btrfs

btrfs, short for B-tree filesystem, is a copy-on-write (COW) filesystem for Linux designed to provide scalable storage with strong data integrity and flexible management features. It uses B-trees to organize data and metadata and includes mechanisms to detect and repair data corruption, aiming to reduce admin overhead on large storage pools.

Key features of btrfs include copy-on-write semantics, built-in data and metadata checksums, and support for snapshots and subvolumes. Snapshots allow point-in-time views of the filesystem, while subvolumes enable separate root-like trees within a single filesystem. It offers transparent compression (zlib, LZO, and newer kernels add zstd), reflinks for efficient cloning, and send/receive for incremental backups. Online defragmentation and balance operations enable maintenance without taking the filesystem offline, and quotas can be implemented using qgroups to manage space usage.

btrfs emphasizes data integrity through per-block checksums; scrubbing can detect and repair errors by re-reading data and re-writing from healthy copies when redundancy is available. The filesystem provides RAID-like data layout across multiple devices, including RAID 0, 1, and 10 in many deployments, with RAID 5/6 support historically present but widely discouraged due to reliability concerns. Regular checks and maintenance are facilitated by built-in tooling and the associated userspace utilities.

Usage and ecosystem: btrfs is supported by the Linux kernel and maintained by the community, with tooling provided by the btrfs-progs package. It is used in various distributions and scenarios where features like snapshots, subvolumes, and integrity checks are valued, though it is not universally adopted as the default filesystem for all workloads. Users are advised to thoroughly test stability and performance for their specific use cases before deployable production use.