0 views collected around this technical thread.
The article explains why employing write‑oriented models for data retrieval in PHP leads to security leaks, unnecessary validation, and performance issues, and proposes dedicated read models, static constructors, and specialized repositories to cleanly separate read and write concerns.
This article outlines the progressive evolution of database architectures—from simple monolithic designs to cold‑hot separation, read‑write splitting, and horizontal sharding—explaining each pattern, its components, and the performance and scalability benefits they bring to large‑scale systems.
This article explains the concept of high concurrency, its key metrics, and presents four practical backend solutions—static/dynamic separation with CDN, read‑write separation using NoSQL and CQRS, various caching techniques, and lightweight read‑optimisation—to improve performance and scalability of distributed systems.
This article presents nine comprehensive performance‑optimization strategies—including load balancing, sharding, read/write separation, caching, indexing, CDN, asynchronous processing, code refinement, and algorithm improvement—aimed at boosting the efficiency and scalability of large‑scale backend systems.
This article explains the essential conditions, metrics, and scenario classifications for building high‑concurrency systems, then details common solutions such as database read/write separation, local and distributed caching, cache‑eviction policies, handling master‑slave lag, preventing cache penetration and avalanche, and applying CQRS to achieve scalable, high‑performance back‑end services.
This article explains Redis master‑slave replication, covering read‑write separation, configuration steps for master and slave nodes, the PSYNC synchronization protocol, and how these mechanisms together provide high availability and consistent data across distributed Redis deployments.
This article explains how internet‑scale services can progressively upgrade their data architecture—from a single read/write instance to read‑write separation, multi‑region “two‑site three‑center” designs, cache fallback, and local decentralised caching—to achieve high availability and scalability.
This article explains the concept of database read‑write separation, why it improves performance in high‑concurrency scenarios, when to apply it, the underlying master‑slave replication mechanism, and provides a MySQL example illustrating the steps of replication and its benefits for scalability and availability.
This article explains the principles and practical designs of high‑performance database clusters, covering read/write separation, master‑slave replication lag, distribution mechanisms, sharding (both vertical and horizontal), business‑level partitioning, and the role of NoSQL technologies such as key‑value stores, document databases, columnar databases, and full‑text search engines.
This article explains MySQL master‑slave replication, covering its basic concepts, the advantages of read‑write separation, data backup and high availability, the underlying binlog‑based mechanism, the four replication modes (full, asynchronous, semi‑synchronous, enhanced semi‑synchronous), common consistency challenges, and practical ways to mitigate them.
This article explains how to achieve database read‑write separation in high‑concurrency Java applications by configuring master‑slave data sources, implementing a routing datasource, managing context with ThreadLocal, and using custom annotations with AOP to switch between read and write databases at runtime.
This article walks through the progressive evolution of a software system—from a traditional monolithic architecture to a task‑oriented monolith, then to Command‑Query Separation (CQS) and finally to a full CQRS solution—explaining the motivations, trade‑offs, and techniques such as eventual consistency, event sourcing, and read‑model generation.
This article details the end‑to‑end design of Baidu App’s personal wallet, covering background requirements, business integration, system architecture, data‑sync mechanisms, multi‑level caching, read‑write separation, data consistency, configuration management, and database sharding to achieve high availability and scalability.
The DBLE 3.22.01.0 release introduces seven new or refactored features—including enhanced read/write separation, ClickHouse support, and hint‑based query planning—while fixing over a hundred bugs, addressing critical community issues, and providing a complete list of changes and incompatibilities.
This article explains MySQL master‑slave read/write separation, the binlog and relay log mechanisms, master‑slave delay, and how different transaction isolation levels affect whether the master or slave may return stale data, illustrated with practical examples and SQL snippets.
The article explains how Redis read/write separation works, compares star and chain replication architectures, and highlights the benefits of transparent compatibility, high availability, and improved performance for read‑heavy, write‑light workloads.
This article explains why and how to apply vertical (business‑oriented) and horizontal (data‑oriented) sharding, covering single‑database, table splitting, database splitting, combined sharding, read‑write separation, routing algorithms, and real‑world case studies to improve scalability and performance.
The article explains how to transition from client‑side sharding with Sharding‑JDBC to proxy‑side sharding using Rainbow Bridge DAL by adding new, switchable data sources, employing AbstractRoutingDataSource and configuration flags for gradual migration, while addressing read‑write separation, unsupported SQL functions, and ensuring backward compatibility.
This article explains Alibaba Cloud's Redis read‑write separation architecture, comparing star and chain replication models, their performance and scalability trade‑offs, and how transparent compatibility, high availability, and high performance are achieved through redis‑proxy, HA monitoring, and optimized binlog replication.
Redis read/write separation introduces read‑only replicas to offload read traffic, offering two replication models—star replication with direct master‑replica sync and chain replication forming a replication chain—each with distinct performance, scalability, and latency trade‑offs, while maintaining high availability and transparent compatibility.