xref: /linux/Documentation/admin-guide/md.rst (revision 6dfafbd0299a60bfb5d5e277fdf100037c7ded07)
1RAID arrays
2===========
3
4Boot time assembly of RAID arrays
5---------------------------------
6
7Tools that manage md devices can be found at
8   https://www.kernel.org/pub/linux/utils/raid/
9
10
11You can boot with your md device with the following kernel command
12lines:
13
14for old raid arrays without persistent superblocks::
15
16  md=<md device no.>,<raid level>,<chunk size factor>,<fault level>,dev0,dev1,...,devn
17
18for raid arrays with persistent superblocks::
19
20  md=<md device no.>,dev0,dev1,...,devn
21
22or, to assemble a partitionable array::
23
24  md=d<md device no.>,dev0,dev1,...,devn
25
26``md device no.``
27+++++++++++++++++
28
29The number of the md device
30
31================= =========
32``md device no.`` device
33================= =========
34              0		md0
35	      1		md1
36	      2		md2
37	      3		md3
38	      4		md4
39================= =========
40
41``raid level``
42++++++++++++++
43
44level of the RAID array
45
46=============== =============
47``raid level``  level
48=============== =============
49-1		linear mode
500		striped mode
51=============== =============
52
53other modes are only supported with persistent super blocks
54
55``chunk size factor``
56+++++++++++++++++++++
57
58(raid-0 and raid-1 only)
59
60Set  the chunk size as 4k << n.
61
62``fault level``
63+++++++++++++++
64
65Totally ignored
66
67``dev0`` to ``devn``
68++++++++++++++++++++
69
70e.g. ``/dev/hda1``, ``/dev/hdc1``, ``/dev/sda1``, ``/dev/sdb1``
71
72A possible loadlin line (Harald Hoyer <HarryH@Royal.Net>)  looks like this::
73
74	e:\loadlin\loadlin e:\zimage root=/dev/md0 md=0,0,4,0,/dev/hdb2,/dev/hdc3 ro
75
76
77Boot time autodetection of RAID arrays
78--------------------------------------
79
80When md is compiled into the kernel (not as module), partitions of
81type 0xfd are scanned and automatically assembled into RAID arrays.
82This autodetection may be suppressed with the kernel parameter
83``raid=noautodetect``.  As of kernel 2.6.9, only drives with a type 0
84superblock can be autodetected and run at boot time.
85
86The kernel parameter ``raid=partitionable`` (or ``raid=part``) means
87that all auto-detected arrays are assembled as partitionable.
88
89Boot time assembly of degraded/dirty arrays
90-------------------------------------------
91
92If a raid5 or raid6 array is both dirty and degraded, it could have
93undetectable data corruption.  This is because the fact that it is
94``dirty`` means that the parity cannot be trusted, and the fact that it
95is degraded means that some datablocks are missing and cannot reliably
96be reconstructed (due to no parity).
97
98For this reason, md will normally refuse to start such an array.  This
99requires the sysadmin to take action to explicitly start the array
100despite possible corruption.  This is normally done with::
101
102   mdadm --assemble --force ....
103
104This option is not really available if the array has the root
105filesystem on it.  In order to support this booting from such an
106array, md supports a module parameter ``start_dirty_degraded`` which,
107when set to 1, bypassed the checks and will allows dirty degraded
108arrays to be started.
109
110So, to boot with a root filesystem of a dirty degraded raid 5 or 6, use::
111
112   md-mod.start_dirty_degraded=1
113
114
115Superblock formats
116------------------
117
118The md driver can support a variety of different superblock formats.
119Currently, it supports superblock formats ``0.90.0`` and the ``md-1`` format
120introduced in the 2.5 development series.
121
122The kernel will autodetect which format superblock is being used.
123
124Superblock format ``0`` is treated differently to others for legacy
125reasons - it is the original superblock format.
126
127
128General Rules - apply for all superblock formats
129------------------------------------------------
130
131An array is ``created`` by writing appropriate superblocks to all
132devices.
133
134It is ``assembled`` by associating each of these devices with an
135particular md virtual device.  Once it is completely assembled, it can
136be accessed.
137
138An array should be created by a user-space tool.  This will write
139superblocks to all devices.  It will usually mark the array as
140``unclean``, or with some devices missing so that the kernel md driver
141can create appropriate redundancy (copying in raid 1, parity
142calculation in raid 4/5).
143
144When an array is assembled, it is first initialized with the
145SET_ARRAY_INFO ioctl.  This contains, in particular, a major and minor
146version number.  The major version number selects which superblock
147format is to be used.  The minor number might be used to tune handling
148of the format, such as suggesting where on each device to look for the
149superblock.
150
151Then each device is added using the ADD_NEW_DISK ioctl.  This
152provides, in particular, a major and minor number identifying the
153device to add.
154
155The array is started with the RUN_ARRAY ioctl.
156
157Once started, new devices can be added.  They should have an
158appropriate superblock written to them, and then be passed in with
159ADD_NEW_DISK.
160
161Devices that have failed or are not yet active can be detached from an
162array using HOT_REMOVE_DISK.
163
164
165Specific Rules that apply to format-0 super block arrays, and arrays with no superblock (non-persistent)
166--------------------------------------------------------------------------------------------------------
167
168An array can be ``created`` by describing the array (level, chunksize
169etc) in a SET_ARRAY_INFO ioctl.  This must have ``major_version==0`` and
170``raid_disks != 0``.
171
172Then uninitialized devices can be added with ADD_NEW_DISK.  The
173structure passed to ADD_NEW_DISK must specify the state of the device
174and its role in the array.
175
176Once started with RUN_ARRAY, uninitialized spares can be added with
177HOT_ADD_DISK.
178
179
180MD devices in sysfs
181-------------------
182
183md devices appear in sysfs (``/sys``) as regular block devices,
184e.g.::
185
186   /sys/block/md0
187
188Each ``md`` device will contain a subdirectory called ``md`` which
189contains further md-specific information about the device.
190
191All md devices contain:
192
193  level
194     a text file indicating the ``raid level``. e.g. raid0, raid1,
195     raid5, linear, multipath, faulty.
196     If no raid level has been set yet (array is still being
197     assembled), the value will reflect whatever has been written
198     to it, which may be a name like the above, or may be a number
199     such as ``0``, ``5``, etc.
200
201  raid_disks
202     a text file with a simple number indicating the number of devices
203     in a fully functional array.  If this is not yet known, the file
204     will be empty.  If an array is being resized this will contain
205     the new number of devices.
206     Some raid levels allow this value to be set while the array is
207     active.  This will reconfigure the array.   Otherwise it can only
208     be set while assembling an array.
209     A change to this attribute will not be permitted if it would
210     reduce the size of the array.  To reduce the number of drives
211     in an e.g. raid5, the array size must first be reduced by
212     setting the ``array_size`` attribute.
213
214  chunk_size
215     This is the size in bytes for ``chunks`` and is only relevant to
216     raid levels that involve striping (0,4,5,6,10). The address space
217     of the array is conceptually divided into chunks and consecutive
218     chunks are striped onto neighbouring devices.
219     The size should be at least PAGE_SIZE (4k) and should be a power
220     of 2.  This can only be set while assembling an array
221
222  layout
223     The ``layout`` for the array for the particular level.  This is
224     simply a number that is interpreted differently by different
225     levels.  It can be written while assembling an array.
226
227  array_size
228     This can be used to artificially constrain the available space in
229     the array to be less than is actually available on the combined
230     devices.  Writing a number (in Kilobytes) which is less than
231     the available size will set the size.  Any reconfiguration of the
232     array (e.g. adding devices) will not cause the size to change.
233     Writing the word ``default`` will cause the effective size of the
234     array to be whatever size is actually available based on
235     ``level``, ``chunk_size`` and ``component_size``.
236
237     This can be used to reduce the size of the array before reducing
238     the number of devices in a raid4/5/6, or to support external
239     metadata formats which mandate such clipping.
240
241  logical_block_size
242     Configure the array's logical block size in bytes. This attribute
243     is only supported for 1.x meta. Write the value before starting
244     array. The final array LBS uses the maximum between this
245     configuration and LBS of all combined devices. Note that
246     LBS cannot exceed PAGE_SIZE before RAID supports folio.
247     WARNING: Arrays created on new kernel cannot be assembled at old
248     kernel due to padding check, Set module parameter 'check_new_feature'
249     to false to bypass, but data loss may occur.
250
251  reshape_position
252     This is either ``none`` or a sector number within the devices of
253     the array where ``reshape`` is up to.  If this is set, the three
254     attributes mentioned above (raid_disks, chunk_size, layout) can
255     potentially have 2 values, an old and a new value.  If these
256     values differ, reading the attribute returns::
257
258        new (old)
259
260     and writing will effect the ``new`` value, leaving the ``old``
261     unchanged.
262
263  component_size
264     For arrays with data redundancy (i.e. not raid0, linear, faulty,
265     multipath), all components must be the same size - or at least
266     there must a size that they all provide space for.  This is a key
267     part or the geometry of the array.  It is measured in sectors
268     and can be read from here.  Writing to this value may resize
269     the array if the personality supports it (raid1, raid5, raid6),
270     and if the component drives are large enough.
271
272  metadata_version
273     This indicates the format that is being used to record metadata
274     about the array.  It can be 0.90 (traditional format), 1.0, 1.1,
275     1.2 (newer format in varying locations) or ``none`` indicating that
276     the kernel isn't managing metadata at all.
277     Alternately it can be ``external:`` followed by a string which
278     is set by user-space.  This indicates that metadata is managed
279     by a user-space program.  Any device failure or other event that
280     requires a metadata update will cause array activity to be
281     suspended until the event is acknowledged.
282
283  resync_start
284     The point at which resync should start.  If no resync is needed,
285     this will be a very large number (or ``none`` since 2.6.30-rc1).  At
286     array creation it will default to 0, though starting the array as
287     ``clean`` will set it much larger.
288
289  new_dev
290     This file can be written but not read.  The value written should
291     be a block device number as major:minor.  e.g. 8:0
292     This will cause that device to be attached to the array, if it is
293     available.  It will then appear at md/dev-XXX (depending on the
294     name of the device) and further configuration is then possible.
295
296  safe_mode_delay
297     When an md array has seen no write requests for a certain period
298     of time, it will be marked as ``clean``.  When another write
299     request arrives, the array is marked as ``dirty`` before the write
300     commences.  This is known as ``safe_mode``.
301     The ``certain period`` is controlled by this file which stores the
302     period as a number of seconds.  The default is 200msec (0.200).
303     Writing a value of 0 disables safemode.
304
305  array_state
306     This file contains a single word which describes the current
307     state of the array.  In many cases, the state can be set by
308     writing the word for the desired state, however some states
309     cannot be explicitly set, and some transitions are not allowed.
310
311     Select/poll works on this file.  All changes except between
312     Active_idle and active (which can be frequent and are not
313     very interesting) are notified.  active->active_idle is
314     reported if the metadata is externally managed.
315
316     clear
317         No devices, no size, no level
318
319         Writing is equivalent to STOP_ARRAY ioctl
320
321     inactive
322         May have some settings, but array is not active
323         all IO results in error
324
325         When written, doesn't tear down array, but just stops it
326
327     suspended (not supported yet)
328         All IO requests will block. The array can be reconfigured.
329
330         Writing this, if accepted, will block until array is quiescent
331
332     readonly
333         no resync can happen.  no superblocks get written.
334
335         Write requests fail
336
337     read-auto
338         like readonly, but behaves like ``clean`` on a write request.
339
340     clean
341         no pending writes, but otherwise active.
342
343         When written to inactive array, starts without resync
344
345         If a write request arrives then
346         if metadata is known, mark ``dirty`` and switch to ``active``.
347         if not known, block and switch to write-pending
348
349         If written to an active array that has pending writes, then fails.
350     active
351         fully active: IO and resync can be happening.
352         When written to inactive array, starts with resync
353
354     write-pending
355         clean, but writes are blocked waiting for ``active`` to be written.
356
357     active-idle
358         like active, but no writes have been seen for a while (safe_mode_delay).
359
360  consistency_policy
361     This indicates how the array maintains consistency in case of unexpected
362     shutdown. It can be:
363
364     none
365       Array has no redundancy information, e.g. raid0, linear.
366
367     resync
368       Full resync is performed and all redundancy is regenerated when the
369       array is started after unclean shutdown.
370
371     bitmap
372       Resync assisted by a write-intent bitmap.
373
374     journal
375       For raid4/5/6, journal device is used to log transactions and replay
376       after unclean shutdown.
377
378     ppl
379       For raid5 only, Partial Parity Log is used to close the write hole and
380       eliminate resync.
381
382     The accepted values when writing to this file are ``ppl`` and ``resync``,
383     used to enable and disable PPL.
384
385  uuid
386     This indicates the UUID of the array in the following format:
387     xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
388
389  bitmap_type
390     [RW] When read, this file will display the current and available
391     bitmap for this array. The currently active bitmap will be enclosed
392     in [] brackets. Writing an bitmap name or ID to this file will switch
393     control of this array to that new bitmap. Note that writing a new
394     bitmap for created array is forbidden.
395
396     none
397         No bitmap
398     bitmap
399         The default internal bitmap
400     llbitmap
401         The lockless internal bitmap
402
403If bitmap_type is not none, then additional bitmap attributes bitmap/xxx or
404llbitmap/xxx will be created after md device KOBJ_CHANGE event.
405
406If bitmap_type is bitmap, then the md device will also contain:
407
408  bitmap/location
409     This indicates where the write-intent bitmap for the array is
410     stored.
411
412     It can be one of ``none``, ``file`` or ``[+-]N``.
413     ``file`` may later be extended to ``file:/file/name``
414     ``[+-]N`` means that many sectors from the start of the metadata.
415
416     This is replicated on all devices.  For arrays with externally
417     managed metadata, the offset is from the beginning of the
418     device.
419
420  bitmap/chunksize
421     The size, in bytes, of the chunk which will be represented by a
422     single bit.  For RAID456, it is a portion of an individual
423     device. For RAID10, it is a portion of the array.  For RAID1, it
424     is both (they come to the same thing).
425
426  bitmap/time_base
427     The time, in seconds, between looking for bits in the bitmap to
428     be cleared. In the current implementation, a bit will be cleared
429     between 2 and 3 times ``time_base`` after all the covered blocks
430     are known to be in-sync.
431
432  bitmap/backlog
433     When write-mostly devices are active in a RAID1, write requests
434     to those devices proceed in the background - the filesystem (or
435     other user of the device) does not have to wait for them.
436     ``backlog`` sets a limit on the number of concurrent background
437     writes.  If there are more than this, new writes will by
438     synchronous.
439
440  bitmap/metadata
441     This can be either ``internal`` or ``external``.
442
443     ``internal``
444       is the default and means the metadata for the bitmap
445       is stored in the first 256 bytes of the allocated space and is
446       managed by the md module.
447
448     ``external``
449       means that bitmap metadata is managed externally to
450       the kernel (i.e. by some userspace program)
451
452  bitmap/can_clear
453     This is either ``true`` or ``false``.  If ``true``, then bits in the
454     bitmap will be cleared when the corresponding blocks are thought
455     to be in-sync.  If ``false``, bits will never be cleared.
456     This is automatically set to ``false`` if a write happens on a
457     degraded array, or if the array becomes degraded during a write.
458     When metadata is managed externally, it should be set to true
459     once the array becomes non-degraded, and this fact has been
460     recorded in the metadata.
461
462If bitmap_type is llbitmap, then the md device will also contain:
463
464  llbitmap/bits
465     This is read-only, show status of bitmap bits, the number of each
466     value.
467
468  llbitmap/metadata
469     This is read-only, show bitmap metadata, include chunksize, chunkshift,
470     chunks, offset and daemon_sleep.
471
472  llbitmap/daemon_sleep
473     This is read-write, time in seconds that daemon function will be
474     triggered to clear dirty bits.
475
476  llbitmap/barrier_idle
477     This is read-write, time in seconds that page barrier will be idled,
478     means dirty bits in the page will be cleared.
479
480As component devices are added to an md array, they appear in the ``md``
481directory as new directories named::
482
483      dev-XXX
484
485where ``XXX`` is a name that the kernel knows for the device, e.g. hdb1.
486Each directory contains:
487
488      block
489        a symlink to the block device in /sys/block, e.g.::
490
491	     /sys/block/md0/md/dev-hdb1/block -> ../../../../block/hdb/hdb1
492
493      super
494        A file containing an image of the superblock read from, or
495        written to, that device.
496
497      state
498	A file recording the current state of the device in the array
499	which can be a comma separated list of:
500
501	      faulty
502			device has been kicked from active use due to
503			a detected fault, or it has unacknowledged bad
504			blocks
505
506	      in_sync
507			device is a fully in-sync member of the array
508
509	      writemostly
510			device will only be subject to read
511			requests if there are no other options.
512
513			This applies only to raid1 arrays.
514
515	      blocked
516			device has failed, and the failure hasn't been
517			acknowledged yet by the metadata handler.
518
519			Writes that would write to this device if
520			it were not faulty are blocked.
521
522	      spare
523			device is working, but not a full member.
524
525			This includes spares that are in the process
526			of being recovered to
527
528	      write_error
529			device has ever seen a write error.
530
531	      want_replacement
532			device is (mostly) working but probably
533			should be replaced, either due to errors or
534			due to user request.
535
536	      replacement
537			device is a replacement for another active
538			device with same raid_disk.
539
540
541	This list may grow in future.
542
543	This can be written to.
544
545	Writing ``faulty``  simulates a failure on the device.
546
547	Writing ``remove`` removes the device from the array.
548
549	Writing ``writemostly`` sets the writemostly flag.
550
551	Writing ``-writemostly`` clears the writemostly flag.
552
553	Writing ``blocked`` sets the ``blocked`` flag.
554
555	Writing ``-blocked`` clears the ``blocked`` flags and allows writes
556	to complete and possibly simulates an error.
557
558	Writing ``in_sync`` sets the in_sync flag.
559
560	Writing ``write_error`` sets writeerrorseen flag.
561
562	Writing ``-write_error`` clears writeerrorseen flag.
563
564	Writing ``want_replacement`` is allowed at any time except to a
565	replacement device or a spare.  It sets the flag.
566
567	Writing ``-want_replacement`` is allowed at any time.  It clears
568	the flag.
569
570	Writing ``replacement`` or ``-replacement`` is only allowed before
571	starting the array.  It sets or clears the flag.
572
573
574	This file responds to select/poll. Any change to ``faulty``
575	or ``blocked`` causes an event.
576
577      errors
578	An approximate count of read errors that have been detected on
579	this device but have not caused the device to be evicted from
580	the array (either because they were corrected or because they
581	happened while the array was read-only).  When using version-1
582	metadata, this value persists across restarts of the array.
583
584	This value can be written while assembling an array thus
585	providing an ongoing count for arrays with metadata managed by
586	userspace.
587
588      slot
589        This gives the role that the device has in the array.  It will
590	either be ``none`` if the device is not active in the array
591        (i.e. is a spare or has failed) or an integer less than the
592	``raid_disks`` number for the array indicating which position
593	it currently fills.  This can only be set while assembling an
594	array.  A device for which this is set is assumed to be working.
595
596      offset
597        This gives the location in the device (in sectors from the
598        start) where data from the array will be stored.  Any part of
599        the device before this offset is not touched, unless it is
600        used for storing metadata (Formats 1.1 and 1.2).
601
602      size
603        The amount of the device, after the offset, that can be used
604        for storage of data.  This will normally be the same as the
605	component_size.  This can be written while assembling an
606        array.  If a value less than the current component_size is
607        written, it will be rejected.
608
609      recovery_start
610        When the device is not ``in_sync``, this records the number of
611	sectors from the start of the device which are known to be
612	correct.  This is normally zero, but during a recovery
613	operation it will steadily increase, and if the recovery is
614	interrupted, restoring this value can cause recovery to
615	avoid repeating the earlier blocks.  With v1.x metadata, this
616	value is saved and restored automatically.
617
618	This can be set whenever the device is not an active member of
619	the array, either before the array is activated, or before
620	the ``slot`` is set.
621
622	Setting this to ``none`` is equivalent to setting ``in_sync``.
623	Setting to any other value also clears the ``in_sync`` flag.
624
625      bad_blocks
626	This gives the list of all known bad blocks in the form of
627	start address and length (in sectors respectively). If output
628	is too big to fit in a page, it will be truncated. Writing
629	``sector length`` to this file adds new acknowledged (i.e.
630	recorded to disk safely) bad blocks.
631
632      unacknowledged_bad_blocks
633	This gives the list of known-but-not-yet-saved-to-disk bad
634	blocks in the same form of ``bad_blocks``. If output is too big
635	to fit in a page, it will be truncated. Writing to this file
636	adds bad blocks without acknowledging them. This is largely
637	for testing.
638
639      ppl_sector, ppl_size
640        Location and size (in sectors) of the space used for Partial Parity Log
641        on this device.
642
643
644An active md device will also contain an entry for each active device
645in the array.  These are named::
646
647    rdNN
648
649where ``NN`` is the position in the array, starting from 0.
650So for a 3 drive array there will be rd0, rd1, rd2.
651These are symbolic links to the appropriate ``dev-XXX`` entry.
652Thus, for example::
653
654       cat /sys/block/md*/md/rd*/state
655
656will show ``in_sync`` on every line.
657
658
659
660Active md devices for levels that support data redundancy (1,4,5,6,10)
661also have
662
663   sync_action
664     a text file that can be used to monitor and control the rebuild
665     process.  It contains one word which can be one of:
666
667       resync
668		redundancy is being recalculated after unclean
669                shutdown or creation
670
671       recover
672		a hot spare is being built to replace a
673		failed/missing device
674
675       idle
676		nothing is happening
677       check
678		A full check of redundancy was requested and is
679                happening.  This reads all blocks and checks
680                them. A repair may also happen for some raid
681                levels.
682
683       repair
684		A full check and repair is happening.  This is
685		similar to ``resync``, but was requested by the
686                user, and the write-intent bitmap is NOT used to
687		optimise the process.
688
689      This file is writable, and each of the strings that could be
690      read are meaningful for writing.
691
692	``idle`` will stop an active resync/recovery etc.  There is no
693	guarantee that another resync/recovery may not be automatically
694	started again, though some event will be needed to trigger
695	this.
696
697	``resync`` or ``recovery`` can be used to restart the
698        corresponding operation if it was stopped with ``idle``.
699
700	``check`` and ``repair`` will start the appropriate process
701	providing the current state is ``idle``.
702
703      This file responds to select/poll.  Any important change in the value
704      triggers a poll event.  Sometimes the value will briefly be
705      ``recover`` if a recovery seems to be needed, but cannot be
706      achieved. In that case, the transition to ``recover`` isn't
707      notified, but the transition away is.
708
709   degraded
710      This contains a count of the number of devices by which the
711      arrays is degraded.  So an optimal array will show ``0``.  A
712      single failed/missing drive will show ``1``, etc.
713
714      This file responds to select/poll, any increase or decrease
715      in the count of missing devices will trigger an event.
716
717   mismatch_count
718      When performing ``check`` and ``repair``, and possibly when
719      performing ``resync``, md will count the number of errors that are
720      found.  The count in ``mismatch_cnt`` is the number of sectors
721      that were re-written, or (for ``check``) would have been
722      re-written.  As most raid levels work in units of pages rather
723      than sectors, this may be larger than the number of actual errors
724      by a factor of the number of sectors in a page.
725
726   bitmap_set_bits
727      If the array has a write-intent bitmap, then writing to this
728      attribute can set bits in the bitmap, indicating that a resync
729      would need to check the corresponding blocks. Either individual
730      numbers or start-end pairs can be written.  Multiple numbers
731      can be separated by a space.
732
733      Note that the numbers are ``bit`` numbers, not ``block`` numbers.
734      They should be scaled by the bitmap_chunksize.
735
736   sync_speed_min, sync_speed_max
737     This are similar to ``/proc/sys/dev/raid/speed_limit_{min,max}``
738     however they only apply to the particular array.
739
740     If no value has been written to these, or if the word ``system``
741     is written, then the system-wide value is used.  If a value,
742     in kibibytes-per-second is written, then it is used.
743
744     When the files are read, they show the currently active value
745     followed by ``(local)`` or ``(system)`` depending on whether it is
746     a locally set or system-wide value.
747
748   sync_completed
749     This shows the number of sectors that have been completed of
750     whatever the current sync_action is, followed by the number of
751     sectors in total that could need to be processed.  The two
752     numbers are separated by a ``/``  thus effectively showing one
753     value, a fraction of the process that is complete.
754
755     A ``select`` on this attribute will return when resync completes,
756     when it reaches the current sync_max (below) and possibly at
757     other times.
758
759   sync_speed
760     This shows the current actual speed, in K/sec, of the current
761     sync_action.  It is averaged over the last 30 seconds.
762
763   suspend_lo, suspend_hi
764     The two values, given as numbers of sectors, indicate a range
765     within the array where IO will be blocked.  This is currently
766     only supported for raid4/5/6.
767
768   sync_min, sync_max
769     The two values, given as numbers of sectors, indicate a range
770     within the array where ``check``/``repair`` will operate. Must be
771     a multiple of chunk_size. When it reaches ``sync_max`` it will
772     pause, rather than complete.
773     You can use ``select`` or ``poll`` on ``sync_completed`` to wait for
774     that number to reach sync_max.  Then you can either increase
775     ``sync_max``, or can write ``idle`` to ``sync_action``.
776
777     The value of ``max`` for ``sync_max`` effectively disables the limit.
778     When a resync is active, the value can only ever be increased,
779     never decreased.
780     The value of ``0`` is the minimum for ``sync_min``.
781
782
783
784Each active md device may also have attributes specific to the
785personality module that manages it.
786These are specific to the implementation of the module and could
787change substantially if the implementation changes.
788
789These currently include:
790
791  stripe_cache_size  (currently raid5 only)
792      number of entries in the stripe cache.  This is writable, but
793      there are upper and lower limits (32768, 17).  Default is 256.
794
795  strip_cache_active (currently raid5 only)
796      number of active entries in the stripe cache
797
798  preread_bypass_threshold (currently raid5 only)
799      number of times a stripe requiring preread will be bypassed by
800      a stripe that does not require preread.  For fairness defaults
801      to 1.  Setting this to 0 disables bypass accounting and
802      requires preread stripes to wait until all full-width stripe-
803      writes are complete.  Valid values are 0 to stripe_cache_size.
804
805  journal_mode (currently raid5 only)
806      The cache mode for raid5. raid5 could include an extra disk for
807      caching. The mode can be "write-through" or "write-back". The
808      default is "write-through".
809
810  ppl_write_hint
811      NVMe stream ID to be set for each PPL write request.
812