Is LXD, Canonical’s open source container hypervisor, still an important part of the container ecosystem? That’s a fair question to ask, because no one has heard much about it lately. However, the project continues to chug along, and may become more important in 2018.
What is LXD?
LXD is an open source container framework based on LXC. To understand what it is, you first have to understand LXC.
LXC is a container technology for the Linux kernel that predates Docker. LXC was designed to run applications or processes inside containerized environments.
LXC saw relatively little adoption, although some vendors including Heroku and Wercker used LXC for production purposes several years ago.
LXC assumed more significance in 2015, when Canonical, the company most famous for Ubuntu Linux, began marketing LXD. It uses the LXC framework to containerize entire operating systems, as opposed to individual applications.
Canonical promoted LXD as a very efficient way to run multiple virtual operating system environments on a single server. It can support many more guest environments on a single host than can full-virtualization hypervisors such as KVM and VMware, according to Canonical.
Today: Quiet but Active
LXD became ready for production use in 2016.
We took it for a spin in the same year and found that it does indeed live up to Canonical’s promises of higher performance than traditional hypervisors.
Since 2016, however, it has been very quiet. Canonical has issued no major announcements of late about the platform.
On the other hand, the LXD GitHub repository remains quite active. It seems that the project is still healthy and remains very much under active development.
2018: The Year of LXD?
It would appear, then, that Canonical remains committed to making LXD an important part of the container ecosystem. But the company is perhaps still deciding how best to promote it, or how to ensure it has a noticeable share of voice within a market where conversation is dominated by Docker and Kubernetes.
From end users’ perspective, LXD clearly has value. Although Docker containers can be used to host guest operating systems, Docker has never pursued that use case in a serious way. It remains the only container platform that caters to this need.
Will 2018 be the year when it finally takes off? Now seems as good a time as any for Canonical to begin promoting the framework more aggressively, and highlighting what it can do that Docker can’t.