Databases 14 min read

Best Practices for Redis Database on the Cloud

In this talk, Tencent Cloud senior engineer Zou Peng outlines cloud‑native best practices for Redis 4.0 clusters, detailing a smart‑proxy architecture with gossip‑based management, lazy‑scan big‑key handling, automatic fail‑over, the high‑performance CKV engine, and practical guidance on scaling, sharding, monitoring, and cost‑effective high‑availability deployment.

Tencent Cloud Developer
Tencent Cloud Developer
Tencent Cloud Developer
Best Practices for Redis Database on the Cloud

Speaker Zou Peng, senior engineer at Tencent Cloud, responsible for Redis product, shares the latest best practices for Redis 4.0 cluster on Tencent Cloud.

He first explains the mission of Redis, its popularity as a high‑performance, low‑latency, in‑memory data store that has become the de‑facto database for the mobile‑Internet era, surpassing Elasticsearch in ranking and capturing 65% of traffic from mainland China.

The talk then describes the architecture of the official Redis Cluster and the enhancements made by Tencent Cloud. The design includes a smart proxy, an autonomous management plane based on gossip protocol, and a load‑balancer that provides a single VIP for developers. The proxy handles sharding, migration commands, hot‑key detection, traffic isolation and custom monitoring.

Key challenges such as data migration, scaling (horizontal and vertical), data skew, large‑key handling and seamless expansion without service interruption are addressed. Tencent Cloud’s solution uses an asynchronous lazy scan for big keys, a proxy‑based migration that hides the process from the application, and automatic fail‑over with voting mechanisms.

He also introduces the self‑developed CKV engine, a Redis‑compatible protocol implementation launched in early 2018, which avoids the cost of a dedicated proxy by allowing each node to forward requests based on global slot information and multithreaded network handling, achieving up to 300 k QPS per node.

The session includes a Q&A where audience members ask about Redis vs MySQL usage ratios, high‑throughput designs, and the feasibility of using DPDK or OS‑less networking. The answers highlight the trade‑offs of custom network stacks and reaffirm the importance of sharding, replica scaling and proxy‑based monitoring.

Overall, the presentation provides practical guidance on deploying, operating and optimizing Redis clusters in a cloud environment, covering performance tuning, monitoring, high availability across availability zones, and cost‑effective architectural choices.

Original Source

Signed-in readers can open the original source through BestHub's protected redirect.

Sign in to view source
Republication Notice

This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactadmin@besthub.devand we will review it promptly.

Data MigrationperformanceProxydatabaseredisClustercloud
Tencent Cloud Developer
Written by

Tencent Cloud Developer

Official Tencent Cloud community account that brings together developers, shares practical tech insights, and fosters an influential tech exchange community.

0 followers
Reader feedback

How this landed with the community

Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.