All Products
Search
Document Center

Tair (Redis® OSS-Compatible):End-of-sale for ApsaraDB for Redis hybrid-storage instances

Last Updated:Jun 16, 2026

ApsaraDB for Redis hybrid-storage instances are no longer available for purchase. Consider persistent memory-optimized instances as an alternative.

End-of-sale date

May 28, 2021

Impacts

Purchased

Impact

Yes, you have purchased hybrid-storage instances.

Existing hybrid-storage instances remain available for use.

No, but you plan to purchase hybrid-storage instances.

Hybrid-storage instances are no longer available as an option when you create ApsaraDB for Redis instances.

Recommendations

Tair (Enterprise Edition) is a high-availability, high-performance distributed NoSQL database service. Tair offers instance series based on DRAM, NVM, and Enterprise SSDs (ESSDs), providing low-latency access, data persistence, and lower overall costs. Tair also delivers enhanced data structures and flexible storage methods for a wide range of scenarios. For more information, see Overview.

  • DRAM-based instances: use a multi-threading model and deliver read/write performance approximately three times that of same-spec Redis Open-Source Edition instances. DRAM-based instances also provide enhanced data modules, allowing you to focus on application development without worrying about storage structures and timeliness.

  • Persistent memory-optimized instances: use persistent memory to provide large-capacity, Redis-compatible in-memory databases that persist each operation without relying on traditional disk storage. Compared with Redis Open-Source Edition instances, persistent memory-optimized instances reduce costs by up to 30% while delivering comparable throughput and latency, improving business data reliability.

  • ESSD/SSD-based instances: store data on ESSDs or standard SSDs and are compatible with the core data structures and APIs of open source Redis. These instances deliver large-capacity, low-cost, persistent database services while improving data reliability and reducing the reserved memory required for open source Redis forks.