in its appropriate system. Federated SQL enables queries across catalogs, reducing the need to create another copy of the data whenever an application needs a combined view. This shared catalog layer gives analytics and AI workloads a clearer path from metadata discovery to governed data access.
Why developers and enterprises should pay attention:
Note: LakeBase Catalog is currently in public preview and requires submitting a ticket to apply for access.
What it is:
AnalyticDB for PostgreSQL introduces data branch management for pay-as-you-go AnalyticDB Supabase projects, enabling isolated data environments with independent connection information.
Description:
Development and testing against shared data can introduce contamination, accidental changes, and lengthy environment preparation. Data branch management lets teams create isolated branches without placing the MAIN branch directly in the path of experiments. It supports nested branches, branch protection, schema viewing, branch reset, snapshots, and point-in-time recovery, giving developers multiple ways to create, protect, and restore working data environments. A reset preserves the branch connection information, reducing the need to reconfigure applications after refreshing a branch from its parent.
Why developers and enterprises should pay attention:
Note: The feature is currently in invitation-only preview. It is free during the preview and does not provide an SLA. The MAIN branch cannot be deleted or reset. A branch reset overwrites rather than merges data and briefly interrupts connections.
What it is:
ApsaraDB for SelectDB enhances DataLens AI with Langfuse-compatible end-to-end tracing, prompt management, evaluation, and scoring for LLM and agent applications.
Description:
AI application teams need reliable ways to understand model calls, prompts, traces, evaluations, and scores as projects grow. DataLens AI brings these observability workflows together and is compatible with Langfuse, reducing the friction of instrumenting and analyzing LLM and agent behavior. This update redesigns query capabilities for large projects and adds an exactly-once ingestion pipeline, project-level table isolation, and improvements to evaluation, scoring, and self-hosted deployment. The result is a more reliable observability data path and clearer separation between projects at scale.
Why developers and enterprises should pay attention:
Note: DataLens AI is currently in invitation-only preview.
PolarDB for PostgreSQL (Compatible with Oracle) adds a post-migration verification tool that compares object structures as well as users, roles, and privileges between an Oracle source database and the PolarDB destination. Two filterable HTML reports help migration teams locate discrepancies before production cutover. The tool performs comparisons only; it does not migrate data or automatically repair differences.
Note: The runtime environment requires Java 1.8 or later, connectivity to both databases, and an execution account with permission to read the required metadata.
ApsaraDB RDS for PostgreSQL releases AliPG 20260830 with broader extension capabilities. PostgreSQL 18 adds tds_fdw, PostgreSQL 17 and 18 upgrade pg_textsearch, PostgreSQL 18 upgrades Apache AGE, and rds_duckdb adds the all, record, and none automatic table-creation modes for applicable versions. These changes expand external data access, text search, graph analytics, and DuckDB-based analysis.
Note: Extension availability depends on the PostgreSQL major version, instance minor engine version, and regional rollout. Applicable versions for rds_duckdb are subject to the console and extension documentation.
Tair adds the France (Paris) region, giving European workloads another nearby deployment option and helping reduce cross-region access latency. The expansion also gives organizations another location to consider when planning a multi-region caching or in-memory database footprint in Europe.
Note: Released on August 21, 2026. The region supports cloud-native instances only and does not support classic instances. Available specifications, zones, and product series are subject to what is displayed on the purchase page.
ApsaraDB for ClickHouse Community-compatible Edition 26.3.2.1-1.3.0 integrates ClickHouse Community Edition 26.3 LTS (v26.3.17.4-lts). The release expands access to Paimon, Iceberg, Delta Lake, and Hudi and adds inverted indexes, intelligent JOIN reordering, materialized CTEs, Bucketed Map, vertical merge for TTL, ALP floating-point compression, and compatibility with multiple SQL dialects. It gives analytical workloads more options to query lake data, optimize joins, and reduce migration rewrites.
Note: Paimon currently supports read access only.
ApsaraDB for SelectDB enhances kernel version upgrades with pre-upgrade parameter compatibility checks and SQL replay to surface risks early. The platform then orchestrates node upgrades and service switching, providing a standardized path to new features, performance improvements, and fixes with less manual coordination.
Note: Brief connection interruptions may occur. A major-version upgrade may cause approximately 10 minutes of downtime, while a parallel upgrade may cause approximately 1 to 5 minutes. Rollback is not supported after an upgrade starts, and other instance lifecycle operations are unavailable during the upgrade. Business impact and retry mechanisms must be evaluated in advance.
ApsaraDB for SelectDB introduces AI Assistant to analyze slow SQL by combining product knowledge with SQL audit logs, execution plans, Query Profiles, and table schemas. It provides evidence-based root-cause analysis, optimization recommendations, and operational guidance, shortening the path from a slow-query alert to a concrete tuning action.
August's verified releases strengthen the database foundation from several directions at once: document workloads gain a path to horizontal scale, lakehouse data becomes easier to discover and query together, development teams gain safer data isolation, and AI applications gain deeper observability. The same release set also improves migration validation, engine capabilities, upgrade workflows, diagnostics, and regional coverage. Follow the next monthly roundup for more updates across the Alibaba Cloud database portfolio.
Learn more about Alibaba Cloud Database services: ApsaraDB Product Page
ApsaraDB - September 27, 2025
ApsaraDB - September 12, 2024
ApsaraDB - March 29, 2026
ApsaraDB - April 20, 2026
ApsaraDB - May 21, 2026
ApsaraDB - February 27, 2026
ApsaraDB RDS for MariaDB
ApsaraDB RDS for MariaDB supports multiple storage engines, including MySQL InnoDB to meet different user requirements.
Learn More
ApsaraDB RDS for MySQL
An on-demand database hosting service for MySQL with automated monitoring, backup and disaster recovery capabilities
Learn More
ApsaraDB RDS for PostgreSQL
An on-demand database hosting service for PostgreSQL with automated monitoring, backup and disaster recovery capabilities
Learn More
ApsaraDB RDS for SQL Server
An on-demand database hosting service for SQL Server with automated monitoring, backup and disaster recovery capabilities
Learn MoreMore Posts by ApsaraDB
