This article explains how to replace invasive manual timing code with Java Instrumentation, demonstrating both premain and agentmain approaches, building and attaching agents, using ASM and Bytekit for bytecode enhancement, and leveraging Arthas for runtime tracing and debugging.
The article introduces a lightweight bytecode‑enhancement framework that weaves interceptors directly into Java methods to record execution times and call hierarchies, avoiding proxy overhead, supporting synchronous, asynchronous and lambda scenarios, and enabling precise performance monitoring and tracing across micro‑service architectures.
This article examines Android Dex DebugInfo line‑number optimization, detailing the structure of DebugInfo, existing solutions, Baidu’s mapping‑based approach, R8 and Alipay techniques, and the resulting APK size reduction, while also describing the end‑to‑end pipeline for line‑number retracing in production.
This article introduces a comprehensive solution for diagnosing and attributing Java Out‑Of‑Memory (OOM) issues in Android apps by capturing Hprof memory snapshots, automatically analyzing heap data, identifying leaks, large objects, and class‑wide memory consumption, and providing privacy‑preserving, automated reporting and remediation workflows.
To overcome native APM’s inability to monitor Flutter pages, the team created a high‑availability, open‑source Flutter SDK that records smoothness, load time, and exceptions through a four‑layer, publish‑subscribe architecture—API, Recorder, Processor, and Uploader—offering accurate, low‑overhead, extensible performance data for integration and future enhancements.
