Rootconf 2015

DevOps and scaling infrastructure

Amit Shah


Live Migration of QEMU/KVM Virtual Machines

Submitted Mar 30, 2015

Live migrating virtual machines is one of the key benefits of virtualization. Knowing the constraints of networks, storage and hypervisors helps in designing and deploying workloads. This session wil help people understand what goes in migrating a virtual machine, and how they can help by getting the configuration right.


Live migrating vitual machines is an interesting ongoing area for virtualization: guests keep getting bigger (more vcpus, more vram), and demands on the uptime for guests keep getting stricter (so no long pauses between a VM migrating from one host to another).

This session will focus on QEMU/KVM as the hypervisor. QEMU/KVM is the default hypervisor for OpenStack and the only hypervisor supported in Red Hat Enterprise Linux 6 and 7, as well as the only hypervisor supported by oVirt and Red Hat Enterprise Virtualization.

This session will go through the simple design from the early days of QEMU/KVM live migration, how it has been tweaked to where it is now, and where we’re going in the future. It will discuss how live migration actually works, the constraints within which it all has to work, and how the design keeps needing new thought to cover the latest requirements.

This session will show what thought needs to be put in while deploying workloads on virtual machines, and designing the surrounding infrastructure for a private cloud.

Speaker bio

Amit has been working on FOSS since 2001, and QEMU/KVM virtualization since 2007. He’s currently employed by Red Hat. He’s worked on several areas within QEMU/KVM virtualization, and live migration is his current focus.

He has spoken extensively on several aspects of virtualization, at several avenues, including KVM Forums, and

Amit blogs at




{{ gettext('Login to leave a comment') }}

{{ gettext('Post a comment…') }}
{{ gettext('New comment') }}
{{ formTitle }}

{{ errorMsg }}

{{ gettext('No comments posted yet') }}