An internal sandbox API kept hanging for a week, prompting a deep dive with top, jstack, and Arthas that revealed Redis connection pool threads stuck in wait, leading to configuration tweaks and proper connection release to restore service stability.
A week‑long API freeze in a sandbox environment was traced to a misconfigured Redis connection pool in a Spring Java application, where threads waited indefinitely for a Jedis resource, and the fix involved setting proper pool timeouts, using RedisCallback for scans, and correctly releasing connections.
The article explains how automating thread dump creation and download in Kubernetes using tools like Fabric8, Prometheus, and CI/CD pipelines dramatically improves fault‑diagnosis speed, data centralization, real‑time capture, and integration with testing frameworks, transforming manual, error‑prone processes into streamlined, intelligent operations.
This article explains what JVM thread dumps are, why they are crucial for fault testing, outlines common scenarios such as deadlocks and resource leaks, and provides step‑by‑step methods—including jstack, kill‑3, VisualVM, programmatic APIs, and Fabric8‑driven Kubernetes automation—to capture and analyze them effectively.
Thread dumps provide a snapshot of all active JVM threads, including their state, stack trace, and metadata, enabling developers to diagnose performance issues such as slow responses, high CPU usage, deadlocks, and thread pool inefficiencies by examining key fields and applying practical analysis techniques.
This guide introduces Java performance analysis fundamentals, covering profiling tools, key metrics such as CPU, memory, thread behavior, and practical methods to detect and resolve common issues like memory leaks, GC pauses, high CPU usage, I/O bottlenecks, and inefficient collection usage.
This guide walks through real‑world scenarios of high CPU and memory alerts, demonstrating how to use Linux's top tool, interpret its detailed output, convert thread IDs, and leverage jstack dumps to pinpoint and resolve performance bottlenecks.
This article explores why using an Integer as a lock can cause synchronized to malfunction, demonstrates the issue with a ticket‑selling example, analyzes thread dumps and JVM bytecode, and presents reliable alternatives such as class‑level locks or map‑based lock objects.
This article explains how to diagnose and resolve severe CPU usage spikes—up to 900%—in MySQL and Java processes, detailing step‑by‑step identification, common causes such as missing indexes or excessive garbage collection, and practical remediation techniques including query optimization, indexing, caching, and thread analysis.
After a new feature rollout caused a sudden CPU surge and thousands of Dubbo threads to stall, this article walks through log inspection, code review, thread‑dump analysis, and memory‑leak detection to uncover the root cause and suggest concrete optimizations.
A week‑long API freeze in a sandbox environment was traced to Redis connection pool deadlocks caused by missing pool configuration and improper connection handling, and the article walks through the diagnosis with top, jstack, Arthas, and the code changes needed to fix it.
This article introduces the jps tool for listing Java processes, explains its common options and usage examples, then provides a detailed overview of jstack, including its purpose, command syntax, thread‑state definitions, sample outputs, and practical tips for detecting deadlocks, counting threads, and diagnosing high CPU usage in Java applications.
A detailed investigation of a slow Elasticsearch cluster revealed that massive Stop‑The‑World (STW) safepoint pauses caused by a buggy ARM‑based JDK version were the root cause, and switching to a proper Kona JDK eliminated the frequent ForceSafepoint interruptions and restored performance.
The article details how a SaaS application's intermittent slowdown was traced to a Log4j logger lock contention issue, discovered via thread dumps and an online analysis tool, and resolved by migrating from the unmaintained Log4j 1.x to Log4j2.
An intermittent slowdown in a Java SaaS app, even during low traffic, was traced to a Log4j locking bug that blocked hundreds of threads; capturing a thread dump, analyzing it with fastthread.io, and upgrading to Log4j2 resolved the issue.
The article outlines a six‑step method for diagnosing sporadic JVM hangs in a high‑concurrency Xianyu service—starting with code review, live state capture, I/O checks, lock analysis, resource‑exhaustion assessment, and finally framework thread‑pool tuning—to uncover lost‑lock behavior and severe thread‑pool imbalance that cause prolonged lock waits despite low CPU load.
This guide walks through the most frequent Java production problems—high CPU, high load, continuous Full GC, thread‑pool saturation, and NoSuchMethodException—explaining their causes, diagnostic commands, and practical remediation steps to restore service stability.
A recent offline performance test of a Dubbo service revealed that enabling Logback’s default packageData option caused massive AladdinClassLoader lock contention, dramatically increasing response times, and the issue was fixed by upgrading Logback to a version where packageData is disabled by default.
This article explains how to identify and resolve common causes of Java application performance degradation—including excessive CPU usage, frequent Full GC events, thread blocking, and deadlocks—by using Linux tools such as top, jstat, jstack, and memory‑dump analysis with Eclipse MAT.
This guide explains how to identify and resolve common Java production problems such as sudden CPU spikes, excessive Full GC, thread blocking, waiting states, and deadlocks by using tools like top, jstack, jstat, and memory‑dump analysis to pinpoint the root cause and apply appropriate fixes.
This article walks through a practical Linux performance investigation, showing how to use the top command to pinpoint high‑CPU processes, examine thread details, convert thread IDs, analyze thread dumps for lock contention, and interpret key top output fields for effective troubleshooting.
The article recounts a step‑by‑step investigation of an intermittent Java service freeze caused by blocked threads, excessive Metaspace usage, and conflicting JVM flags, showing how thread dumps, GC statistics, and startup‑parameter tweaks revealed a hidden HotSpot bug that was resolved by upgrading the JDK.
