DiDi's HUATUO project presents a kernel observability framework for agent sandboxes using eBPF and dynamic tracing to capture system-wide metrics, exceptions, auto-tracing, continuous profiling, and hardware faults, correlating kernel events with container identities for root-cause analysis in cloud-native and AI infrastructure.
The article explains how JDK Flight Recorder (JFR) provides low‑overhead, continuous JVM event recording that outperforms traditional snapshot tools like top, jstack, and jmap, and offers practical commands and use‑cases for diagnosing CPU spikes, GC issues, thread contention, and latency problems in production Java services.
Vtrace’s new continuous profiling feature fills previous Trace blind spots by sampling method‑level execution via JMX, delivering detailed intra‑service insights comparable to leading APM tools while keeping CPU overhead under five percent even at high transaction rates.
This article explains the evolution of observability, outlines a step‑by‑step diagnosis workflow using metrics, logs and tracing, highlights the blind spots of traditional tracing, and demonstrates how Alibaba Cloud ARMS continuous profiling and code‑hotspot features can pinpoint slow call‑chain issues in Java applications.
This guide explains how to install and configure the open‑source Pyroscope continuous profiler on a Kubernetes cluster, integrate it with Python, .NET, and Go microservices, and visualize low‑overhead flame graphs to identify performance bottlenecks and reduce cloud costs.
This article explains why traditional log‑based performance debugging is inefficient, introduces continuous profiling with the open‑source Pyroscope tool, and provides step‑by‑step instructions for installing Pyroscope on Kubernetes, integrating it with Python, .NET, and Go microservices, and visualizing low‑overhead flame graphs to optimize code and reduce cloud costs.
This guide explains how to use the open‑source Pyroscope continuous profiler on Kubernetes to collect, store, and visualize performance data from microservices written in Go, Python, .NET, and more, enabling developers to pinpoint bottlenecks, reduce CPU and memory usage, and lower cloud costs.
Developers can identify performance bottlenecks in production microservices by leveraging continuous profiling tools like Pyroscope, which collects, stores, and visualizes CPU and memory usage across Kubernetes clusters, and the article provides a detailed walkthrough of installation, configuration, and integration with various language services.
Grafana Phlare is an open‑source, horizontally scalable continuous profiling database that integrates with Grafana, offering easy installation, multi‑tenant support, and object‑storage‑backed long‑term storage, with detailed deployment instructions for both monolithic and micro‑service modes on Kubernetes using Helm.
Grafana Phlare is an open‑source, horizontally scalable continuous profiling database that integrates with Grafana, offering easy installation, multi‑tenant support, cheap object‑storage persistence, and both monolithic and microservice deployment modes, with detailed Helm‑based Kubernetes setup and usage instructions.
This guide explains how to integrate the open‑source Pyroscope platform with Rainbond to continuously profile microservices, detect CPU bottlenecks and memory leaks, and visualize performance data through its UI and various view modes.
Learn how to use the open‑source Pyroscope continuous profiler to identify performance bottlenecks in Kubernetes‑deployed microservices, compare it with other profilers, and follow detailed installation, configuration, and integration steps for Python, Go, and .NET services, including Helm deployment and resource monitoring.
