OpenZFS, the project that develops the implementation of the ZFS file system for Linux and FreeBSD, has recently released version 2.2.0 after more than two years of development. Previously known as “ZFS on Linux,” the project has gained recognition as the main implementation of OpenZFS and has expanded its support to FreeBSD.
The OpenZFS project has been tested with various Linux cores from 3.10 to 6.5 and all FreeBSD branches starting from 12.2-Release. The code is distributed under the free CDDL license and is already being used in FreeBSD, as well as included in major Linux distributions such as Debian, Ubuntu, Gentoo, Sabayon Linux, and Alt Linux. Packages with the new version will soon be available for other popular Linux distributions such as Debian, Ubuntu, Fedora, and RHEL/CentOS.
OpenZFS provides the implementation of ZFS components related to both the file system and the volume manager. This includes components such as SPA (Storage Pool Allocator), DMU (Data Management Unit), Zvol (ZFS Emulated Volume), and ZPL (ZFS Posix Layer). Additionally, the project also has the capability to use ZFS as a backend for the Lustre cluster file system. The project builds upon the original ZFS code imported from the Openslaris project and incorporates improvements and corrections from the Illumos community. The development of OpenZFS is carried out by employees of the Liverrand National Laboratory under a contract with the US Department of Energy.
Due to licensing incompatibility, OpenZFS cannot be integrated into the main branch of the Linux kernel, as it is distributed under the CDDL license which is incompatible with GPLV2. To address this, OpenZFS is distributed as a separately loaded module, which is supplied separately from the Linux kernel. Nevertheless, the stability of the OpenZFS code base is said to be comparable to other file systems for Linux.
The latest release of OpenZFS, version 2.2.0, brings several notable changes. Firstly, a block cloning mechanism has been added that allows for the creation of file copies or parts without duplicating data by utilizing links to existing data blocks. This mechanism, known as Copy-on-Write mode at the file level, ensures that actual copying of blocks only occurs when there are amendments to the source file or its copies. Additionally, support has been added for container isolation technologies in Linux, including the system call renameat, fs overlayfs, mapping user identifiers during mounting, and delegation namespaces for containers.
For more information on the release and to access the