Bcachefs 1.37 Released With Stable Erasure Coding and Linux 7 Support

Bcachefs 1.37: A Major Leap in Modern Linux Filesystem Innovation

In the ever-evolving landscape of Linux filesystems, Kent Overstreet has unveiled Bcachefs 1.37, a cutting-edge, copy-on-write filesystem that is rapidly positioning itself as a formidable contender against established names like Btrfs and ZFS. This latest release is packed with groundbreaking features, performance enhancements, and a renewed focus on data integrity, making it a must-know for sysadmins, developers, and tech enthusiasts alike.

Erasure Coding: From Experimental to Enterprise-Ready

One of the most significant milestones in Bcachefs 1.37 is the graduation of erasure coding from its experimental phase. This feature, now considered stable, is fully integrated with the filesystem’s reconcile process, enabling automatic repair of degraded data. This means that if a drive fails or data becomes corrupted, Bcachefs can intelligently reconstruct the missing pieces using redundant information stored across multiple devices. The integration also supports tiered and mixed-device setups, allowing users to optimize storage for both performance and capacity. While stripe allocation improvements are still on the roadmap, the core functionality is now robust enough for production use.

Unprecedented Data Integrity and Recovery

Data integrity has received a major boost in this release. Journal rewind is now fully safe, with the filesystem meticulously tracking safe rollback points. This ensures that even if a crash occurs during a rewind operation, the process remains transactionally consistent and can be resumed without risk of data loss. Additionally, a new targeted recovery mechanism has been introduced. After an unclean shutdown, Bcachefs now scans recently written data for corruption and attempts immediate repair. If recovery isn’t possible, the filesystem automatically rewinds to the last known good state within a configurable time window, minimizing downtime and data loss.

Enhanced Resilience to Hardware Faults

In an era where hardware failures can be catastrophic, Bcachefs 1.37 introduces automatic recovery from devices with faulty flush or FUA (Force Unit Access) support. This means that even if a drive doesn’t properly acknowledge write commands, Bcachefs can detect and recover from potential silent data corruption, a common but often undetected issue. Recovery from unclean shutdowns is now faster and more predictable, giving administrators greater confidence in the reliability of their storage systems.

New Tools and Usability Improvements

Bcachefs 1.37 also brings a suite of new tooling features designed to make filesystem management more intuitive and powerful. Users can now list subvolumes with advanced filtering and sorting options, explore snapshot hierarchies with detailed disk usage information, and propagate file I/O options across reflinked data while maintaining strict ownership constraints. The fs top command now features a terminal user interface (TUI) with per-device statistics, offering real-time insights into filesystem performance. Tabular outputs have also been improved for better readability, making it easier to parse complex data at a glance.

Expanding Compatibility and Embracing Rust

This release expands compatibility by adding support for the Linux kernel 7.0, ensuring that Bcachefs remains at the forefront of kernel development. In a bold move toward modernization, the userspace component has been fully converted to Rust, a language renowned for its safety and performance. This shift not only reduces the likelihood of memory-related bugs but also aligns Bcachefs with contemporary software engineering practices.

Availability and Installation

Currently, Bcachefs tooling is officially packaged in a limited number of distributions, including Arch Linux, Gentoo, Void, and Fedora. For users of Debian and Ubuntu, Bcachefs is not included in the default repositories, but support can be added via the external upstream APT repository at apt.bcachefs.org. openSUSE provides Bcachefs through its Build Service, while NixOS includes only the userspace tools. Since Bcachefs is no longer part of the mainline kernel, all distributions rely on an out-of-tree module—typically DKMS—to enable filesystem support. Module availability may vary between systems, so users should consult their distribution’s documentation for specific instructions.

Looking Ahead

With Bcachefs 1.37, Kent Overstreet and the development team have delivered a release that not only enhances the filesystem’s core capabilities but also sets the stage for future innovations. The graduation of erasure coding, the focus on data integrity, and the embrace of modern tooling and languages signal a clear commitment to making Bcachefs a top-tier choice for Linux storage.

For those eager to dive deeper, the full changelog is available on the official repository.

Tags: Bcachefs, Linux filesystem, erasure coding, data integrity, snapshots, compression, encryption, Rust, kernel 7.0, DKMS, storage, open source, Kent Overstreet

Viral Phrases: “game-changing release,” “enterprise-ready,” “silent data corruption,” “transactionally consistent,” “modern copy-on-write,” “cutting-edge innovation,” “future-proof storage,” “robust data recovery,” “terminal user interface,” “out-of-tree module,” “advanced filesystem features,” “production-ready,” “data integrity revolution,” “next-gen Linux storage,” “unparalleled reliability,” “automatic repair,” “tiered storage setups,” “hardware fault tolerance,” “real-time filesystem stats,” “open source storage evolution.”

,