The article chronicles how 58.com’s web architecture evolved from a tiny, single‑machine setup to a multi‑layer, Java‑based, highly available service‑oriented platform, detailing the technical decisions, scaling challenges, and automation practices adopted at each traffic milestone.
This article outlines the UIOC (Urgent Incident Office Center) process, detailing a six‑step workflow, multi‑team collaboration tactics, common troubleshooting methods, and pitfalls to avoid, helping operations teams achieve 99.99% availability while handling emergencies efficiently.
Tencent’s operations team explains how its SET‑based NoSQL clusters deliver ultra‑low latency, high availability, and seamless disaster recovery for billions of users, detailing deployment models, synchronization mechanisms, cost‑saving techniques, and the Data‑as‑Service approach that underpins its massive social platforms.
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 details the evolution and redesign of JD.com’s product detail page architecture, describing the transition from static HTML generation to a dynamic, high‑performance, multi‑datacenter system built on key‑value storage, Nginx + Lua, asynchronous processing, multi‑level caching, and robust scaling and reliability strategies.
This article describes how JD.com redesigned its product detail page system from a static, cache‑heavy architecture to a fully dynamic, multi‑level cached service using Nginx+Lua, JIMDB, and asynchronous workers, addressing scalability, performance, and high‑availability challenges for billions of daily page views.
The article analyzes the use of a master/slave architecture for distributed caching, explains why two clusters, CAS, and proxy are employed, discusses consistency and availability challenges, and evaluates possible mitigation strategies for cache failures.
DRBD (Distributed Replicated Block Device) is a software‑based, network‑driven block replication solution that mirrors disks, partitions, or logical volumes across servers in real time, offering synchronous and asynchronous modes, transparent failover, and a middle‑layer between the filesystem and physical storage.
The article explains how distributed systems split a task into multiple subtasks across separate servers to reduce execution time, while clusters replicate the same service on many servers to increase throughput, highlighting their complementary roles in scalability and high availability.
This article explains how Keepalived automatically monitors web servers at the IP, TCP, and application layers, removes unhealthy nodes, and reintegrates them once recovered, thereby maintaining high availability and reducing manual maintenance in complex server clusters.
When read traffic overwhelms a single MySQL master, cascading replication and dual-master setups can distribute slave load, reduce replication lag, and simplify failover, offering a scalable solution that balances read/write pressure across multiple replication layers.
An in‑depth look at Weibo’s evolution to its third‑generation backend system, detailing the orthogonal decomposition model, three‑tier horizontal layering, key middleware such as MCQ, Motan RPC, SSDCache, and the WatchMan tracing platform that together enable high‑availability, massive concurrency, and low‑latency services for billions of users.
This article presents Sogou's design for a highly available, high‑performance, and horizontally scalable business platform, illustrating key architectural components and scalability strategies through a series of diagrams.
This article compiles recent high‑availability architecture discussions covering plugin‑based system design, simple kanban tools, independent scheduling services, challenges of managing 60‑billion‑row tables, and the trade‑offs between UUIDs and custom distributed ID generators, offering practical insights for backend engineers.
This article introduces MySQL Cluster’s share‑nothing architecture, explains the NDB storage engine’s advantages and drawbacks, describes the management, data, and API node roles, and provides a step‑by‑step configuration and deployment tutorial with command‑line examples.
This article explains the motivations behind key‑value stores, compares Redis with the newly created QDB, and details the architecture, dynamic resharding, high‑availability mechanisms, and future improvements of the distributed database solution RebornDB.
The article details Uber senior engineer Zhao Lei’s presentation on building a highly available messaging platform, covering single‑point failure mitigation, sharding approaches, large‑scale outage handling, cross‑region failover, and the practical engineering practices and protocols used to keep billions of users online.
Keepalived, built on LVS, automatically monitors service nodes, isolates failed machines, and performs seamless failover between load balancers using VRRPv2, allowing web servers to be added or removed without manual intervention, thus ensuring high availability and efficient load distribution.
This article presents an in‑depth overview of Codis and its next‑generation project RebornDB, covering Redis, Redis Cluster, the proxy‑based architecture, consistency trade‑offs, production deployment experiences, operational pitfalls, and broader perspectives on distributed databases and architectures.
This article examines WhatsApp’s high‑reliability architecture that supports 450 million users, detailing its Erlang‑based backend, hardware choices, scaling techniques, performance metrics, monitoring tools, and lessons learned from achieving up to two million concurrent connections on a single server.
The article explains why Kafka introduced high availability starting with version 0.8, detailing the need for data replication and leader election, describing replica distribution algorithms, replication mechanics, ISR handling, ZooKeeper structures, and the broker failover process to ensure fault‑tolerant streaming.
The 2015 Beijing QCon highlighted four technical talks covering asynchronous processing in distributed systems, the DRC data‑replication infrastructure, minute‑level high‑availability fault recovery, and cloud‑era database operations, illustrating Alibaba's approaches to scalability and reliability in modern cloud platforms.
This article examines common NoSQL key‑value stores such as Memcached and Redis, compares their strengths and limitations, and proposes a distributed architecture with routing, storage, management, and migration nodes that achieves high availability, automatic fault‑tolerance, load balancing, and elastic scaling.
This article explains Hadoop 2.0's High Availability architecture, comparing the NFS and Quorum Journal Manager methods, detailing their principles, failover mechanisms, and practical tips for reliable NameNode redundancy in big‑data deployments.
This article compares Redis with MySQL, outlines their similarities and differences, examines Redis alongside Memcached, EhCache, and OSCache, and proposes a simple high‑availability architecture for Redis, highlighting performance, data model, scalability, and operational considerations.
This article explains what a Linux cluster is, contrasts it with distributed systems, outlines its two main characteristics—scalability and high availability—along with essential capabilities like load balancing and error recovery, and details common cluster types such as high‑availability, load‑balancing, and high‑performance computing clusters.
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.
At Baidu’s 49th Technical Salon, Cheng Yishi explained how the company revamped its MooseFS and Redis systems by adding a Shadow Master to split reads from writes, introducing Slave nodes for failover, and deploying a Redis proxy middleware, thereby dramatically improving performance, scalability, and high‑availability for critical services.
