Nearly three months after the previous 2.4.1 release, OpenZFS, an open-source file system and volume manager with advanced data protection features like snapshots, checksums, and replication, has rolled out version 2.4.2 as the second maintenance release in the 2.4 series, offering compatibility with Linux kernels 4.18 through 7.0 and FreeBSD versions 13.3 and 14.0 or later.
For users running ZFS on newer Linux systems, updates include improvements to the fs_context-based mount API, mount option handling, lease handlers, ACL-related kernel changes, and block queue API changes.
In addition to kernel compatibility, OpenZFS 2.4.2 delivers several storage reliability fixes. These address rare checksum errors after rebuilds, dRAID checksum issues with degraded disks, and data corruption after clearing a disk in dRAID setups. The release also resolves a deadlock in vdev_rebuild() and an import failure that could occur after disk replacements in dRAID pools.
The release also resolves a read corruption issue that could occur after block cloning followed by truncation. Snapshot and mount handling have been improved as well, with OpenZFS 2.4.2 addressing a snapshot automount deadlock during concurrent zfs recv operations, an options memory leak in zfsctl_snapshot_mount, and an s_active leak in zfsvfs_hold() when a filesystem is already unmounted.
Additional changes include support for POSIX_FADV_DONTNEED, improved handling of POSIX_FADV_DONTNEED on single-block files, allocation class selection cleanup, memory leak fixes, build enhancements, and expanded CI coverage for newer Fedora and FreeBSD releases.
For more details, see the changelog.
