This article details a backend redesign of an instant‑messaging platform, covering Java interfaces for unread messages, Redis‑based unread counters, persistent queue processing, MySQL sharding and partitioning, SQL retrieval logic, and performance tuning to achieve high insert throughput.
This article explores the essential design considerations for building a short URL service, covering data structures, encoding algorithms, key length choices, capacity planning, sharding strategies, concurrency handling, network architecture, security measures, and a real‑world example.
Designing a short URL service involves choosing appropriate key‑value storage, simple incremental or base‑36 encoding for keys, estimating data capacity and sharding strategies, handling concurrent reads/writes with thread‑safe structures or Redis, selecting network event loops, and addressing security concerns such as abuse prevention.
The article describes how the limitations of an in‑memory NoSQL HOLD platform led to the creation of TDSQL, a distributed MySQL‑based SQL system featuring ZooKeeper‑coordinated Scheduler, Agent, and Gateway modules, automatic cross‑IDC disaster recovery, transparent horizontal scaling, strong synchronous replication, and future integration with container technologies.
This article explains how TDSQL transforms MySQL into a distributed, high‑availability SQL system by addressing NoSQL limitations, introducing a Scheduler‑Agent‑Gateway architecture, automatic scaling, sharding, robust disaster‑recovery mechanisms, and future integration with container technologies.
This article explains how scalability ties to concurrency and walks through architectural evolution—from a single MySQL instance to vertical partitioning, master‑slave replication, horizontal sharding, and finally cloud‑native SaaS databases—highlighting key challenges and solutions.
This article shares practical Elasticsearch experience covering index and query performance optimization, shard routing strategies, JVM tuning, daily maintenance, and answers to common production questions, providing actionable guidance for building high‑availability search clusters.
MyCat is an open‑source distributed database middleware that transparently shreds tables across multiple backend databases, solves connection overload, provides ER‑based sharding, global partitioning, AI‑driven catlets, and advanced read‑write separation, enabling low‑cost migration of single‑node databases to the cloud.
This article compiles expert insights from a high‑efficiency operations forum, covering MySQL performance optimization techniques such as read‑write separation, sharding, schema and parameter tuning, hardware and SSD enhancements, as well as a detailed FAQ addressing replication lag, Aurora challenges, and backup tools.
The article describes the evolution, architecture, automatic scaling, sharding, disaster‑recovery mechanisms, and strong synchronous replication strategy of TDSQL, a MySQL‑engine based distributed database built to meet the high‑availability and high‑consistency requirements of Tencent's billing platform.
This article presents 58.com’s experience with MySQL under massive data volumes, covering core concepts such as single‑instance, sharding, replication and grouping, common availability and read‑write challenges, detailed sharding implementations for user, post, friend and order tables, and post‑sharding business practices including IN queries, non‑partition key queries, and cross‑database pagination.
MongoDB 3.0 introduces a pluggable storage engine API with WiredTiger support, expands backup set capacity, adds sharding enhancements, improves security with SCRAM‑SHA‑1 authentication, and upgrades tooling and logging, while retaining MMAPv1 as the default engine.
The article explains how to achieve horizontal scalability by making applications stateless, using client‑side cookies for session data, applying various caching layers, splitting services and databases with sharding, adopting asynchronous messaging, storing unstructured data, and integrating monitoring with alerting.
This article outlines eBay's practical scalability best practices—including functional partitioning, horizontal sharding, avoiding distributed transactions, asynchronous design, virtualization, and judicious caching—to help large‑scale web systems achieve reliable, cost‑effective growth.
This article compiles practical discussions on database sharding and partitioning, covering criteria such as user ID, time, and hash, redundancy techniques, caching with Redis, query handling, pagination, consistency, and distributed transaction approaches for scalable backend systems.
This comprehensive tutorial walks you through Redis Cluster fundamentals, including installation, hash slot sharding, node replication, fault tolerance, configuration files, cluster creation commands, and client testing, all without delving into complex distributed theory.
This article traces MySQL’s architectural evolution—from a single‑instance setup to vertical partitioning, master‑slave replication, horizontal sharding, and finally cloud‑native SaaS—highlighting scalability strategies, sharding routing methods, consistency challenges, and automated scaling techniques for high‑concurrency web applications.
The article introduces SessionDB, a Java‑implemented, LSM‑tree‑based key/value storage engine designed to eliminate sticky sessions by providing high‑throughput, durable, and scalable session data handling, and details its architecture, optimizations, sharding strategy, and benchmark comparisons with BerkeleyDB, LevelDB, and RocksDB.
