Design Filesystem Using Tree Should Use B-tree

btrfs is a modern copy on write (CoW) filesystem for Linux aimed at implementing advanced features while also focusing on fault tolerance, repair and easy administration. Its main features and benefits are:

• Snapshots which do not make the full copy of files
• RAID - support for software-based RAID 0, RAID 1, RAID 10
• Self-healing - checksums for data and metadata, automatic detection of silent data corruptions

Development of Btrfs started in 2007. Since that time, Btrfs is a part of the Linux kernel and is under active development.

Jointly developed at multiple companies, Btrfs is licensed under the GPL and open for contribution from anyone.

List of companies using btrfs in production.

Development and Issue Reporting

For feature status, please refer to the Status page.

The Btrfs code base is stable. However, new features are still under development. Every effort is made to ensure that it remains stable and fast at each and every commit. This rapid pace of development means that the filesystem improves noticeably with every new Linux release so it's highly recommended that users run the most modern kernel possible.

For benchmarks, it's recommended to test the latest stable Linux version, and not any older, as well as the latest Linux development versions. Also, it's recommended to test the various mount options such as different compression options.

If you find any behavior you suspect to be caused by a bug, performance issues, or have any questions about using Btrfs, please email the Btrfs mailing list (no subscription required). Please report bugs on Bugzilla.

Features

Linux has a wealth of filesystems from which to choose, but we are facing a number of challenges with scaling to the large storage subsystems that are becoming common in today's data centers. Filesystems need to scale in their ability to address and manage large storage, and also in their ability to detect, repair and tolerate errors in the data stored on disk.

Major Features Currently Implemented

• Extent based file storage
• 2^64 byte == 16 EiB maximum file size (practical limit is 8 EiB due to Linux VFS)
• Space-efficient packing of small files
• Space-efficient indexed directories
• Dynamic inode allocation
• Writable snapshots, read-only snapshots
• Subvolumes (separate internal filesystem roots)
• Checksums on data and metadata (crc32c, xxhash, sha256, blake2b)
• Compression (ZLIB, LZO, ZSTD), heuristics
• Integrated multiple device support
  • File Striping
  • File Mirroring
  • File Striping+Mirroring
  • Single and Dual Parity implementations (experimental, not production-ready)
• SSD (flash storage) awareness (TRIM/Discard for reporting free blocks for reuse) and optimizations (e.g. avoiding unnecessary seek optimizations, sending writes in clusters, even if they are from unrelated files. This results in larger write operations and faster write throughput)
• Efficient incremental backup
• Background scrub process for finding and repairing errors of files with redundant copies
• Online filesystem defragmentation
• Offline filesystem check
• In-place conversion of existing ext2/3/4 and reiserfs file systems
• Seed devices. Create a (readonly) filesystem that acts as a template to seed other Btrfs filesystems. The original filesystem and devices are included as a readonly starting point for the new filesystem. Using copy on write, all modifications are stored on different devices; the original is unchanged.
• Subvolume-aware quota support
• Send/receive of subvolume changes
  • Efficient incremental filesystem mirroring
• Batch, or out-of-band deduplication (happens after writes, not during)
• Swapfile support
• Tree-checker, post-read and pre-write metadata verification
• Zoned mode support (SMR/ZBC/ZNS friendly allocation)

Features by kernel version

As part of the changelog you can also review

• features by kernel version

Features Currently in Development or Planned for Future Implementation

• DAX/persistent memory support
• The file/directory -level encryption support (fscrypt)
• fsverity integration

Documentation

Guides and usage information

• Getting started — first steps, distributions with btrfs support
  • Mount options
• FAQ — About the btrfs project and filesystem
• UseCases — Recipes for how to do stuff with btrfs
• SysadminGuide — A more in-depth guide to btrfs's concepts and a bit of its internals, to answer all those "but what is a subvolume?" kind of questions.
• Multiple devices – A guide to the RAID features of Btrfs
• Conversion from Ext3 and Ext4 or reiserfs
• Problem FAQ — Commonly-encountered problems and solutions.
  • Gotchas — lists known bugs and issues, but not necessarily solutions.

External Btrfs Documentation / Guides

Links to Btrfs documentation of various Linux distributions:

• "The Btrfs File System" chapter in the Oracle Linux 6 Administrator's Solutions Guide
• Major File Systems in Linux chapter in the SLES 15 Storage Administration Guide
• Btrfs Wiki page on the Ubuntu Community Help Wiki
• Btrfs Wiki page on the Arch Linux Wiki
• Marc MERLIN's Btrfs talk at Linuxcon JP 2014 which gives an overview of Btrfs, best practices, and its more interesting features.

Manual pages

• Manual pages generated from git (complete list):
  • btrfs — main administration tool
  • mkfs.btrfs — creating the filesystem
  • btrfs check — repairing file systems
  • btrfs-convert — tool to convert in-place from ext2/3/4 filesystems to btrfs. For a greater detail of how the algorithm works, please see the Conversion from Ext3 page.

• Original wiki documentation (obsolete, will be removed)
  • restore and find-root — utilities to find and restore data from an unmountable filesystem

Developer documentation

• Development setup - how to build btrfs from sources and prepare a development environment

• Developer documentation on Github - collection of documents describing internals of BTRFS
• Developer's FAQ — hints and answers for contributors and developers, general information about patch formatting
• Development notes — notes, hints, checklists for specific implementation tasks (eg. adding new ioctls)
• Code documentation — trees, source files, sample code for manipulating trees
• Data Structures — detailed on-disk data structures
• On-disk Format - Details of data structures written on disk
• Trees — detailed in-tree representation of files and directories

• Btrfs design — design notes (possibly out of date in places)
• Multiple Device Support — design notes
• ENOSPC — Current ENOSPC design issues
• Design_notes_on_Send/Receive — notes from initial impelentation, protocol V2 updates draft
• Qgroups status quo - notes on some qgroups observations/pain points

• Debugging Btrfs with GDB
• Writing patch for btrfs
• Btree_Items - Mapping from Btrfs key to item-data
• Resolving_Extent_Backrefs - How back references are resolved to root owners

• Original COW B-tree: Source code in C that implements the COW B-tree algorithms repository. Written by Ohad Rodeh at IBM Research in 2006, and released under a BSD license. This is a reference implementation, that works in user space.

• Unmerged features
  • In-band (write) time deduplication
    • User notes on dedupe — User/tester notes for using in-band deduplication feature

News

New location for documentation

The new place for documentation will be at https://btrfs.readthedocs.org or https://btrfs.rtfd.io , will contents is going to be migrated

IRC channel at libera.chat

The #btrfs channel is at libera.chat, matrix.org bridge works (persistent room #btrfs:matrix.org).

zstd (Nov 15)

The zstd implementation 1.4.10 in kernel has been merged to 5.15-rc1, speedups and sync with upstream version

btrfs-progs v5.15 (Nov 2021)

• mkfs: new defaults!
  • no-holes
  • free-space-tree
  • DUP for metadata unconditionally
• libbtrfsutil: add missing profile defines
• libbtrfs: minimize its impact on the other code, refactor and separate implementation where needed, cleanup afterwards, reduced header exports
• documentation: introduce sphinx build and RST versions of manual pages, will become the new format and replace asciidoc
• fixes: fix warning regarding v1 space cache when only v2 (free space tree) is enabled

linux v5.15 (Nov 2021)

Features:

• fs-verity support, using standard ioctls, backward compatible with read-only limitation on inodes with previously enabled fs-verity
• idmapped mount support
• make mount with rescue=ibadroots more tolerant to partially damaged trees
• allow raid0 on a single device and raid10 on two devices, degenerate cases but might be useful as an intermediate step during conversion to other profiles
• zoned mode block group auto reclaim can be disabled via sysfs knob

Performance improvements:

• continue readahead of node siblings even if target node is in memory, could speed up full send (on sample test +11%)
• batching of delayed items can speed up creating many files
• fsync/tree-log speedups
  • avoid unnecessary work (gains +2% throughput, -2% run time on sample load)
  • reduced lock contention on renames (on dbench +4% throughput, up to -30% latency)

Fixes:

• various zoned mode fixes
• preemptive flushing threshold tuning, avoid excessive work on almost full filesystems

Core:

• continued subpage support, preparation for implementing remaining features like compression and defragmentation; with some limitations, write is now enabled on 64K page systems with 4K sectors, still considered experimental
  • no readahead on compressed reads
  • inline extents disabled
  • disabled raid56 profile conversion and mount
• improved flushing logic, fixing early ENOSPC on some workloads
• inode flags have been internally split to read-only and read-write incompat bit parts, used by fs-verity
• new tree items for fs-verity: descriptor item, Merkle tree item
• inode operations extended to be namespace-aware
• cleanups and refactoring

Articles, presentations, podcasts

• Video: Deploying Btrfs at Facebook Scale by Josef Bacik at the Open Source Summit 2020 (2020-06-29)
• Video: btrfs is awesome, except when it isn't by Richard Brown at openSUSE Conferece 2018 (2018-05-25)
• Video: btrfs: The Best Filesystem You've Never Heard Of by poiupoiu at PhreakNIC 21 (2017-11-3)
• Video TUT91782 Getting the most out of the btrfs filesystem by Thorsthen Kukuk and Jeff Mahoney (SUSECON, 2017)
• Video: NYLUG Presents: Chris Mason on Btrfs (May 14th 2015) by Chris Mason at the 192nd meeting of the NYLUG
• Video: Why you should consider using btrfs ... like Google does. by Marc Merlin at linux.conf.au 2015. talk slides
• Article: Bitrot and atomic COWs: Inside "next-gen" filesystems (ars technica, 2014/01)
• Article: Btrfs: Subvolumes and snapshots (LWN.net, 2014/01)
• Article: Btrfs: Working with multiple devices (LWN.net, 2013/12)
• Article: Btrfs: Getting started (LWN.net, 2013/12)
• Article: Btrfs hands on: An extremely cool file system (ZDNet, 2013/11)
• Technical report: Visualizating Block IO Workloads. Section six shows a visual comparison of the IO patterns for BTRFS, XFS, and EXT4. Submitted to ACM Transactions on Storage, November 2013.
• Paper: BTRFS: The Linux B-Tree Filesystem describing the overall concepts and architecture, appeared in ACM Transactions on Storage, August 2013. Includes a detailed comparison with ZFS. There is a free ACM authorized link, from O. Rodeh's [1] page. Otherwise, try IBM Research link

Project information/Contact

• Changelog — history of changes in linux kernel wrt btrfs
  • features added by release
• Development statistics — contributors, commits, lines
• Glossary
• Contact information:
  • Btrfs mailing list
  • IRC on libera.chat in the channel #btrfs
  • Reporting bugs:
    • for kernel code see the Bugzilla FAQ, quick tip: use product File System and component btrfs.
    • for btrfs-progs it's either bugzilla or github issues
    • for read-only documentation exported on wiki eg. manual pages as github issues
• Project ideas
• Cleanup ideas
• Userspace tools projects

Wiki accounts, editing

The wiki contributions are welcome! Please create an account and wait for approval (this is a necessary spam protection and we cannot remove it). You can try to catch some of the wiki admins on IRC (or ping user 'kdave' in a query) to expedite the account creation.

The registration requires full name for account but it's not mandatory from our perspective. The wiki User and User talk pages are created automatically but removed after account is approved. If you want to use the pages, create them manually, they won't be deleted.

Design Filesystem Using Tree Should Use B-tree

Source: https://btrfs.wiki.kernel.org/index.php/Main%5FPage

Share this post