Persistent memory-optimized instances adopt Intel® Optane™ DC persistent memory (AEP) to provide you with in-memory databases that have a large capacity and are compatible with open source Redis. A persistent memory-optimized instance does not use disks to implement data persistence. It saves up to 30% of costs compared with an ApsaraDB for Redis Community Edition instance and provides almost the same performance in terms of throughput and latency. This helps improve the reliability of business data.

Purchase an instance

Create a persistent memory-optimized or storage-optimized instance of ApsaraDB for Redis Enhanced Edition (Tair)

Background information

The high price and low capacity of memory limit the large-scale use of memory in specific scenarios. Alibaba Cloud began to invest in the research and implementation of Intel AEP in 2018. Intel AEP was applied to the core cluster of e-commerce products during Double 11 that year. Costs were remarkably reduced. The cluster became the first product in China that officially deployed Intel persistent memory in a production environment.

Maturer cloud environments and Improved AEP-related technologies help Alibaba Cloud develop a new engine for data persistence based on AEP. Alibaba Cloud integrates the new engine with ECS Bare Metal instances and introduces persistent memory-optimized instances of ApsaraDB for Redis Enterprise Edition (Tair). These instances replace the volatile memory of traditional Redis with the native persistent memory of AEP. This significantly reduces the risk of data loss.

Persistent memory-optimized instances provide not only memory-level access latency and throughput but also data persistence. In addition to reducing costs, persistent memory-optimized instances can simplify the application architecture. The popular architecture consisting of applications, cache, and persistent storage can be simplified to an architecture consisting of applications and persistent memory-optimized instances, as shown in the following figure.

Architecture evolution of persistent memory-optimized instances

Benefits

Persistent memory-optimized instances of ApsaraDB for Redis Enhanced Edition (Tair) adopt Intel AEP to provide you with in-memory databases that have a large capacity and are compatible with open source Redis. A persistent memory-optimized instance does not use disks to implement data persistence. It provides almost the same performance as an ApsaraDB for Redis Community Edition instance in terms of throughput and latency. This helps improve the reliability of business data. Persistent memory-optimized instances apply to scenarios that store hot and warm data, have high requirements for data persistence, and require compatibility with Redis, large capacity, and stable and controllable jitter.

Item Description
High cost-effectiveness
  • A persistent memory-optimized instance saves about 30% of costs compared with an ApsaraDB for Redis Community Edition instance.
  • A persistent memory-optimized instance can provide 90% of the performance of a native Redis database.
Optimization for high specifications
  • In scenarios that adopt high specifications of memory, the issues of latency jitter and slow service data loading during fork operations triggered by append-only file (AOF) rewrites are resolved. You do not need to choose between performance and persistence.
  • The memory size of master-replica instances ranges from 128 GB to 512 GB. The memory size of cluster instances can reach up to 32 TB.
Data loss prevention during power outages
  • Persistent memory-optimized instances support command-level persistence. A response is returned after data persistence is complete for each write operation.
  • If you use open source Redis, you must take the risk of second-level data loss. Persistent memory-optimized instances are more reliable when the QPS is high.
High compatibility
  • Persistent memory-optimized instances are compatible with the ApsaraDB for Redis database system and support high availability, auto scaling, logging, intelligent diagnostics, and flexible backup and restoration.
  • Persistent memory-optimized instances are compatible with most of the data structures and interfaces of native Redis.

Scenarios

  • Scenarios that require high performance and reduced costs for processing a large amount of data

    Data computing requires high performance. ApsaraDB for Redis Community Edition is costly. Other database types such as HBase cannot meet the performance requirements. Persistent memory-optimized instances not only ensure data persistence but also provide almost the same performance as ApsaraDB for Redis Community Edition instances in terms of throughput and latency. This can help you get a balance between performance and costs.

  • Scenarios that have high requirements for data persistence

    For gaming services, persistent memory-optimized instances are used for data storage. Compared with the architecture that uses Redis together with MySQL, persistent memory-optimized instances provide a simpler architecture, higher performance, higher cost-effectiveness, and higher data reliability.

Instance specifications

Persistent memory-optimized instances