Bpf Performance Tools (Addison-Wesley Professional Computing Series) Paperback – 25 February 2020

Lieferung in sehr gutem Zustand und mehr als pünktlich angekommen.Sehr gut

Not useful. Just a bunch of program listing and generic statments.

Este libro es una referencia indispensable si trabajas en IT y eres SysAdmin. No es requisito saber programar en el ambiente Linux pero es deseable tener fundamentos de arquitectura del kernel y haber programado en algún lenguage de bajo nivel para sacarle el máximo provecho a todo el material que se presenta.

Teaching and learning about operating system theory is one thing, while actually using an OS, verifying that it is doing what theory specifies, both safely and efficiently, is quite another. Understanding how a particular OS performs is a daunting task, especially in the current world of multicore processors, virtualization, and massive scale. Powerful and usable observability tools are needed to examine every aspect of OS behaviour. Such tools have evolved dramatically in the past two decades, driven by commercial efforts like DTrace for Solaris and its open-source adoption and adaptation to other operating sytems like FreeBSD, OS X, Linux, and even Windows(!).Brendan Gregg is a leader in OS performance theory and practice, well-known for his earlier work and publications on DTrace; his diagnostic and graphical tools are now essential for discovering bugs and bottlenecks in OS and application execution. His latest volume, BPF Performance Tools, in size and scope is a veritable encyclopedia of OS observability methods and programs for both the Linux kernel itself and its application execution environment. Exploiting his insights gained from DTrace, Gregg has extended and enhanced BPF (Berkeley Packet Filter) technology, creating a massive collection of probes and monitors used to dynamically trace the execution of compute, memory, network, and I/O functions.Linux system administrators, kernel hackers, security researchers, and application developers should thoroughly study and keep Gregg’s reference handy to both diagnose performance problems and to learn how to avoid them in the first place. Anyone teaching about operating systems, both in industry and in academia, should enhance their instruction with the invaluable practical advice and techniques in this thorough and comprehensive text. Gregg also provides additional material, source code, and examples on the book’s companion web site. And to help understand the observability purpose of each of the standard and BPF-specific Linux tools, Gregg includes helpful diagrams showing which parts of the kernel each tool addresses.

El contenido de maravilla, la versión impresa no está a color, y varios flamecharts y diagramas no se aprecian muy bien en escala de grises, por lo demás uno de los mejores libros sobre el tema.