Selection guide and feature comparison between Tair (Enterprise Edition) and Redis Open-Source Edition - Tair (Redis® OSS-Compatible)

This topic compares the features of different series of Tair (Enterprise Edition) with the Redis Open-Source Edition to help you select the right product.

Selection guide

Edition	Series	Features	Scenarios
Tair (Enterprise Edition)	DRAM-based instance	Ultra-high performance: Its multi-threaded model delivers up to three times the read and write performance of a Redis Open-Source Edition instance with the same specifications. For more information, see Performance white paper for Tair DRAM-based, Persistent memory-optimized, and Redis Open-Source Edition instances. Provides proprietary Tair extended data structures, including exString (with enhanced Redis String commands), exHash, exZset, GIS, Bloom, Doc, TS, Cpc, Roaring, Search, and Vector. These data structures help you streamline code, improve application performance, and focus on business innovation. Supports many enterprise-grade features, such as point-in-time recovery (PITR), Global Distributed Cache, and proxy query cache. Supports advanced enterprise-grade encryption, such as TLS encryption and Transparent Data Encryption (TDE). Supports semi-synchronous mode, which improves data consistency over asynchronous replication. Supports real-time tracing to track read and write requests to an instance, helping you quickly identify and troubleshoot issues. Supports high-consumption key statistics with custom thresholds to provide real-time statistics and alerts. Supports server-side RBAL rate limiting to ensure instance stability.	Mission-critical, performance-centric business workloads.
	Persistent memory-optimized instance	Offers ultra-high cost-effectiveness. The price of persistent memory-optimized instances is approximately 30% lower than that of Redis Open-Source Edition instances that have the same capacity. The performance of persistent memory-optimized instances reaches 90% of that of Redis Open-Source Edition instances. For more information, see Performance whitepaper of persistent memory-optimized instances. Supports enhanced data modules. These modules include exString (including commands that enhance Redis string functionality), exHash, and Cpc. Prevents data loss caused by power failure. Persistent memory-optimized instances implement persistence for each command. The system returns a success response for each write operation only after the data is persistently stored. You can use persistent memory-optimized instances as in-memory databases instead of caches.	Data caching and storage scenarios that require high performance and strong data persistence, where cost is a secondary concern.
	ESSD/SSD-based instance	Reduced costs: ESSD/SSD-based instances reduce up to 85% of costs compared with Redis Open-Source Edition instances. High performance: An ESSD/SSD-based instance delivers approximately 60% of the performance of a Redis Open-Source Edition instance. For more information, see Performance whitepaper of ESSD-based instances and Performance whitepaper of SSD-based instances. Storage in disks: ESSD/SSD-based instances store data in ESSDs or SSDs to ensure high data reliability. The capacity of an ESSD/SSD-based instance can reach hundreds of terabytes. Data distribution: ESSD/SSD-based instances use the Alibaba Cloud TairDB storage engine and combine disks and memory to provide an optimal balance between data persistence and quick access to data. High compatibility: ESSD/SSD-based instances are compatible with most data structures and commands of Redis 6.0.	Data storage scenarios that require large capacity and low costs, with low access density and no ultra-low latency requirement.
Redis Open-Source Edition	N/A	High performance and compatibility with open-source Redis.	Standard Redis use cases and migration scenarios.

Note

For more information about product selection, see Product selection guide.

Feature comparison

In the following table, ✔️ indicates that a feature is supported, and ❌ indicates that it is not supported.

Category	Item	Tair (Enterprise Edition)				Redis Open-Source Edition
Category	Item	DRAM-based instance	Persistent memory instance	ESSD instance	SSD instance	Versions 2.8, 4.0, and 5.0	Versions 6.0 and 7.0	Redis Yitian Edition
Baseline performance	Performance benchmark (relative to Redis Open-Source Edition)	300%	90%	Read: 40%	Read: 60%	Same	120%	120%
		300%	90%	Write: 30%	Write: 40%	Same	120%	120%
	Maximum number of connections per data node	30,000	10,000	10,000	40,000	10,000	10,000	10,000
	Single-key processing capability (QPS reference) ①	450,000	130,000	30,000–60,000	50,000–60,000	140,000	160,000	160,000
	Maximum bandwidth (MB/s)	96–2,048	96–2,048	187.5–1,000	187.5–2,048	10–2048	48–2,048	96–2,048
Specifications	I/O and worker model	Real Multi-I/O (multiple I/O threads) ③	Single-threaded	Real Multi-I/O (multiple I/O threads + multiple worker threads)	Real Multi-I/O (multiple I/O threads + multiple worker threads)	Single-threaded	Multiple I/O threads + single worker thread	Multiple I/O threads + single worker thread
Data structure	Basic data structures and command support	Supported commands vary by instance series. For more information, see Tair (Enterprise Edition) command support and limits.				Some commands are not supported. For more information, see Redis Open-Source Edition command support.
Data structure	Tair extended data structures	✔️	✔️ (Partial)	❌	❌	❌	❌	❌

Note

Tair (Enterprise Edition) does not support Redis Stack modules (such as RediSearch, RedisJSON, or RedisTimeSeries). Instead, it provides proprietary Tair extended data structures with equivalent functionality: TairSearch for full-text search, TairDoc for JSON operations, and TairVector for AISearch. These data structures cover similar use cases but use different implementations and commands. For example, the FT.SEARCH (RediSearch) and TS.RANGE (RedisTimeSeries) commands are not supported in Tair (Enterprise Edition).

Category	Item	Tair (Enterprise Edition)				Redis Open-Source Edition
Category	Item	DRAM-based instance	Persistent memory instance	ESSD instance	SSD instance	Versions 2.8, 4.0, and 5.0	Versions 6.0 and 7.0	Redis Yitian Edition
Data persistence	Primary-replica replication consistency	eventual consistency	eventual consistency	eventual consistency	eventual consistency	eventual consistency	eventual consistency	eventual consistency
	Persistence consistency ④	Write Back	Write Through	Write Through	Write Through	Write Back	Write Back	Write Back
	Persistence level	Second-level	Command-level	Command-level	Command-level	Second-level	Second-level	Second-level
Security	TLS encryption	✔️	✔️	❌	❌	✔️	✔️	✔️
	Transparent Data Encryption (TDE)	✔️	❌	❌	❌	❌	❌	❌
	IP whitelist	✔️	✔️	✔️	✔️	✔️	✔️	✔️
Performance analysis	Top key statistics	✔️	✔️	❌	❌	✔️	✔️	✔️
	Historical hotkey query	✔️	✔️	✔️	✔️	✔️	✔️	✔️
	Offline large key analysis	✔️	✔️	❌	❌	✔️	✔️	✔️
	Audit log	✔️	✔️	✔️	✔️	✔️	✔️	✔️
Advanced features	Point-in-time recovery (PITR)	✔️	✔️	❌	✔️	❌	❌	❌
	Semi-synchronous mode	✔️	❌	✔️	✔️	❌	❌	❌
	Proxy query cache	✔️	✔️	❌	❌	❌	❌	❌
	Global Distributed Cache	✔️	❌	❌	❌	❌	❌	❌
	DTS one-way synchronization	✔️	✔️	✔️	✔️	✔️	✔️	✔️
	DTS two-way synchronization	✔️	✔️	✔️	❌	❌	❌	❌

The following notes correspond to the numbers in the table:

①: The queries per second (QPS) reference value is based on commands with a time complexity of O(1). The higher the time complexity, the lower the QPS.
②: Performance depends on the hot/cold data access distribution. A higher memory hit ratio results in performance closer to the Redis Open-Source Edition benchmark.
③: Unlike the I/O multi-threading in Redis Open-Source Edition 6.0, the Real Multi-I/O in DRAM-based instances thoroughly accelerates both I/O and command execution. This provides greater resistance to connection surges and allows for linear throughput scaling.
④: Data is persisted to disk in one of two ways:
- Write Through: The system returns a success response only after the data is synchronously persisted to disk.
- Write Back: The system returns a success response as soon as the data is written to the cache. The data is then asynchronously flushed to disk.