All Products
Search
Document Center

Tair (Redis® OSS-Compatible):References for selecting an appropriate Tair (Redis OSS-compatible) instance

Last Updated:Jan 23, 2025

Before you create a Tair (Redis OSS-compatible) instance, you need to select the most cost-effective and stable instance type that suits your performance, price, scenario, and workload needs. For example, you may want to use instances as high-speed caches or in-memory databases. This topic describes the instance types, disaster recovery solutions, architectures, and instance specifications of Tair (Redis OSS-compatible) to help you select the appropriate instances.

Selection references

Select an instance type

Tair (Redis OSS-compatible) provides Redis Open-Source Edition and Tair (Enterprise Edition). Tair (Enterprise Edition) is an enterprise-level, in-memory database service that is developed based on the business practices of Alibaba. Tair (Enterprise Edition) introduces multiple series types of instances based on storage mediums such as dynamic random-access memory (DRAM), NVM, and SSDs or Enterprise SSDs (ESSDs) to meet your requirements for low-latency access, data persistence, and reduced overall costs. Tair (Enterprise Edition) provides you with higher performance, more data structures, and more flexible storage methods to help you meet business requirements in different scenarios.

Important

For information about the commands and parameters supported by each series type, see Command overview and Configure instance parameters.

The following table describes the instance types.

Comparison item

Redis Open-Source Edition

Tair (Enterprise Edition)

Deployment mode

Cloud-native (recommended)

Classic

Cloud-native

Storage medium

Memory

DRAM

NVM

ESSD/SSD

Compatible Redis versions

5.0, 6.0, and 7.0

5.0, 6.0, and 7.0

6.0

6.0

Performance

100% (benchmark)

300%

90%

Up to 60%

Cost

100% (benchmark)

Approximately 117% (calculated at a monthly discount of 30% off)

Approximately 70%

15% to 20%

Features

Open source Redis services on the cloud.

  • High cost efficiency: provides up to 30% cost reduction compared with Redis Open-Source Edition.

  • High reliability: supports command-level persistence. A response is returned after data persistence is completed for each write operation.

Combines disks and memory to provide an optimal balance between data persistence and quick access to data.

Scenarios

Use scenarios of open source Redis.

  • Scenarios that require short response time, such as live streaming, online education, online gaming, and RTA.

  • Caching scenarios that involve tens of millions of queries per second (QPS), such as online shopping and social networking.

Scenarios that require high performance and reduced costs for processing a large amount of data, such as Internet of Things (IoT).

Scenarios that require large storage space and high access performance to store warm and cold data and take cost as the primary consideration, such as indexing of file storage and long-term storage of historical messages.

Note

For more information about the differences in features and performance between Redis Open-Source Edition and Tair (Enterprise Edition), see Comparison between Tair (Enterprise Edition) and Redis Open-Source Edition.

Select a disaster recovery solution

Tair (Redis OSS-compatible) provides three disaster recovery solutions: single-zone high-availability (HA) solution, zone-disaster recovery solution, and cross-region disaster recovery solution. You can select a disaster recovery solution based on your business requirements.

Disaster recovery solution

Description

Operation

Single-zone HA solution

The master and replica nodes are deployed on different machines in the same zone to provide machine-level fault recovery capabilities.

On the buy page, set Deployment Type to Single-zone Deployment.

Zone-disaster recovery (multi-zone) solution

The master and replica nodes are deployed in different zones (data centers) of the same region to provide data center-level fault recovery capabilities.

On the buy page, set Deployment Type to Dual-zone Deployment.

Cross-region disaster recovery solution

Multiple child instances are deployed in different regions to form a Global Distributed Cache instance. This solution provides region-level (natural disaster) fault recovery capabilities. For more information, see Global Distributed Cache.

For more information, see Create a distributed instance.

Select an architecture

Tair (Redis OSS-compatible) supports the standard architecture (with cluster mode disabled), the cluster architecture, and the read/write splitting feature (optional). You can select an architecture that suits your business data volume and your requirements for read and write capabilities and business performance.

Instance architecture

Architecture model

Data distribution

Scenario

Read/write splitting

Standard architecture

One shard, using a master-replica model.

All data is stored in a single shard.

  • A single node can handle business data and traffic.

  • Commands are simple with few operations such as sorting and calculation.

You can enable or disable the read/write splitting feature. If you enable read/write splitting, you can specify the number of read replicas. Each instance can consist of one master node and up to nine read replicas.

Cluster architecture

Consists of a proxy node and multiple shards. Each shard uses a master-replica model.

Data is distributed across shards.

  • A single shard cannot handle all business data and traffic.

  • The commands involved are complex and time-consuming to run.

You can enable or disable the read/write splitting feature. Each shard can consist of one master node and up to four read replicas.

Select instance specifications

You must select the appropriate instance specifications and quantities (shard specifications and number of shards) based on a comprehensive estimation of business metrics such as capacity, bandwidth, number of connections, and QPS. We recommend that you keep your usage below 80% of the total resources you acquire.

When you estimate the capacity, you do not need to consider the memory overhead caused by the write-time replication of forks and advanced features such as whitelist configuration, audit logging, large key management, and hotkey management. The preceding memory overhead is borne by Alibaba Cloud and is not included in the memory size of the purchased instance.

Important

Large keys are a common issue in Redis usage. If the total capacity of a cluster is large and the capacity of a single shard is small, when large keys are generated, the capacity of the shard where the large keys reside is more likely to be exhausted.

The following table describes the recommended shard specifications for the cluster architecture.

Total instance capacity

Recommended shard specifications

16 GB to 64 GB

2 GB or higher

64 GB to 256 GB

4 GB or higher

Greater than 256 GB

8 GB or higher

Note

If the selected instance type does not meet your business requirements, you can change the instance configurations at any time. For more information, see Change the configurations of an instance.

References on what to do next