The second release of the Incus project has been introduced, in which the Linux Containers community develops a fork of the LXD container management system, created by the original team of developers who once built LXD. The Incus code is written in Go and is distributed under the Apache 2.0 license.
Recall that the Linux Containers community oversaw the development of LXD until Canonical decided to develop LXD separately as a corporate project. The goal of the fork is to provide a managed alternative to the LXD project, controlled by Canonical, by an independent community. The Incus project also aims to address some conceptual errors made in LXD's development, which could previously not be corrected without breaking backward compatibility.
Incus provides tools for centralized management of containers and virtual machines deployed on both a single host and a cluster of multiple hosts. serversThe project is implemented as a background process that accepts network requests through a REST API and supports various storage backends (directory trees, ZFS, Btrfs, LVM), stateful snapshots, live migration of running containers from one machine to another, and tools for storing container images. LXC is used as the runtime for launching containers, including the liblxc library, a set of utilities (lxc-create, lxc-start, lxc-stop, lxc-ls, etc.), templates for building containers, and a set of bindings for various programming languages. Isolation is achieved using standard Linux kernel mechanisms (namespaces, cgroups, AppArmor, SELinux, Seccomp).
The most notable changes:
- The lxd-to-incus utility, which automates the transition from LXD to Incus, has added support for LXD 5.19 and enables migration not only of individual hosts but also of LXD clusters.
- Virtual machines now support the creation of storage based on NVME technology. A new configuration parameter, 'io.bus', has been added to specify the disk type, defaulting to 'virtio-scsi'. Changing this value to 'nvme' makes the storage in the virtual machine appear as NVME SSD.
- For container images, the parameter 'requirements.privileged' has been implemented, which can be set to 'false' to prohibit using the image in privileged containers. This setting, for example, can be applied to images with the NixOS distribution, which is currently incompatible with privileged containers.
- The implementation of the section copy mode has been migrated from LXD, serverwhich significantly speeds up section copy operations by eliminating data redirection through the client.
- Statically compiled builds for ARM64 systems are now provided.
Source: opennet.ru
