Focused on system technology, sharing cutting‑edge developments, innovation and practice, and analysis of industry tech hotspots.
This article explains the meaning of JVM Non‑Heap metrics, clarifies why Direct/Mapped memory is not part of Non‑Heap, analyzes a Flink taskmanager OOM case, and demonstrates how Native Memory Tracking (NMT) can reveal hidden memory gaps caused by allocator strategies such as TCMalloc and PTMalloc.
This article reviews the fundamentals of token‑bucket rate limiters, identifies precision, cascading compensation, and TCP‑loss sensitivity issues, and details two major improvements—port‑loan backpressure and the Carousel algorithm—while outlining future directions for more reliable network traffic shaping.
This article explains ByteDance’s end‑to‑end chaos engineering practice for cloud‑native environments, covering its background, principles, comparison with traditional testing, the evolution of its internal platforms, and a detailed look at the Application Resilience Enhancement Service (ARES) and its core features.
This article dives deep into the OpenJDK G1 FullGC implementation, explaining its architecture, trigger conditions, preparation steps, and the four‑phase collection process while comparing single‑threaded and multi‑threaded implementations in JDK 8 and JDK 11, and highlighting code paths that can cause long GC pauses.
This article explains common Linux memory‑related performance bottlenecks—such as memory reclamation, page‑cache pressure, huge‑page usage, and cross‑NUMA access—and provides practical tuning methods to improve latency and throughput on servers and applications.
DPDK’s libgraph, inspired by VPP’s vector packet processing, replaces traditional scalar packet handling with a graph‑based pipeline that improves i‑cache locality, reduces pointer copies, and enables flexible, high‑throughput networking across multiple CPUs and NICs, while introducing trade‑offs such as memory safety risks.
This article, based on ByteDance STE’s 2022 Linux Kernel Developer Conference talk, explains the HVO (HugeTLB vmemmap Optimization) feature that dramatically reduces struct page memory for huge pages, details its latest enhancements, cross‑architecture support, usage steps, performance trade‑offs, and future plans.
This article explains ByteDance's large‑scale adoption of profile‑guided optimization (PGO), covering its principles, instrumentation and sampling methods, the automated platform built for data collection and compilation, and the resulting performance gains across dozens of critical services.
This article explains Linux's Pressure Stall Information (PSI) mechanism, its /proc and cgroup interfaces, how to monitor CPU, memory, and I/O pressure, and presents code‑level optimizations to reduce PSI overhead and improve system performance.
This article details the exploration and implementation of macOS virtualization on Apple’s ARM‑based M1 Mac mini, covering the use of AsahiLinux, QEMU, KVM, device modeling, memory layout, boot process, and future optimization plans for a functional cloud‑native macOS environment.
