The article explains how to combine Claude Code's Skill framework with the Model Context Protocol (MCP) to let AI automatically fetch logs, extract key information, cross‑reference code, and pinpoint root‑cause bugs in a single command, illustrated by a hidden SQL bug case and efficiency tips.
When production logs grow to gigabytes, opening them with editors like vi or cat can crash your system; this article systematically presents essential Linux commands—grep, awk, sed, tail, less, and their pipelines—showing how to efficiently search, filter, and analyze massive logs for rapid backend issue resolution.
This guide shows practical Linux commands—tail, less, grep, sed, and awk—illustrated with real‑world scenarios to help backend developers quickly monitor, search, filter, and extract log data during production incidents.
An unexpected ‘Broken pipe’ error halted production, prompting a deep dive into logs, trace IDs, and monitoring tools like Kibana and SkyWalking, which uncovered Feign client timeouts, a costly SQL update, and redundant microservice calls, leading to targeted fixes that restored normal device operation.
An Nginx‑based data collection gateway began crashing with OOM kills, prompting a detailed investigation that uncovered memory spikes caused by aggressive batch processing, oversized protobuf payloads, and insufficient memory‑pool management, leading to a custom core‑dump solution and several mitigation strategies.
A special request caused memory usage to spike dramatically, leading to an OOM‑killer termination of an Nginx‑based data‑collection gateway, and the investigation uncovered a compression‑bomb style payload and coarse memory‑pool allocation as the root causes.
A production incident where a poorly implemented lazy‑loading mechanism for KMSClient caused repeated initialization, blocking threads, exhausting the shared thread pool, and triggering RejectedExecutionException alerts, was investigated step‑by‑step, leading to a concrete code fix, improved monitoring, and better thread‑pool isolation.
This article walks you through using the open‑source Java diagnostic tool Arthas to locate and analyze performance problems in Java applications, covering basic and advanced tracing, filtering, and flame‑graph visualization without modifying source code.
The article recounts a developer's experience with FastJson, detailing a date‑format configuration bug caused by an outdated library version, a $ref circular‑reference serialization issue, the debugging steps taken, and broader observations about FastJson's maintenance and feature defaults.
This article explains the meaning of HTTP 502 Bad Gateway errors, how reverse proxies like nginx generate them, common causes such as premature connection termination, server crashes, misconfigured upstreams, and practical steps to diagnose and fix the problem.
This guide introduces Alibaba's open‑source Arthas tool for Java, explains when to use it, shows how to install it, and provides detailed examples of its most useful commands such as stack, jad, sc, watch, trace, jobs, logger, dashboard, and redefine for live JVM debugging and performance analysis.
After taking over a system via a server image, the author uncovered malicious modifications that used a ternary expression to force id=1, causing MyBatis to generate a “where 1=1” clause that deleted the entire T_QUART_DATA table, and later discovered cron jobs that removed core jars, illustrating a thorough forensic debugging process.
After a late‑night incident where a Dubbo 2.7.12 service crashed, the author traced high memory and CPU usage to a full GC spike, identified a HashedWheelTimer thread‑pool bug causing request timeouts to be missed, reproduced the leak, and confirmed the issue was fixed in Dubbo 2.7.13.
A recent P2P box SDK upgrade introduced video screen‑tearing and gray‑screen issues for app users, which were traced to a faulty C++ vector initialization that caused multiple data buffers to share the same memory, and resolved by rolling back the SDK.
This article details ICBC's development of an online diagnostic platform using Arthas, covering the challenges of traditional Java debugging, the architectural design of the platform, gateway, and diagnostic process, as well as real‑world usage results and future outlook.
This article details a production incident where a Go backend service returned "connection reset by peer" due to exhausted process file descriptors caused by a saturated database connection pool, and describes the step‑by‑step troubleshooting, socket internals, and the eventual fix.
Rdebug, Didi’s open‑source real‑debugger, records live traffic and replays it locally, letting developers debug, test and develop micro‑service backends without deploying downstream dependencies, thereby boosting efficiency, cutting testing costs, and reducing production incidents.
The article analyzes a production incident where a Java web application using Apollo Config Center triggered "Too many open files" and NoClassDefFoundError errors after a configuration push, tracing the root cause to low OS file‑descriptor limits and Tomcat's class‑loading behavior.
