What is it about?
This paper evaluates the latencies introduced by two commonly-used open source hypervisors. Introducing low latencies is a very important requirement when serving real-time applications, and it is commonly believed that real-time hypervisors are needed to achieve this result. However, this paper shows that even KVM and Xen, if properly configured, can provide low latencies compatible with many real-time applications. The Xen hypervisor, which originally introduced higher latencies related to timers, has been fixed (and the fix has been accepted in mainline) to support real-time applications.
Featured Image
Why is it important?
This paper is important because it shows that in many situations, real-time applications can be executed in a VM even without using a specifically-designed real-time hypervisor. Of course, the system must be properly configured, and the paper provides a detailed discussion about the most important configuration options and the best choices.
Read the Original
This page is a summary of: Using Xen and KVM as real-time hypervisors, Journal of Systems Architecture, June 2020, Elsevier,
DOI: 10.1016/j.sysarc.2020.101709.
You can read the full text:
Contributors
The following have contributed to this page







