Book Image

KVM Virtualization Cookbook

By : Konstantin Ivanov
Book Image

KVM Virtualization Cookbook

By: Konstantin Ivanov

Overview of this book

Virtualization technologies such as KVM allow for better control over the available server resources, by deploying multiple virtual instances on the same physical host, or clusters of compute resources. With KVM it is possible to run various workloads in isolation with the hypervisor layer providing better tenant isolation and higher degree of security. This book will provide a deep dive into deploying KVM virtual machines using qemu and libvirt and will demonstrate practical examples on how to run, scale, monitor, migrate and backup such instances. You will also discover real production ready recipes on deploying KVM instances with OpenStack and how to programatically manage the life cycle of KVM virtual machines using Python. You will learn numerous tips and techniques which will help you deploy & plan the KVM infrastructure. Next, you will be introduced to the working of libvirt libraries and the iPython development environment. Finally, you will be able to tune your Linux kernel for high throughput and better performance. By the end of this book, you will gain all the knowledge needed to be an expert in working with the KVM virtualization infrastructure.
Table of Contents (15 chapters)
Title Page
About the Author
About the Reviewer
Customer Feedback

NUMA tuning with libvirt

NUMA is a technology that allows the system memory to be divided into zones, also named nodes. The NUMA nodes are then allocated to particular CPUs or sockets. In contrast to the traditional monolithic memory approach, where each CPU/core can access all the memory regardless of its locality, usually resulting in larger latencies, NUMA bound processes can access memory that is local to the CPU they are being executed on. In most cases, this is much faster than the memory connected to the remote CPUs on the system.

Libvirt uses the libnuma library to enable NUMA functionality for virtual machines, as we can see here:

root@kvm:~# ldd /usr/sbin/libvirtd | grep numa => /usr/lib/x86_64-linux-gnu/ (0x00007fd12d49e000)

Libvirt NUMA supports the following memory allocation policies to place virtual machines to NUMA nodes:

  • strict: The placement will fail if the memory cannot be allocated on the target node
  • interleave: Memory pages are allocated...