Backend technology sharing, architecture design, performance optimization, source code reading, troubleshooting, and pitfall practices
This article explores a bug in an enhanced singleflight event‑merging system, illustrates problematic scenarios with overlapping events, and presents a time‑stamp based compensation mechanism that detects and re‑pushes inconsistent data to ensure correct final event states.
This article explains the Go singleflight package, its original implementation, and a custom WriteGroup enhancement that merges concurrent read and write events to dramatically improve registration center performance under high load.
This article walks through a hands‑on Go RPC demo, explains the underlying server and client mechanisms, explores registration, HTTP handling, request processing, object pooling, and the gob codec, and clarifies why Go includes RPC alongside HTTP.
This article walks through building a minimal Go HTTP server, explains how ListenAndServe manages connections, details the internal accept‑serve loop, shows request handling, routing rules, and customization hooks, providing clear code examples and diagrams to demystify Go’s built‑in web server.
This article explains how to batch multiple cluster endpoint requests in Go by using a token‑based queue and concurrent workers, dramatically reducing load on the registration center while requiring minimal changes to existing code.
This article details the redesign of a service health‑check component, covering its original limitations, industry alternatives, the chosen centralized active checking approach, architectural modules, concurrency model, scaling mechanisms, gray‑release strategy, and performance optimizations for reliable distributed systems.
This article compares Java’s HttpClient timeout implementation with Go’s built‑in HttpClient, explains Go’s Context‑based timeout mechanism, walks through the underlying source code, and shows why Java cannot easily replicate the same approach due to differences in concurrency primitives.
This article explains the differences between short and long connections, why long‑connection services need load balancing, and presents practical strategies for connection‑level balancing, handling server restarts, heterogeneous hardware, and scaling challenges.
Explore the complete lifecycle of graceful shutdown in Java backend systems, covering kernel signal handling, JVM shutdown hooks, Spring’s ContextClosedEvent, Dubbo’s integration and bug fix, and Netty’s Reactor shutdown process, with detailed code examples and state transition diagrams.
This article explains why payment orders can disappear after a successful wallet transaction, outlines the complete payment flow, distinguishes internal and external drop‑order scenarios, and provides practical server‑side and client‑side strategies—including retry mechanisms, reliable async messaging, scheduled queries, and delayed‑message approaches—to reliably prevent such issues.
