1Tools that manage md devices can be found at
   5Boot time assembly of RAID arrays
   8You can boot with your md device with the following kernel command
  11for old raid arrays without persistent superblocks:
  12  md=<md device no.>,<raid level>,<chunk size factor>,<fault level>,dev0,dev1,...,devn
  14for raid arrays with persistent superblocks
  15  md=<md device no.>,dev0,dev1,...,devn
  16or, to assemble a partitionable array:
  17  md=d<md device no.>,dev0,dev1,...,devn
  19md device no. = the number of the md device ... 
  20              0 means md0, 
  21              1 md1,
  22              2 md2,
  23              3 md3,
  24              4 md4
  26raid level = -1 linear mode
  27              0 striped mode
  28              other modes are only supported with persistent super blocks
  30chunk size factor = (raid-0 and raid-1 only)
  31              Set  the chunk size as 4k << n.
  33fault level = totally ignored
  35dev0-devn: e.g. /dev/hda1,/dev/hdc1,/dev/sda1,/dev/sdb1
  37A possible loadlin line (Harald Hoyer <HarryH@Royal.Net>)  looks like this:
  39e:\loadlin\loadlin e:\zimage root=/dev/md0 md=0,0,4,0,/dev/hdb2,/dev/hdc3 ro
  42Boot time autodetection of RAID arrays
  45When md is compiled into the kernel (not as module), partitions of
  46type 0xfd are scanned and automatically assembled into RAID arrays.
  47This autodetection may be suppressed with the kernel parameter
  48"raid=noautodetect".  As of kernel 2.6.9, only drives with a type 0
  49superblock can be autodetected and run at boot time.
  51The kernel parameter "raid=partitionable" (or "raid=part") means
  52that all auto-detected arrays are assembled as partitionable.
  54Boot time assembly of degraded/dirty arrays
  57If a raid5 or raid6 array is both dirty and degraded, it could have
  58undetectable data corruption.  This is because the fact that it is
  59'dirty' means that the parity cannot be trusted, and the fact that it
  60is degraded means that some datablocks are missing and cannot reliably
  61be reconstructed (due to no parity).
  63For this reason, md will normally refuse to start such an array.  This
  64requires the sysadmin to take action to explicitly start the array
  65despite possible corruption.  This is normally done with
  66   mdadm --assemble --force ....
  68This option is not really available if the array has the root
  69filesystem on it.  In order to support this booting from such an
  70array, md supports a module parameter "start_dirty_degraded" which,
  71when set to 1, bypassed the checks and will allows dirty degraded
  72arrays to be started.
  74So, to boot with a root filesystem of a dirty degraded raid[56], use
  76   md-mod.start_dirty_degraded=1
  79Superblock formats
  82The md driver can support a variety of different superblock formats.
  83Currently, it supports superblock formats "0.90.0" and the "md-1" format
  84introduced in the 2.5 development series.
  86The kernel will autodetect which format superblock is being used.
  88Superblock format '0' is treated differently to others for legacy
  89reasons - it is the original superblock format.
  92General Rules - apply for all superblock formats
  95An array is 'created' by writing appropriate superblocks to all
  98It is 'assembled' by associating each of these devices with an
  99particular md virtual device.  Once it is completely assembled, it can
 100be accessed.
 102An array should be created by a user-space tool.  This will write
 103superblocks to all devices.  It will usually mark the array as
 104'unclean', or with some devices missing so that the kernel md driver
 105can create appropriate redundancy (copying in raid1, parity
 106calculation in raid4/5).
 108When an array is assembled, it is first initialized with the
 109SET_ARRAY_INFO ioctl.  This contains, in particular, a major and minor
 110version number.  The major version number selects which superblock
 111format is to be used.  The minor number might be used to tune handling
 112of the format, such as suggesting where on each device to look for the
 115Then each device is added using the ADD_NEW_DISK ioctl.  This
 116provides, in particular, a major and minor number identifying the
 117device to add.
 119The array is started with the RUN_ARRAY ioctl.
 121Once started, new devices can be added.  They should have an
 122appropriate superblock written to them, and then passed be in with
 125Devices that have failed or are not yet active can be detached from an
 126array using HOT_REMOVE_DISK.
 129Specific Rules that apply to format-0 super block arrays, and
 130       arrays with no superblock (non-persistent).
 133An array can be 'created' by describing the array (level, chunksize
 134etc) in a SET_ARRAY_INFO ioctl.  This must has major_version==0 and
 135raid_disks != 0.
 137Then uninitialized devices can be added with ADD_NEW_DISK.  The
 138structure passed to ADD_NEW_DISK must specify the state of the device
 139and its role in the array.
 141Once started with RUN_ARRAY, uninitialized spares can be added with
 146MD devices in sysfs
 148md devices appear in sysfs (/sys) as regular block devices,
 150   /sys/block/md0
 152Each 'md' device will contain a subdirectory called 'md' which
 153contains further md-specific information about the device.
 155All md devices contain:
 156  level
 157     a text file indicating the 'raid level'. e.g. raid0, raid1,
 158     raid5, linear, multipath, faulty.
 159     If no raid level has been set yet (array is still being
 160     assembled), the value will reflect whatever has been written
 161     to it, which may be a name like the above, or may be a number
 162     such as '0', '5', etc.
 164  raid_disks
 165     a text file with a simple number indicating the number of devices
 166     in a fully functional array.  If this is not yet known, the file
 167     will be empty.  If an array is being resized this will contain
 168     the new number of devices.
 169     Some raid levels allow this value to be set while the array is
 170     active.  This will reconfigure the array.   Otherwise it can only
 171     be set while assembling an array.
 172     A change to this attribute will not be permitted if it would
 173     reduce the size of the array.  To reduce the number of drives
 174     in an e.g. raid5, the array size must first be reduced by
 175     setting the 'array_size' attribute.
 177  chunk_size
 178     This is the size in bytes for 'chunks' and is only relevant to
 179     raid levels that involve striping (0,4,5,6,10). The address space
 180     of the array is conceptually divided into chunks and consecutive
 181     chunks are striped onto neighbouring devices.
 182     The size should be at least PAGE_SIZE (4k) and should be a power
 183     of 2.  This can only be set while assembling an array
 185  layout
 186     The "layout" for the array for the particular level.  This is
 187     simply a number that is interpretted differently by different
 188     levels.  It can be written while assembling an array.
 190  array_size
 191     This can be used to artificially constrain the available space in
 192     the array to be less than is actually available on the combined
 193     devices.  Writing a number (in Kilobytes) which is less than
 194     the available size will set the size.  Any reconfiguration of the
 195     array (e.g. adding devices) will not cause the size to change.
 196     Writing the word 'default' will cause the effective size of the
 197     array to be whatever size is actually available based on
 198     'level', 'chunk_size' and 'component_size'.
 200     This can be used to reduce the size of the array before reducing
 201     the number of devices in a raid4/5/6, or to support external
 202     metadata formats which mandate such clipping.
 204  reshape_position
 205     This is either "none" or a sector number within the devices of
 206     the array where "reshape" is up to.  If this is set, the three
 207     attributes mentioned above (raid_disks, chunk_size, layout) can
 208     potentially have 2 values, an old and a new value.  If these
 209     values differ, reading the attribute returns
 210        new (old)
 211     and writing will effect the 'new' value, leaving the 'old'
 212     unchanged.
 214  component_size
 215     For arrays with data redundancy (i.e. not raid0, linear, faulty,
 216     multipath), all components must be the same size - or at least
 217     there must a size that they all provide space for.  This is a key
 218     part or the geometry of the array.  It is measured in sectors
 219     and can be read from here.  Writing to this value may resize
 220     the array if the personality supports it (raid1, raid5, raid6),
 221     and if the component drives are large enough.
 223  metadata_version
 224     This indicates the format that is being used to record metadata
 225     about the array.  It can be 0.90 (traditional format), 1.0, 1.1,
 226     1.2 (newer format in varying locations) or "none" indicating that
 227     the kernel isn't managing metadata at all.
 228     Alternately it can be "external:" followed by a string which
 229     is set by user-space.  This indicates that metadata is managed
 230     by a user-space program.  Any device failure or other event that
 231     requires a metadata update will cause array activity to be
 232     suspended until the event is acknowledged.
 234  resync_start
 235     The point at which resync should start.  If no resync is needed,
 236     this will be a very large number (or 'none' since 2.6.30-rc1).  At
 237     array creation it will default to 0, though starting the array as
 238     'clean' will set it much larger.
 240   new_dev
 241     This file can be written but not read.  The value written should
 242     be a block device number as major:minor.  e.g. 8:0
 243     This will cause that device to be attached to the array, if it is
 244     available.  It will then appear at md/dev-XXX (depending on the
 245     name of the device) and further configuration is then possible.
 247   safe_mode_delay
 248     When an md array has seen no write requests for a certain period
 249     of time, it will be marked as 'clean'.  When another write
 250     request arrives, the array is marked as 'dirty' before the write
 251     commences.  This is known as 'safe_mode'.
 252     The 'certain period' is controlled by this file which stores the
 253     period as a number of seconds.  The default is 200msec (0.200).
 254     Writing a value of 0 disables safemode.
 256   array_state
 257     This file contains a single word which describes the current
 258     state of the array.  In many cases, the state can be set by
 259     writing the word for the desired state, however some states
 260     cannot be explicitly set, and some transitions are not allowed.
 262     Select/poll works on this file.  All changes except between
 263        active_idle and active (which can be frequent and are not
 264        very interesting) are notified.  active->active_idle is
 265        reported if the metadata is externally managed.
 267     clear
 268         No devices, no size, no level
 269         Writing is equivalent to STOP_ARRAY ioctl
 270     inactive
 271         May have some settings, but array is not active
 272            all IO results in error
 273         When written, doesn't tear down array, but just stops it
 274     suspended (not supported yet)
 275         All IO requests will block. The array can be reconfigured.
 276         Writing this, if accepted, will block until array is quiessent
 277     readonly
 278         no resync can happen.  no superblocks get written.
 279         write requests fail
 280     read-auto
 281         like readonly, but behaves like 'clean' on a write request.
 283     clean - no pending writes, but otherwise active.
 284         When written to inactive array, starts without resync
 285         If a write request arrives then
 286           if metadata is known, mark 'dirty' and switch to 'active'.
 287           if not known, block and switch to write-pending
 288         If written to an active array that has pending writes, then fails.
 289     active
 290         fully active: IO and resync can be happening.
 291         When written to inactive array, starts with resync
 293     write-pending
 294         clean, but writes are blocked waiting for 'active' to be written.
 296     active-idle
 297         like active, but no writes have been seen for a while (safe_mode_delay).
 299  bitmap/location
 300     This indicates where the write-intent bitmap for the array is
 301     stored.
 302     It can be one of "none", "file" or "[+-]N".
 303     "file" may later be extended to "file:/file/name"
 304     "[+-]N" means that many sectors from the start of the metadata.
 305       This is replicated on all devices.  For arrays with externally
 306       managed metadata, the offset is from the beginning of the
 307       device.
 308  bitmap/chunksize
 309     The size, in bytes, of the chunk which will be represented by a
 310     single bit.  For RAID456, it is a portion of an individual
 311     device. For RAID10, it is a portion of the array.  For RAID1, it
 312     is both (they come to the same thing).
 313  bitmap/time_base
 314     The time, in seconds, between looking for bits in the bitmap to
 315     be cleared. In the current implementation, a bit will be cleared
 316     between 2 and 3 times "time_base" after all the covered blocks
 317     are known to be in-sync.
 318  bitmap/backlog
 319     When write-mostly devices are active in a RAID1, write requests
 320     to those devices proceed in the background - the filesystem (or
 321     other user of the device) does not have to wait for them.
 322     'backlog' sets a limit on the number of concurrent background
 323     writes.  If there are more than this, new writes will by
 324     synchronous.
 325  bitmap/metadata
 326     This can be either 'internal' or 'external'.
 327     'internal' is the default and means the metadata for the bitmap
 328     is stored in the first 256 bytes of the allocated space and is
 329     managed by the md module.
 330     'external' means that bitmap metadata is managed externally to
 331     the kernel (i.e. by some userspace program)
 332  bitmap/can_clear
 333     This is either 'true' or 'false'.  If 'true', then bits in the
 334     bitmap will be cleared when the corresponding blocks are thought
 335     to be in-sync.  If 'false', bits will never be cleared.
 336     This is automatically set to 'false' if a write happens on a
 337     degraded array, or if the array becomes degraded during a write.
 338     When metadata is managed externally, it should be set to true
 339     once the array becomes non-degraded, and this fact has been
 340     recorded in the metadata.
 345As component devices are added to an md array, they appear in the 'md'
 346directory as new directories named
 347      dev-XXX
 348where XXX is a name that the kernel knows for the device, e.g. hdb1.
 349Each directory contains:
 351      block
 352        a symlink to the block device in /sys/block, e.g.
 353             /sys/block/md0/md/dev-hdb1/block -> ../../../../block/hdb/hdb1
 355      super
 356        A file containing an image of the superblock read from, or
 357        written to, that device.
 359      state
 360        A file recording the current state of the device in the array
 361        which can be a comma separated list of
 362              faulty   - device has been kicked from active use due to
 363                         a detected fault, or it has unacknowledged bad
 364                         blocks
 365              in_sync  - device is a fully in-sync member of the array
 366              writemostly - device will only be subject to read
 367                         requests if there are no other options.
 368                         This applies only to raid1 arrays.
 369              blocked  - device has failed, and the failure hasn't been
 370                         acknowledged yet by the metadata handler.
 371                         Writes that would write to this device if
 372                         it were not faulty are blocked.
 373              spare    - device is working, but not a full member.
 374                         This includes spares that are in the process
 375                         of being recovered to
 376              write_error - device has ever seen a write error.
 377              want_replacement - device is (mostly) working but probably
 378                         should be replaced, either due to errors or
 379                         due to user request.
 380              replacement - device is a replacement for another active
 381                         device with same raid_disk.
 384        This list may grow in future.
 385        This can be written to.
 386        Writing "faulty"  simulates a failure on the device.
 387        Writing "remove" removes the device from the array.
 388        Writing "writemostly" sets the writemostly flag.
 389        Writing "-writemostly" clears the writemostly flag.
 390        Writing "blocked" sets the "blocked" flag.
 391        Writing "-blocked" clears the "blocked" flags and allows writes
 392                to complete and possibly simulates an error.
 393        Writing "in_sync" sets the in_sync flag.
 394        Writing "write_error" sets writeerrorseen flag.
 395        Writing "-write_error" clears writeerrorseen flag.
 396        Writing "want_replacement" is allowed at any time except to a
 397                replacement device or a spare.  It sets the flag.
 398        Writing "-want_replacement" is allowed at any time.  It clears
 399                the flag.
 400        Writing "replacement" or "-replacement" is only allowed before
 401                starting the array.  It sets or clears the flag.
 404        This file responds to select/poll. Any change to 'faulty'
 405        or 'blocked' causes an event.
 407      errors
 408        An approximate count of read errors that have been detected on
 409        this device but have not caused the device to be evicted from
 410        the array (either because they were corrected or because they
 411        happened while the array was read-only).  When using version-1
 412        metadata, this value persists across restarts of the array.
 414        This value can be written while assembling an array thus
 415        providing an ongoing count for arrays with metadata managed by
 416        userspace.
 418      slot
 419        This gives the role that the device has in the array.  It will
 420        either be 'none' if the device is not active in the array
 421        (i.e. is a spare or has failed) or an integer less than the
 422        'raid_disks' number for the array indicating which position
 423        it currently fills.  This can only be set while assembling an
 424        array.  A device for which this is set is assumed to be working.
 426      offset
 427        This gives the location in the device (in sectors from the
 428        start) where data from the array will be stored.  Any part of
 429        the device before this offset us not touched, unless it is
 430        used for storing metadata (Formats 1.1 and 1.2).
 432      size
 433        The amount of the device, after the offset, that can be used
 434        for storage of data.  This will normally be the same as the
 435        component_size.  This can be written while assembling an
 436        array.  If a value less than the current component_size is
 437        written, it will be rejected.
 439      recovery_start
 440        When the device is not 'in_sync', this records the number of
 441        sectors from the start of the device which are known to be
 442        correct.  This is normally zero, but during a recovery
 443        operation is will steadily increase, and if the recovery is
 444        interrupted, restoring this value can cause recovery to
 445        avoid repeating the earlier blocks.  With v1.x metadata, this
 446        value is saved and restored automatically.
 448        This can be set whenever the device is not an active member of
 449        the array, either before the array is activated, or before
 450        the 'slot' is set.
 452        Setting this to 'none' is equivalent to setting 'in_sync'.
 453        Setting to any other value also clears the 'in_sync' flag.
 455      bad_blocks
 456        This gives the list of all known bad blocks in the form of
 457        start address and length (in sectors respectively). If output
 458        is too big to fit in a page, it will be truncated. Writing
 459        "sector length" to this file adds new acknowledged (i.e.
 460        recorded to disk safely) bad blocks.
 462      unacknowledged_bad_blocks
 463        This gives the list of known-but-not-yet-saved-to-disk bad
 464        blocks in the same form of 'bad_blocks'. If output is too big
 465        to fit in a page, it will be truncated. Writing to this file
 466        adds bad blocks without acknowledging them. This is largely
 467        for testing.
 471An active md device will also contain and entry for each active device
 472in the array.  These are named
 474    rdNN
 476where 'NN' is the position in the array, starting from 0.
 477So for a 3 drive array there will be rd0, rd1, rd2.
 478These are symbolic links to the appropriate 'dev-XXX' entry.
 479Thus, for example,
 480       cat /sys/block/md*/md/rd*/state
 481will show 'in_sync' on every line.
 485Active md devices for levels that support data redundancy (1,4,5,6)
 486also have
 488   sync_action
 489     a text file that can be used to monitor and control the rebuild
 490     process.  It contains one word which can be one of:
 491       resync        - redundancy is being recalculated after unclean
 492                       shutdown or creation
 493       recover       - a hot spare is being built to replace a
 494                       failed/missing device
 495       idle          - nothing is happening
 496       check         - A full check of redundancy was requested and is
 497                       happening.  This reads all block and checks
 498                       them. A repair may also happen for some raid
 499                       levels.
 500       repair        - A full check and repair is happening.  This is
 501                       similar to 'resync', but was requested by the
 502                       user, and the write-intent bitmap is NOT used to
 503                       optimise the process.
 505      This file is writable, and each of the strings that could be
 506      read are meaningful for writing.
 508       'idle' will stop an active resync/recovery etc.  There is no
 509           guarantee that another resync/recovery may not be automatically
 510           started again, though some event will be needed to trigger
 511           this.
 512        'resync' or 'recovery' can be used to restart the
 513           corresponding operation if it was stopped with 'idle'.
 514        'check' and 'repair' will start the appropriate process
 515           providing the current state is 'idle'.
 517      This file responds to select/poll.  Any important change in the value
 518      triggers a poll event.  Sometimes the value will briefly be
 519      "recover" if a recovery seems to be needed, but cannot be
 520      achieved. In that case, the transition to "recover" isn't
 521      notified, but the transition away is.
 523   degraded
 524      This contains a count of the number of devices by which the
 525      arrays is degraded.  So an optimal array with show '0'.  A
 526      single failed/missing drive will show '1', etc.
 527      This file responds to select/poll, any increase or decrease
 528      in the count of missing devices will trigger an event.
 530   mismatch_count
 531      When performing 'check' and 'repair', and possibly when
 532      performing 'resync', md will count the number of errors that are
 533      found.  The count in 'mismatch_cnt' is the number of sectors
 534      that were re-written, or (for 'check') would have been
 535      re-written.  As most raid levels work in units of pages rather
 536      than sectors, this my be larger than the number of actual errors
 537      by a factor of the number of sectors in a page.
 539   bitmap_set_bits
 540      If the array has a write-intent bitmap, then writing to this
 541      attribute can set bits in the bitmap, indicating that a resync
 542      would need to check the corresponding blocks. Either individual
 543      numbers or start-end pairs can be written.  Multiple numbers
 544      can be separated by a space.
 545      Note that the numbers are 'bit' numbers, not 'block' numbers.
 546      They should be scaled by the bitmap_chunksize.
 548   sync_speed_min
 549   sync_speed_max
 550     This are similar to /proc/sys/dev/raid/speed_limit_{min,max}
 551     however they only apply to the particular array.
 552     If no value has been written to these, of if the word 'system'
 553     is written, then the system-wide value is used.  If a value,
 554     in kibibytes-per-second is written, then it is used.
 555     When the files are read, they show the currently active value
 556     followed by "(local)" or "(system)" depending on whether it is
 557     a locally set or system-wide value.
 559   sync_completed
 560     This shows the number of sectors that have been completed of
 561     whatever the current sync_action is, followed by the number of
 562     sectors in total that could need to be processed.  The two
 563     numbers are separated by a '/'  thus effectively showing one
 564     value, a fraction of the process that is complete.
 565     A 'select' on this attribute will return when resync completes,
 566     when it reaches the current sync_max (below) and possibly at
 567     other times.
 569   sync_max
 570     This is a number of sectors at which point a resync/recovery
 571     process will pause.  When a resync is active, the value can
 572     only ever be increased, never decreased.  The value of 'max'
 573     effectively disables the limit.
 576   sync_speed
 577     This shows the current actual speed, in K/sec, of the current
 578     sync_action.  It is averaged over the last 30 seconds.
 580   suspend_lo
 581   suspend_hi
 582     The two values, given as numbers of sectors, indicate a range
 583     within the array where IO will be blocked.  This is currently
 584     only supported for raid4/5/6.
 586   sync_min
 587   sync_max
 588     The two values, given as numbers of sectors, indicate a range
 589     within the array where 'check'/'repair' will operate. Must be
 590     a multiple of chunk_size. When it reaches "sync_max" it will
 591     pause, rather than complete.
 592     You can use 'select' or 'poll' on "sync_completed" to wait for
 593     that number to reach sync_max.  Then you can either increase
 594     "sync_max", or can write 'idle' to "sync_action".
 597Each active md device may also have attributes specific to the
 598personality module that manages it.
 599These are specific to the implementation of the module and could
 600change substantially if the implementation changes.
 602These currently include
 604  stripe_cache_size  (currently raid5 only)
 605      number of entries in the stripe cache.  This is writable, but
 606      there are upper and lower limits (32768, 16).  Default is 128.
 607  strip_cache_active (currently raid5 only)
 608      number of active entries in the stripe cache
 609  preread_bypass_threshold (currently raid5 only)
 610      number of times a stripe requiring preread will be bypassed by
 611      a stripe that does not require preread.  For fairness defaults
 612      to 1.  Setting this to 0 disables bypass accounting and
 613      requires preread stripes to wait until all full-width stripe-
 614      writes are complete.  Valid values are 0 to stripe_cache_size.