BPF Performance Tools

BPF Performance Tools provides a practical and comprehensive guide to using eBPF and modern observability tools to gain deep insight into Linux systems and applications. As software environments become increasingly complex, observability has become essential for understanding performance, troubleshooting issues, improving reliability, and strengthening security. This text introduces readers to the concepts, techniques, and tools that enable unprecedented visibility into system behaviour, helping developers, engineers, and administrators identify and solve performance challenges with confidence. Written by renowned performance expert Brendan Gregg, the text combines foundational knowledge of operating systems, kernel internals, and performance engineering with hands-on exploration of industry-leading BPF technologies. Readers learn how to use eBPF tracing and observability tools to analyse CPUs, memory, storage, file systems, networking, applications, containers, hypervisors, and the Linux kernel. Progressing from fundamental concepts to advanced analysis techniques, the text demonstrates how to generate custom metrics, capture stack traces, build latency histograms, and investigate real-world performance issues across modern computing environments. Featuring extensive coverage of open-source tools from the Linux Foundation's iovisor project, including bcc and bpftrace , it includes more than 150 ready-to-use tracing tools and examples that can be used immediately or adapted to meet specific requirements. By combining practical examples with powerful diagnostic methodologies, BPF Performance Tools equips readers with the skills needed to improve application performance, enhance system reliability, and gain deeper visibility into virtually every aspect of Linux-based systems.

BPF Performance Tools provides a practical and comprehensive guide to using eBPF and modern observability tools to gain deep insight into Linux systems and applications. As software environments become increasingly complex, observability has become essential for understanding performance, troubleshooting issues, improving reliability, and strengthening security. This text introduces readers to the concepts, techniques, and tools that enable unprecedented visibility into system behaviour, helping developers, engineers, and administrators identify and solve performance challenges with confidence. Written by renowned performance expert Brendan Gregg, the text combines foundational knowledge of operating systems, kernel internals, and performance engineering with hands-on exploration of industry-leading BPF technologies. Readers learn how to use eBPF tracing and observability tools to analyse CPUs, memory, storage, file systems, networking, applications, containers, hypervisors, and the Linux kernel. Progressing from fundamental concepts to advanced analysis techniques, the text demonstrates how to generate custom metrics, capture stack traces, build latency histograms, and investigate real-world performance issues across modern computing environments. Featuring extensive coverage of open-source tools from the Linux Foundation's iovisor project, including bcc and bpftrace , it includes more than 150 ready-to-use tracing tools and examples that can be used immediately or adapted to meet specific requirements. By combining practical examples with powerful diagnostic methodologies, BPF Performance Tools equips readers with the skills needed to improve application performance, enhance system reliability, and gain deeper visibility into virtually every aspect of Linux-based systems.