Documentation/filesystems/btrfs.txt

   1
   2 BTRFS
   3 =====
   4
   5 Btrfs is a copy on write filesystem for Linux aimed at
   6 implementing advanced features while focusing on fault tolerance,
   7 repair and easy administration. Initially developed by Oracle, Btrfs
   8 is licensed under the GPL and open for contribution from anyone.
   9
  10 Linux has a wealth of filesystems to choose from, but we are facing a
  11 number of challenges with scaling to the large storage subsystems that
  12 are becoming common in today's data centers. Filesystems need to scale
  13 in their ability to address and manage large storage, and also in
  14 their ability to detect, repair and tolerate errors in the data stored
  15 on disk.  Btrfs is under heavy development, and is not suitable for
  16 any uses other than benchmarking and review. The Btrfs disk format is
  17 not yet finalized.
  18
  19 The main Btrfs features include:
  20
  21     * Extent based file storage (2^64 max file size)
  22     * Space efficient packing of small files
  23     * Space efficient indexed directories
  24     * Dynamic inode allocation
  25     * Writable snapshots
  26     * Subvolumes (separate internal filesystem roots)
  27     * Object level mirroring and striping
  28     * Checksums on data and metadata (multiple algorithms available)
  29     * Compression
  30     * Integrated multiple device support, with several raid algorithms
  31     * Online filesystem check (not yet implemented)
  32     * Very fast offline filesystem check
  33     * Efficient incremental backup and FS mirroring (not yet implemented)
  34     * Online filesystem defragmentation
  35
  36
  37 Mount Options
  38 =============
  39
  40 When mounting a btrfs filesystem, the following option are accepted.
  41 Unless otherwise specified, all options default to off.
  42
  43   alloc_start=<bytes>
  44         Debugging option to force all block allocations above a certain
  45         byte threshold on each block device.  The value is specified in
  46         bytes, optionally with a K, M, or G suffix, case insensitive.
  47         Default is 1MB.
  48
  49   autodefrag
  50         Detect small random writes into files and queue them up for the
  51         defrag process.  Works best for small files; Not well suited for
  52         large database workloads.
  53
  54   check_int
  55   check_int_data
  56   check_int_print_mask=<value>
  57         These debugging options control the behavior of the integrity checking
  58         module (the BTRFS_FS_CHECK_INTEGRITY config option required).
  59
  60         check_int enables the integrity checker module, which examines all
  61         block write requests to ensure on-disk consistency, at a large
  62         memory and CPU cost.
  63
  64         check_int_data includes extent data in the integrity checks, and
  65         implies the check_int option.
  66
  67         check_int_print_mask takes a bitmask of BTRFSIC_PRINT_MASK_* values
  68         as defined in fs/btrfs/check-integrity.c, to control the integrity
  69         checker module behavior.
  70
  71         See comments at the top of fs/btrfs/check-integrity.c for more info.
  72
  73   commit=<seconds>
  74         Set the interval of periodic commit, 30 seconds by default. Higher
  75         values defer data being synced to permanent storage with obvious
  76         consequences when the system crashes. The upper bound is not forced,
  77         but a warning is printed if it's more than 300 seconds (5 minutes).
  78
  79   compress
  80   compress=<type>
  81   compress-force
  82   compress-force=<type>
  83         Control BTRFS file data compression.  Type may be specified as "zlib"
  84         "lzo" or "no" (for no compression, used for remounting).  If no type
  85         is specified, zlib is used.  If compress-force is specified,
  86         all files will be compressed, whether or not they compress well.
  87         If compression is enabled, nodatacow and nodatasum are disabled.
  88
  89   degraded
  90         Allow mounts to continue with missing devices.  A read-write mount may
  91         fail with too many devices missing, for example if a stripe member
  92         is completely missing.
  93
  94   device=<devicepath>
  95         Specify a device during mount so that ioctls on the control device
  96         can be avoided.  Especially useful when trying to mount a multi-device
  97         setup as root.  May be specified multiple times for multiple devices.
  98
  99   discard
 100         Issue frequent commands to let the block device reclaim space freed by
 101         the filesystem.  This is useful for SSD devices, thinly provisioned
 102         LUNs and virtual machine images, but may have a significant
 103         performance impact.  (The fstrim command is also available to
 104         initiate batch trims from userspace).
 105
 106   enospc_debug
 107         Debugging option to be more verbose in some ENOSPC conditions.
 108
 109   fatal_errors=<action>
 110         Action to take when encountering a fatal error:
 111           "bug" - BUG() on a fatal error.  This is the default.
 112           "panic" - panic() on a fatal error.
 113
 114   flushoncommit
 115         The 'flushoncommit' mount option forces any data dirtied by a write in a
 116         prior transaction to commit as part of the current commit.  This makes
 117         the committed state a fully consistent view of the file system from the
 118         application's perspective (i.e., it includes all completed file system
 119         operations).  This was previously the behavior only when a snapshot is
 120         created.
 121
 122   inode_cache
 123         Enable free inode number caching.   Defaults to off due to an overflow
 124         problem when the free space crcs don't fit inside a single page.
 125
 126   max_inline=<bytes>
 127         Specify the maximum amount of space, in bytes, that can be inlined in
 128         a metadata B-tree leaf.  The value is specified in bytes, optionally
 129         with a K, M, or G suffix, case insensitive.  In practice, this value
 130         is limited by the root sector size, with some space unavailable due
 131         to leaf headers.  For a 4k sectorsize, max inline data is ~3900 bytes.
 132
 133   metadata_ratio=<value>
 134         Specify that 1 metadata chunk should be allocated after every <value>
 135         data chunks.  Off by default.
 136
 137   noacl
 138         Disable support for Posix Access Control Lists (ACLs).  See the
 139         acl(5) manual page for more information about ACLs.
 140
 141   nobarrier
 142         Disables the use of block layer write barriers.  Write barriers ensure
 143         that certain IOs make it through the device cache and are on persistent
 144         storage.  If used on a device with a volatile (non-battery-backed)
 145         write-back cache, this option will lead to filesystem corruption on a
 146         system crash or power loss.
 147
 148   nodatacow
 149         Disable data copy-on-write for newly created files.  Implies nodatasum,
 150         and disables all compression.
 151
 152   nodatasum
 153         Disable data checksumming for newly created files.
 154
 155   notreelog
 156         Disable the tree logging used for fsync and O_SYNC writes.
 157
 158   recovery
 159         Enable autorecovery attempts if a bad tree root is found at mount time.
 160         Currently this scans a list of several previous tree roots and tries to
 161         use the first readable.
 162
 163   rescan_uuid_tree
 164         Force check and rebuild procedure of the UUID tree. This should not
 165         normally be needed.
 166
 167   skip_balance
 168         Skip automatic resume of interrupted balance operation after mount.
 169         May be resumed with "btrfs balance resume."
 170
 171   space_cache (*)
 172         Enable the on-disk freespace cache.
 173   nospace_cache
 174         Disable freespace cache loading without clearing the cache.
 175   clear_cache
 176         Force clearing and rebuilding of the disk space cache if something
 177         has gone wrong.
 178
 179   ssd
 180   nossd
 181   ssd_spread
 182         Options to control ssd allocation schemes.  By default, BTRFS will
 183         enable or disable ssd allocation heuristics depending on whether a
 184         rotational or nonrotational disk is in use.  The ssd and nossd options
 185         can override this autodetection.
 186
 187         The ssd_spread mount option attempts to allocate into big chunks
 188         of unused space, and may perform better on low-end ssds.  ssd_spread
 189         implies ssd, enabling all other ssd heuristics as well.
 190
 191   subvol=<path>
 192         Mount subvolume at <path> rather than the root subvolume.  <path> is
 193         relative to the top level subvolume.
 194
 195   subvolid=<ID>
 196         Mount subvolume specified by an ID number rather than the root subvolume.
 197         This allows mounting of subvolumes which are not in the root of the mounted
 198         filesystem.
 199         You can use "btrfs subvolume list" to see subvolume ID numbers.
 200
 201   subvolrootid=<objectid> (deprecated)
 202         Mount subvolume specified by <objectid> rather than the root subvolume.
 203         This allows mounting of subvolumes which are not in the root of the mounted
 204         filesystem.
 205         You can use "btrfs subvolume show " to see the object ID for a subvolume.
 206
 207   thread_pool=<number>
 208         The number of worker threads to allocate.  The default number is equal
 209         to the number of CPUs + 2, or 8, whichever is smaller.
 210
 211   user_subvol_rm_allowed
 212         Allow subvolumes to be deleted by a non-root user. Use with caution.
 213
 214 MAILING LIST
 215 ============
 216
 217 There is a Btrfs mailing list hosted on vger.kernel.org. You can
 218 find details on how to subscribe here:
 219
 220 http://vger.kernel.org/vger-lists.html#linux-btrfs
 221
 222 Mailing list archives are available from gmane:
 223
 224 http://dir.gmane.org/gmane.comp.file-systems.btrfs
 225
 226
 227
 228 IRC
 229 ===
 230
 231 Discussion of Btrfs also occurs on the #btrfs channel of the Freenode
 232 IRC network.
 233
 234
 235
 236         UTILITIES
 237         =========
 238
 239 Userspace tools for creating and manipulating Btrfs file systems are
 240 available from the git repository at the following location:
 241
 242  http://git.kernel.org/?p=linux/kernel/git/mason/btrfs-progs.git
 243  git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-progs.git
 244
 245 These include the following tools:
 246
 247 * mkfs.btrfs: create a filesystem
 248
 249 * btrfs: a single tool to manage the filesystems, refer to the manpage for more details
 250
 251 * 'btrfsck' or 'btrfs check': do a consistency check of the filesystem
 252
 253 Other tools for specific tasks:
 254
 255 * btrfs-convert: in-place conversion from ext2/3/4 filesystems
 256
 257 * btrfs-image: dump filesystem metadata for debugging