1MDADM(8) System Manager's Manual MDADM(8)
2
6 mdadm - manage MD devices aka Linux Software Raid.
7
10 mdadm [mode] <raiddevice> [options] <component-devices>
11
14 RAID devices are virtual devices created from two or more real block
15 devices. This allows multiple devices (typically disk drives or parti‐
16 tions thereof) to be combined into a single device to hold (for exam‐
17 ple) a single filesystem. Some RAID levels include redundancy and so
18 can survive some degree of device failure.
19 Linux Software RAID devices are implemented through the md (Multiple
21 Devices) device driver.
22 Currently, Linux supports LINEAR md devices, RAID0 (striping), RAID1
24 (mirroring), RAID4, RAID5, RAID6, RAID10, MULTIPATH, and FAULTY.
25 MULTIPATH is not a Software RAID mechanism, but does involve multiple
27 devices. For MULTIPATH each device is a path to one common physical
28 storage device.
29 FAULTY is also not true RAID, and it only involves one device. It pro‐
31 vides a layer over a true device that can be used to inject faults.
32
36 mdadm has several major modes of operation:
37 Assemble
39 Assemble the parts of a previously created array into an active
40 array. Components can be explicitly given or can be searched
41 for. mdadm checks that the components do form a bona fide
42 array, and can, on request, fiddle superblock information so as
43 to assemble a faulty array.
44 Build Build an array that doesn't have per-device superblocks. For
47 these sorts of arrays, mdadm cannot differentiate between ini‐
48 tial creation and subsequent assembly of an array. It also can‐
49 not perform any checks that appropriate devices have been
50 requested. Because of this, the Build mode should only be used
51 together with a complete understanding of what you are doing.
52 Create Create a new array with per-device superblocks.
55 Follow or Monitor
58 Monitor one or more md devices and act on any state changes.
59 This is only meaningful for raid1, 4, 5, 6, 10 or multipath
60 arrays as only these have interesting state. raid0 or linear
61 never have missing, spare, or failed drives, so there is nothing
62 to monitor.
63 Grow Grow (or shrink) an array, or otherwise reshape it in some way.
66 Currently supported growth options including changing the active
67 size of component devices in RAID level 1/4/5/6 and changing the
68 number of active devices in RAID1/5/6.
69 Incremental Assembly
72 Add a single device to an appropriate array. If the addition of
73 the device makes the array runnable, the array will be started.
74 This provides a convenient interface to a hot-plug system. As
75 each device is detected, mdadm has a chance to include it in
76 some array as appropriate.
77 Manage This is for doing things to specific components of an array such
80 as adding new spares and removing faulty devices.
81 Misc This is an 'everything else' mode that supports operations on
84 active arrays, operations on component devices such as erasing
85 old superblocks, and information gathering operations.
86 Auto-detect
89 This mode does not act on a specific device or array, but rather
90 it requests the Linux Kernel to activate any auto-detected
91 arrays.
92
95 -A, --assemble
96 Assemble a pre-existing array.
97 -B, --build
100 Build a legacy array without superblocks.
101 -C, --create
104 Create a new array.
105 -F, --follow, --monitor
108 Select Monitor mode.
109 -G, --grow
112 Change the size or shape of an active array.
113 -I, --incremental
116 Add a single device into an appropriate array, and possibly
117 start the array.
118 --auto-detect
121 Request that the kernel starts any auto-detected arrays. This
122 can only work if md is compiled into the kernel — not if it is a
123 module. Arrays can be auto-detected by the kernel if all the
124 components are in primary MS-DOS partitions with partition type
125 FD. In-kernel autodetect is not recommended for new installa‐
126 tions. Using mdadm to detect and assemble arrays — possibly in
127 an initrd — is substantially more flexible and should be pre‐
128 ferred.
129 If a device is given before any options, or if the first option is
132 --add, --fail, or --remove, then the MANAGE mode is assume. Anything
133 other than these will cause the Misc mode to be assumed.
134
137 -h, --help
138 Display general help message or, after one of the above options,
139 a mode specific help message.
140 --help-options
143 Display more detailed help about command line parsing and some
144 commonly used options.
145 -V, --version
148 Print version information for mdadm.
149 -v, --verbose
152 Be more verbose about what is happening. This can be used twice
153 to be extra-verbose. The extra verbosity currently only affects
154 --detail --scan and --examine --scan.
155 -q, --quiet
158 Avoid printing purely informative messages. With this, mdadm
159 will be silent unless there is something really important to
160 report.
161 -b, --brief
164 Be less verbose. This is used with --detail and --examine.
165 Using --brief with --verbose gives an intermediate level of ver‐
166 bosity.
167 -f, --force
170 Be more forceful about certain operations. See the various
171 modes of the exact meaning of this option in different contexts.
172 -c, --config=
175 Specify the config file. Default is to use /etc/mdadm.conf, or
176 if that is missing, then /etc/mdadm/mdadm.conf. If the config
177 file given is partitions then nothing will be read, but mdadm
178 will act as though the config file contained exactly DEVICE par‐
179 titions and will read /proc/partitions to find a list of devices
180 to scan. If the word none is given for the config file, then
181 mdadm will act as though the config file were empty.
182 -s, --scan
185 scan config file or /proc/mdstat for missing information. In
186 general, this option gives mdadm permission to get any missing
187 information, like component devices, array devices, array iden‐
188 tities, and alert destination from the configuration file:
189 /etc/mdadm.conf. One exception is MISC mode when using --detail
190 or --stop in which case --scan says to get a list of array
191 devices from /proc/mdstat.
192 -e , --metadata=
195 Declare the style of superblock (raid metadata) to be used. The
196 default is 0.90 for --create, and to guess for other operations.
197 The default can be overridden by setting the metadata value for
198 the CREATE keyword in mdadm.conf.
199 Options are:
201 0, 0.90, default
203 Use the original 0.90 format superblock. This format
204 limits arrays to 28 componenet devices and limits compo‐
205 nent devices of levels 1 and greater to 2 terabytes.
206 1, 1.0, 1.1, 1.2
208 Use the new version-1 format superblock. This has few
209 restrictions. The different subversion store the
210 superblock at different locations on the device, either
211 at the end (for 1.0), at the start (for 1.1) or 4K from
212 the start (for 1.2).
213 --homehost=
216 This will override any HOMEHOST setting in the config file and
217 provides the identify of the host which should be considered the
218 home for any arrays.
219 When creating an array, the homehost will be recorded in the
221 superblock. For version-1 superblocks, it will be prefixed to
222 the array name. For version-0.90 superblocks part of the SHA1
223 hash of the hostname will be stored in the later half of the
224 UUID.
225 When reporting information about an array, any array which is
227 tagged for the given homehost will be reported as such.
228 When using Auto-Assemble, only arrays tagged for the given home‐
230 host will be assembled.
231
234 -n, --raid-devices=
235 Specify the number of active devices in the array. This, plus
236 the number of spare devices (see below) must equal the number of
237 component-devices (including "missing" devices) that are listed
238 on the command line for --create. Setting a value of 1 is prob‐
239 ably a mistake and so requires that --force be specified first.
240 A value of 1 will then be allowed for linear, multipath, raid0
241 and raid1. It is never allowed for raid4 or raid5.
242 This number can only be changed using --grow for RAID1, RAID5
243 and RAID6 arrays, and only on kernels which provide necessary
244 support.
245 -x, --spare-devices=
248 Specify the number of spare (eXtra) devices in the initial
249 array. Spares can also be added and removed later. The number
250 of component devices listed on the command line must equal the
251 number of raid devices plus the number of spare devices.
252 -z, --size=
256 Amount (in Kibibytes) of space to use from each drive in
257 RAID1/4/5/6. This must be a multiple of the chunk size, and
258 must leave about 128Kb of space at the end of the drive for the
259 RAID superblock. If this is not specified (as it normally is
260 not) the smallest drive (or partition) sets the size, though if
261 there is a variance among the drives of greater than 1%, a warn‐
262 ing is issued.
263 This value can be set with --grow for RAID level 1/4/5/6. If the
265 array was created with a size smaller than the currently active
266 drives, the extra space can be accessed using --grow. The size
267 can be given as max which means to choose the largest size that
268 fits on all current drives.
269 -c, --chunk=
272 Specify chunk size of kibibytes. The default is 64.
273 --rounding=
276 Specify rounding factor for linear array (==chunk size)
277 -l, --level=
280 Set raid level. When used with --create, options are: linear,
281 raid0, 0, stripe, raid1, 1, mirror, raid4, 4, raid5, 5, raid6,
282 6, raid10, 10, multipath, mp, faulty. Obviously some of these
283 are synonymous.
284 When used with --build, only linear, stripe, raid0, 0, raid1,
286 multipath, mp, and faulty are valid.
287 Not yet supported with --grow.
289 -p, --layout=
292 This option configures the fine details of data layout for
293 raid5, and raid10 arrays, and controls the failure modes for
294 faulty.
295 The layout of the raid5 parity block can be one of left-asymmet‐
297 ric, left-symmetric, right-asymmetric, right-symmetric, la, ra,
298 ls, rs. The default is left-symmetric.
299 When setting the failure mode for faulty the options are:
301 write-transient, wt, read-transient, rt, write-persistent, wp,
302 read-persistent, rp, write-all, read-fixable, rf, clear, flush,
303 none.
304 Each mode can be followed by a number which is used as a period
306 between fault generation. Without a number, the fault is gener‐
307 ated once on the first relevant request. With a number, the
308 fault will be generated after that many request, and will con‐
309 tinue to be generated every time the period elapses.
310 Multiple failure modes can be current simultaneously by using
312 the --grow option to set subsequent failure modes.
313 "clear" or "none" will remove any pending or periodic failure
315 modes, and "flush" will clear any persistent faults.
316 To set the parity with --grow, the level of the array ("faulty")
318 must be specified before the fault mode is specified.
319 Finally, the layout options for RAID10 are one of 'n', 'o' or
321 'f' followed by a small number. The default is 'n2'.
322 n signals 'near' copies. Multiple copies of one data block are
324 at similar offsets in different devices.
325 o signals 'offset' copies. Rather than the chunks being dupli‐
327 cated within a stripe, whole stripes are duplicated but are
328 rotated by one device so duplicate blocks are on different
329 devices. Thus subsequent copies of a block are in the next
330 drive, and are one chunk further down.
331 f signals 'far' copies (multiple copies have very different off‐
333 sets). See md(4) for more detail about 'near' and 'far'.
334 The number is the number of copies of each datablock. 2 is nor‐
336 mal, 3 can be useful. This number can be at most equal to the
337 number of devices in the array. It does not need to divide
338 evenly into that number (e.g. it is perfectly legal to have an
339 'n2' layout for an array with an odd number of devices).
340 --parity=
343 same as --layout (thus explaining the p of -p).
344 -b, --bitmap=
347 Specify a file to store a write-intent bitmap in. The file
348 should not exist unless --force is also given. The same file
349 should be provided when assembling the array. If the word
350 internal is given, then the bitmap is stored with the metadata
351 on the array, and so is replicated on all devices. If the word
352 none is given with --grow mode, then any bitmap that is present
353 is removed.
354 To help catch typing errors, the filename must contain at least
356 one slash ('/') if it is a real file (not 'internal' or 'none').
357 Note: external bitmaps are only known to work on ext2 and ext3.
359 Storing bitmap files on other filesystems may result in serious
360 problems.
361 --bitmap-chunk=
364 Set the chunksize of the bitmap. Each bit corresponds to that
365 many Kilobytes of storage. When using a file based bitmap, the
366 default is to use the smallest size that is atleast 4 and
367 requires no more than 2^21 chunks. When using an internal bit‐
368 map, the chunksize is automatically determined to make best use
369 of available space.
370 -W, --write-mostly
374 subsequent devices lists in a --build, --create, or --add com‐
375 mand will be flagged as 'write-mostly'. This is valid for RAID1
376 only and means that the 'md' driver will avoid reading from
377 these devices if at all possible. This can be useful if mirror‐
378 ing over a slow link.
379 --write-behind=
382 Specify that write-behind mode should be enabled (valid for
383 RAID1 only). If an argument is specified, it will set the maxi‐
384 mum number of outstanding writes allowed. The default value is
385 256. A write-intent bitmap is required in order to use write-
386 behind mode, and write-behind is only attempted on drives marked
387 as write-mostly.
388 --assume-clean
391 Tell mdadm that the array pre-existed and is known to be clean.
392 It can be useful when trying to recover from a major failure as
393 you can be sure that no data will be affected unless you actu‐
394 ally write to the array. It can also be used when creating a
395 RAID1 or RAID10 if you want to avoid the initial resync, however
396 this practice — while normally safe — is not recommended. Use
397 this ony if you really know what you are doing.
398 --backup-file=
401 This is needed when --grow is used to increase the number of
402 raid-devices in a RAID5 if there are no spare devices avail‐
403 able. See the section below on RAID_DEVICE CHANGES. The file
404 should be stored on a separate device, not on the raid array
405 being reshaped.
406 -N, --name=
409 Set a name for the array. This is currently only effective when
410 creating an array with a version-1 superblock. The name is a
411 simple textual string that can be used to identify array compo‐
412 nents when assembling.
413 -R, --run
416 Insist that mdadm run the array, even if some of the components
417 appear to be active in another array or filesystem. Normally
418 mdadm will ask for confirmation before including such components
419 in an array. This option causes that question to be suppressed.
420 -f, --force
423 Insist that mdadm accept the geometry and layout specified with‐
424 out question. Normally mdadm will not allow creation of an
425 array with only one device, and will try to create a raid5 array
426 with one missing drive (as this makes the initial resync work
427 faster). With --force, mdadm will not try to be so clever.
428 -a, --auto{=no,yes,md,mdp,part,p}{NN}
431 Instruct mdadm to create the device file if needed, possibly
432 allocating an unused minor number. "md" causes a non-partition‐
433 able array to be used. "mdp", "part" or "p" causes a partition‐
434 able array (2.6 and later) to be used. "yes" requires the named
435 md device to have a 'standard' format, and the type and minor
436 number will be determined from this. See DEVICE NAMES below.
437 The argument can also come immediately after "-a". e.g. "-ap".
439 If --auto is not given on the command line or in the config
441 file, then the default will be --auto=yes.
442 If --scan is also given, then any auto= entries in the config
444 file will override the --auto instruction given on the command
445 line.
446 For partitionable arrays, mdadm will create the device file for
448 the whole array and for the first 4 partitions. A different
449 number of partitions can be specified at the end of this option
450 (e.g. --auto=p7). If the device name ends with a digit, the
451 partition names add a 'p', and a number, e.g. "/dev/home1p3".
452 If there is no trailing digit, then the partition names just
453 have a number added, e.g. "/dev/scratch3".
454 If the md device name is in a 'standard' format as described in
456 DEVICE NAMES, then it will be created, if necessary, with the
457 appropriate number based on that name. If the device name is
458 not in one of these formats, then a unused minor number will be
459 allocated. The minor number will be considered unused if there
460 is no active array for that number, and there is no entry in
461 /dev for that number and with a non-standard name.
462 --symlink=no
465 Normally when --auto causes mdadm to create devices in /dev/md/
466 it will also create symlinks from /dev/ with names starting with
467 md or md_. Use --symlink=no to suppress this, or --symlink=yes
468 to enforce this even if it is suppressing mdadm.conf.
469
473 -u, --uuid=
474 uuid of array to assemble. Devices which don't have this uuid
475 are excluded
476 -m, --super-minor=
479 Minor number of device that array was created for. Devices
480 which don't have this minor number are excluded. If you create
481 an array as /dev/md1, then all superblocks will contain the
482 minor number 1, even if the array is later assembled as
483 /dev/md2.
484 Giving the literal word "dev" for --super-minor will cause mdadm
486 to use the minor number of the md device that is being assem‐
487 bled. e.g. when assembling /dev/md0, will look for super blocks
488 with a minor number of 0.
489 -N, --name=
492 Specify the name of the array to assemble. This must be the
493 name that was specified when creating the array. It must either
494 match then name stored in the superblock exactly, or it must
495 match with the current homehost is added to the start of the
496 given name.
497 -f, --force
500 Assemble the array even if some superblocks appear out-of-date
501 -R, --run
504 Attempt to start the array even if fewer drives were given than
505 were present last time the array was active. Normally if not
506 all the expected drives are found and --scan is not used, then
507 the array will be assembled but not started. With --run an
508 attempt will be made to start it anyway.
509 --no-degraded
512 This is the reverse of --run in that it inhibits the started if
513 array unless all expected drives are present. This is only
514 needed with --scan and can be used if you physical connections
515 to devices are not as reliable as you would like.
516 -a, --auto{=no,yes,md,mdp,part}
519 See this option under Create and Build options.
520 -b, --bitmap=
523 Specify the bitmap file that was given when the array was cre‐
524 ated. If an array has an internal bitmap, there is no need to
525 specify this when assembling the array.
526 --backup-file=
529 If --backup-file was used to grow the number of raid-devices in
530 a RAID5, and the system crashed during the critical section,
531 then the same --backup-file must be presented to --assemble to
532 allow possibly corrupted data to be restored.
533 -U, --update=
536 Update the superblock on each device while assembling the array.
537 The argument given to this flag can be one of sparc2.2, sum‐
538 maries, uuid, name, homehost, resync, byteorder, devicesize, or
539 super-minor.
540 The sparc2.2 option will adjust the superblock of an array what
542 was created on a Sparc machine running a patched 2.2 Linux ker‐
543 nel. This kernel got the alignment of part of the superblock
544 wrong. You can use the --examine --sparc2.2 option to mdadm to
545 see what effect this would have.
546 The super-minor option will update the preferred minor field on
548 each superblock to match the minor number of the array being
549 assembled. This can be useful if --examine reports a different
550 "Preferred Minor" to --detail. In some cases this update will
551 be performed automatically by the kernel driver. In particular
552 the update happens automatically at the first write to an array
553 with redundancy (RAID level 1 or greater) on a 2.6 (or later)
554 kernel.
555 The uuid option will change the uuid of the array. If a UUID is
557 given with the --uuid option that UUID will be used as a new
558 UUID and will NOT be used to help identify the devices in the
559 array. If no --uuid is given, a random UUID is chosen.
560 The name option will change the name of the array as stored in
562 the superblock. This is only supported for version-1
563 superblocks.
564 The homehost option will change the homehost as recorded in the
566 superblock. For version-0 superblocks, this is the same as
567 updating the UUID. For version-1 superblocks, this involves
568 updating the name.
569 The resync option will cause the array to be marked dirty mean‐
571 ing that any redundancy in the array (e.g. parity for raid5,
572 copies for raid1) may be incorrect. This will cause the raid
573 system to perform a "resync" pass to make sure that all redun‐
574 dant information is correct.
575 The byteorder option allows arrays to be moved between machines
577 with different byte-order. When assembling such an array for
578 the first time after a move, giving --update=byteorder will
579 cause mdadm to expect superblocks to have their byteorder
580 reversed, and will correct that order before assembling the
581 array. This is only valid with original (Version 0.90)
582 superblocks.
583 The summaries option will correct the summaries in the
585 superblock. That is the counts of total, working, active,
586 failed, and spare devices.
587 The devicesize will rarely be of use. It applies to version 1.1
589 and 1.2 metadata only (where the metadata is at the start of the
590 device) and is only useful when the component device has changed
591 size (typically become larger). The version 1 metadata records
592 the amount of the device that can be used to store data, so if a
593 device in a version 1.1 or 1.2 array becomes larger, the meta‐
594 data will still be visible, but the extra space will not. In
595 this case it might be useful to assemble the array with
596 --update=devicesize. This will cause mdadm to determine the
597 maximum usable amount of space on each device and update the
598 relevant field in the metadata.
599 --auto-update-homehost
602 This flag is only meaning with auto-assembly (see discussion
603 below). In that situation, if no suitable arrays are found for
604 this homehost, mdadm will recan for any arrays at all and will
605 assemble them and update the homehost to match the current host.
606
609 -a, --add
610 hot-add listed devices.
611 --re-add
614 re-add a device that was recently removed from an array.
615 -r, --remove
618 remove listed devices. They must not be active. i.e. they
619 should be failed or spare devices. As well as the name of a
620 device file (e.g. /dev/sda1) the words failed and detached can
621 be given to --remove. The first causes all failed device to be
622 removed. The second causes any device which is no longer con‐
623 nected to the system (i.e and open returns ENXIO) to be removed.
624 This will only succeed for devices that are spares or have
625 already been marked as failed.
626 -f, --fail
629 mark listed devices as faulty. As well as the name of a device
630 file, the word detached can be given. This will cause any
631 device that has been detached from the system to be marked as
632 failed. It can then be removed.
633 --set-faulty
636 same as --fail.
637 Each of these options require that the first device list is the array
640 to be acted upon and the remainder are component devices to be added,
641 removed, or marked as fault. Several different operations can be spec‐
642 ified for different devices, e.g.
643 mdadm /dev/md0 --add /dev/sda1 --fail /dev/sdb1 --remove /dev/sdb1
644 Each operation applies to all devices listed until the next operations.
645 If an array is using a write-intent bitmap, then devices which have
647 been removed can be re-added in a way that avoids a full reconstruction
648 but instead just updated the blocks that have changed since the device
649 was removed. For arrays with persistent metadata (superblocks) this is
650 done automatically. For arrays created with --build mdadm needs to be
651 told that this device we removed recently with --re-add.
652 Devices can only be removed from an array if they are not in active
654 use. i.e. that must be spares or failed devices. To remove an active
655 device, it must be marked as faulty first.
656
659 -Q, --query
660 Examine a device to see (1) if it is an md device and (2) if it
661 is a component of an md array. Information about what is dis‐
662 covered is presented.
663 -D, --detail
666 Print detail of one or more md devices.
667 -Y, --export
670 When used with --detail, output will be formatted as key=value
671 pairs for easy import into the environment.
672 -E, --examine
675 Print content of md superblock on device(s).
676 --sparc2.2
678 If an array was created on a 2.2 Linux kernel patched with RAID
679 support, the superblock will have been created incorrectly, or
680 at least incompatibly with 2.4 and later kernels. Using the
681 --sparc2.2 flag with --examine will fix the superblock before
682 displaying it. If this appears to do the right thing, then the
683 array can be successfully assembled using --assemble
684 --update=sparc2.2.
685 -X, --examine-bitmap
688 Report information about a bitmap file.
689 -R, --run
692 start a partially built array.
693 -S, --stop
696 deactivate array, releasing all resources.
697 -o, --readonly
700 mark array as readonly.
701 -w, --readwrite
704 mark array as readwrite.
705 --zero-superblock
708 If the device contains a valid md superblock, the block is over‐
709 written with zeros. With --force the block where the superblock
710 would be is overwritten even if it doesn't appear to be valid.
711 -t, --test
714 When used with --detail, the exit status of mdadm is set to
715 reflect the status of the device.
716 -W, --wait
719 For each md device given, wait for any resync, recovery, or
720 reshape activity to finish before returning. mdadm will return
721 with success if it actually waited for every device listed, oth‐
722 erwise it will return failure.
723
726 --rebuild-map, -r
727 Rebuild the map file (/var/run/mdadm/map) that mdadm uses to
728 help track which arrays are currently being assembled.
729 --run, -R
732 Run any array assembled as soon as a minimal number of devices
733 are available, rather than waiting until all expected devices
734 are present.
735 --scan, -s
738 Only meaningful with -R this will scan the map file for arrays
739 that are being incrementally assembled and will try to start any
740 that are not already started. If any such array is listed in
741 mdadm.conf as requiring an external bitmap, that bitmap will be
742 attached first.
743
746 -m, --mail
747 Give a mail address to send alerts to.
748 -p, --program, --alert
751 Give a program to be run whenever an event is detected.
752 -y, --syslog
755 Cause all events to be reported through 'syslog'. The messages
756 have facility of 'daemon' and varying priorities.
757 -d, --delay
760 Give a delay in seconds. mdadm polls the md arrays and then
761 waits this many seconds before polling again. The default is 60
762 seconds.
763 -f, --daemonise
766 Tell mdadm to run as a background daemon if it decides to moni‐
767 tor anything. This causes it to fork and run in the child, and
768 to disconnect form the terminal. The process id of the child is
769 written to stdout. This is useful with --scan which will only
770 continue monitoring if a mail address or alert program is found
771 in the config file.
772 -i, --pid-file
775 When mdadm is running in daemon mode, write the pid of the dae‐
776 mon process to the specified file, instead of printing it on
777 standard output.
778 -1, --oneshot
781 Check arrays only once. This will generate NewArray events and
782 more significantly DegradedArray and SparesMissing events. Run‐
783 ning
784 mdadm --monitor --scan -1
785 from a cron script will ensure regular notification of any
786 degraded arrays.
787 -t, --test
790 Generate a TestMessage alert for every array found at startup.
791 This alert gets mailed and passed to the alert program. This
792 can be used for testing that alert message do get through suc‐
793 cessfully.
794
797 Usage: mdadm --assemble md-device options-and-component-devices...
798 Usage: mdadm --assemble --scan md-devices-and-options...
800 Usage: mdadm --assemble --scan options...
802 This usage assembles one or more raid arrays from pre-existing compo‐
805 nents. For each array, mdadm needs to know the md device, the identity
806 of the array, and a number of component-devices. These can be found in
807 a number of ways.
808 In the first usage example (without the --scan) the first device given
810 is the md device. In the second usage example, all devices listed are
811 treated as md devices and assembly is attempted. In the third (where
812 no devices are listed) all md devices that are listed in the configura‐
813 tion file are assembled.
814 If precisely one device is listed, but --scan is not given, then mdadm
816 acts as though --scan was given and identify information is extracted
817 from the configuration file.
818 The identity can be given with the --uuid option, with the
820 --super-minor option, can be found in the config file, or will be
821 taken from the super block on the first component-device listed on the
822 command line.
823 Devices can be given on the --assemble command line or in the config
825 file. Only devices which have an md superblock which contains the right
826 identity will be considered for any array.
827 The config file is only used if explicitly named with --config or
829 requested with (a possibly implicit) --scan. In the later case,
830 /etc/mdadm.conf is used.
831 If --scan is not given, then the config file will only be used to find
833 the identity of md arrays.
834 Normally the array will be started after it is assembled. However if
836 --scan is not given and insufficient drives were listed to start a com‐
837 plete (non-degraded) array, then the array is not started (to guard
838 against usage errors). To insist that the array be started in this
839 case (as may work for RAID1, 4, 5, 6, or 10), give the --run flag.
840 If the md device does not exist, then it will be created providing the
842 intent is clear. i.e. the name must be in a standard form, or the
843 --auto option must be given to clarify how and whether the device
844 should be created.
845 This can be useful for handling partitioned devices (which don't have a
847 stable device number — it can change after a reboot) and when using
848 "udev" to manage your /dev tree (udev cannot handle md devices because
849 of the unusual device initialisation conventions).
850 If the option to "auto" is "mdp" or "part" or (on the command line
852 only) "p", then mdadm will create a partitionable array, using the
853 first free one that is not in use, and does not already have an entry
854 in /dev (apart from numeric /dev/md* entries).
855 If the option to "auto" is "yes" or "md" or (on the command line) noth‐
857 ing, then mdadm will create a traditional, non-partitionable md array.
858 It is expected that the "auto" functionality will be used to create
860 device entries with meaningful names such as "/dev/md/home" or
861 "/dev/md/root", rather than names based on the numerical array number.
862 When using this option to create a partitionable array, the device
864 files for the first 4 partitions are also created. If a different num‐
865 ber is required it can be simply appended to the auto option. e.g.
866 "auto=part8". Partition names are created by appending a digit string
867 to the device name, with an intervening "p" if the device name ends
868 with a digit.
869 The --auto option is also available in Build and Create modes. As
871 those modes do not use a config file, the "auto=" config option does
872 not apply to these modes.
873 Auto Assembly
876 When --assemble is used with --scan and no devices are listed, mdadm
877 will first attempt to assemble all the arrays listed in the config
878 file.
879 If a homehost has been specified (either in the config file or on the
881 command line), mdadm will look further for possible arrays and will try
882 to assemble anything that it finds which is tagged as belonging to the
883 given homehost. This is the only situation where mdadm will assemble
884 arrays without being given specific device name or identify information
885 for the array.
886 If mdadm finds a consistent set of devices that look like they should
888 comprise an array, and if the superblock is tagged as belonging to the
889 given home host, it will automatically choose a device name and try to
890 assemble the array. If the array uses version-0.90 metadata, then the
891 minor number as recorded in the superblock is used to create a name in
892 /dev/md/ so for example /dev/md/3. If the array uses version-1 meta‐
893 data, then the name from the superblock is used to similarly create a
894 name in /dev/md. The name will have any 'host' prefix stripped first.
895 If mdadm cannot find any array for the given host at all, and if
897 --auto-update-homehost is given, then mdadm will search again for any
898 array (not just an array created for this host) and will assemble each
899 assuming --update=homehost. This will change the host tag in the
900 superblock so that on the next run, these arrays will be found without
901 the second pass. The intention of this feature is to support transi‐
902 tioning a set of md arrays to using homehost tagging.
903 The reason for requiring arrays to be tagged with the homehost for auto
905 assembly is to guard against problems that can arise when moving
906 devices from one host to another.
907
910 Usage: mdadm --build device --chunk=X --level=Y --raid-devices=Z
911 devices
912 This usage is similar to --create. The difference is that it creates
915 an array without a superblock. With these arrays there is no difference
916 between initially creating the array and subsequently assembling the
917 array, except that hopefully there is useful data there in the second
918 case.
919 The level may raid0, linear, multipath, or faulty, or one of their syn‐
921 onyms. All devices must be listed and the array will be started once
922 complete.
923
926 Usage: mdadm --create device --chunk=X --level=Y
927 --raid-devices=Z devices
928 This usage will initialise a new md array, associate some devices with
931 it, and activate the array.
932 If the --auto option is given (as described in more detail in the sec‐
934 tion on Assemble mode), then the md device will be created with a suit‐
935 able device number if necessary.
936 As devices are added, they are checked to see if they contain raid
938 superblocks or filesystems. They are also checked to see if the vari‐
939 ance in device size exceeds 1%.
940 If any discrepancy is found, the array will not automatically be run,
942 though the presence of a --run can override this caution.
943 To create a "degraded" array in which some devices are missing, simply
945 give the word "missing" in place of a device name. This will cause
946 mdadm to leave the corresponding slot in the array empty. For a RAID4
947 or RAID5 array at most one slot can be "missing"; for a RAID6 array at
948 most two slots. For a RAID1 array, only one real device needs to be
949 given. All of the others can be "missing".
950 When creating a RAID5 array, mdadm will automatically create a degraded
952 array with an extra spare drive. This is because building the spare
953 into a degraded array is in general faster than resyncing the parity on
954 a non-degraded, but not clean, array. This feature can be overridden
955 with the --force option.
956 When creating an array with version-1 metadata a name for the host is
958 required. If this is not given with the --name option, mdadm will
959 chose a name based on the last component of the name of the device
960 being created. So if /dev/md3 is being created, then the name 3 will
961 be chosen. If /dev/md/home is being created, then the name home will
962 be used.
963 A new array will normally get a randomly assigned 128bit UUID which is
965 very likely to be unique. If you have a specific need, you can choose
966 a UUID for the array by giving the --uuid= option. Be warned that cre‐
967 ating two arrays with the same UUID is a recipe for disaster. Also,
968 using --uuid= when creating a v0.90 array will silently override any
969 --homehost= setting.
970 The General Management options that are valid with --create are:
972 --run insist on running the array even if some devices look like they
974 might be in use.
975 --readonly
978 start the array readonly — not supported yet.
979
983 Usage: mdadm device options... devices...
984 This usage will allow individual devices in an array to be failed,
986 removed or added. It is possible to perform multiple operations with
987 on command. For example:
988 mdadm /dev/md0 -f /dev/hda1 -r /dev/hda1 -a /dev/hda1
989 will firstly mark /dev/hda1 as faulty in /dev/md0 and will then remove
990 it from the array and finally add it back in as a spare. However only
991 one md array can be affected by a single command.
992
995 Usage: mdadm options ... devices ...
996 MISC mode includes a number of distinct operations that operate on dis‐
998 tinct devices. The operations are:
999 --query
1001 The device is examined to see if it is (1) an active md array,
1002 or (2) a component of an md array. The information discovered
1003 is reported.
1004 --detail
1007 The device should be an active md device. mdadm will display a
1008 detailed description of the array. --brief or --scan will cause
1009 the output to be less detailed and the format to be suitable for
1010 inclusion in /etc/mdadm.conf. The exit status of mdadm will
1011 normally be 0 unless mdadm failed to get useful information
1012 about the device(s). However if the --test option is given,
1013 then the exit status will be:
1014 0 The array is functioning normally.
1016 1 The array has at least one failed device.
1018 2 The array has multiple failed devices such that it is
1020 unusable.
1021 4 There was an error while trying to get information about
1023 the device.
1024 --examine
1027 The device should be a component of an md array. mdadm will
1028 read the md superblock of the device and display the contents.
1029 If --brief is given, or --scan then multiple devices that are
1030 components of the one array are grouped together and reported in
1031 a single entry suitable for inclusion in /etc/mdadm.conf.
1032 Having --scan without listing any devices will cause all devices
1034 listed in the config file to be examined.
1035 --stop The devices should be active md arrays which will be deacti‐
1038 vated, as long as they are not currently in use.
1039 --run This will fully activate a partially assembled md array.
1042 --readonly
1045 This will mark an active array as read-only, providing that it
1046 is not currently being used.
1047 --readwrite
1050 This will change a readonly array back to being read/write.
1051 --scan For all operations except --examine, --scan will cause the oper‐
1054 ation to be applied to all arrays listed in /proc/mdstat. For
1055 --examine, --scan causes all devices listed in the config file
1056 to be examined.
1057
1061 Usage: mdadm --monitor options... devices...
1062 This usage causes mdadm to periodically poll a number of md arrays and
1065 to report on any events noticed. mdadm will never exit once it decides
1066 that there are arrays to be checked, so it should normally be run in
1067 the background.
1068 As well as reporting events, mdadm may move a spare drive from one
1070 array to another if they are in the same spare-group and if the desti‐
1071 nation array has a failed drive but no spares.
1072 If any devices are listed on the command line, mdadm will only monitor
1074 those devices. Otherwise all arrays listed in the configuration file
1075 will be monitored. Further, if --scan is given, then any other md
1076 devices that appear in /proc/mdstat will also be monitored.
1077 The result of monitoring the arrays is the generation of events. These
1079 events are passed to a separate program (if specified) and may be
1080 mailed to a given E-mail address.
1081 When passing event to program, the program is run once for each event
1083 and is given 2 or 3 command-line arguments. The first is the name of
1084 the event (see below). The second is the name of the md device which
1085 is affected, and the third is the name of a related device if relevant,
1086 such as a component device that has failed.
1087 If --scan is given, then a program or an E-mail address must be speci‐
1089 fied on the command line or in the config file. If neither are avail‐
1090 able, then mdadm will not monitor anything. Without --scan mdadm will
1091 continue monitoring as long as something was found to monitor. If no
1092 program or email is given, then each event is reported to stdout.
1093 The different events are:
1095 DeviceDisappeared
1098 An md array which previously was configured appears to no
1099 longer be configured. (syslog priority: Critical)
1100 If mdadm was told to monitor an array which is RAID0 or Lin‐
1102 ear, then it will report DeviceDisappeared with the extra
1103 information Wrong-Level. This is because RAID0 and Linear
1104 do not support the device-failed, hot-spare and resync oper‐
1105 ations which are monitored.
1106 RebuildStarted
1109 An md array started reconstruction. (syslog priority: Warn‐
1110 ing)
1111 RebuildNN
1114 Where NN is 20, 40, 60, or 80, this indicates that rebuild
1115 has passed that many percentage of the total. (syslog prior‐
1116 ity: Warning)
1117 RebuildFinished
1120 An md array that was rebuilding, isn't any more, either
1121 because it finished normally or was aborted. (syslog prior‐
1122 ity: Warning)
1123 Fail An active component device of an array has been marked as
1126 faulty. (syslog priority: Critical)
1127 FailSpare
1130 A spare component device which was being rebuilt to replace
1131 a faulty device has failed. (syslog priority: Critial)
1132 SpareActive
1135 A spare component device which was being rebuilt to replace
1136 a faulty device has been successfully rebuilt and has been
1137 made active. (syslog priority: Info)
1138 NewArray
1141 A new md array has been detected in the /proc/mdstat file.
1142 (syslog priority: Info)
1143 DegradedArray
1146 A newly noticed array appears to be degraded. This message
1147 is not generated when mdadm notices a drive failure which
1148 causes degradation, but only when mdadm notices that an
1149 array is degraded when it first sees the array. (syslog
1150 priority: Critial)
1151 MoveSpare
1154 A spare drive has been moved from one array in a spare-group
1155 to another to allow a failed drive to be replaced. (syslog
1156 priority: Info)
1157 SparesMissing
1160 If mdadm has been told, via the config file, that an array
1161 should have a certain number of spare devices, and mdadm
1162 detects that it has fewer that this number when it first
1163 sees the array, it will report a SparesMissing message.
1164 (syslog priority: Warning)
1165 TestMessage
1168 An array was found at startup, and the --test flag was
1169 given. (syslog priority: Info)
1170 Only Fail , FailSpare , DegradedArray , SparesMissing , and TestMessage
1172 cause Email to be sent. All events cause the program to be run. The
1173 program is run with two or three arguments, they being the event name,
1174 the array device and possibly a second device.
1175 Each event has an associated array device (e.g. /dev/md1) and possibly
1177 a second device. For Fail, FailSpare, and SpareActive the second
1178 device is the relevant component device. For MoveSpare the second
1179 device is the array that the spare was moved from.
1180 For mdadm to move spares from one array to another, the different
1182 arrays need to be labelled with the same spare-group in the configura‐
1183 tion file. The spare-group name can be any string. It is only neces‐
1184 sary that different spare groups use different names.
1185 When mdadm detects that an array which is in a spare group has fewer
1187 active devices than necessary for the complete array, and has no spare
1188 devices, it will look for another array in the same spare group that
1189 has a full complement of working drive and a spare. It will then
1190 attempt to remove the spare from the second drive and add it to the
1191 first. If the removal succeeds but the adding fails, then it is added
1192 back to the original array.
1193
1196 The GROW mode is used for changing the size or shape of an active
1197 array. For this to work, the kernel must support the necessary change.
1198 Various types of growth are being added during 2.6 development, includ‐
1199 ing restructuring a raid5 array to have more active devices.
1200 Currently the only support available is to
1202 · change the "size" attribute for RAID1, RAID5 and RAID6.
1204 · increase the "raid-disks" attribute of RAID1, RAID5, and RAID6.
1206 · add a write-intent bitmap to any array which support these bitmaps,
1208 or remove a write-intent bitmap from such an array.
1209 SIZE CHANGES
1211 Normally when an array is built the "size" it taken from the smallest
1212 of the drives. If all the small drives in an arrays are, one at a
1213 time, removed and replaced with larger drives, then you could have an
1214 array of large drives with only a small amount used. In this situa‐
1215 tion, changing the "size" with "GROW" mode will allow the extra space
1216 to start being used. If the size is increased in this way, a "resync"
1217 process will start to make sure the new parts of the array are synchro‐
1218 nised.
1219 Note that when an array changes size, any filesystem that may be stored
1221 in the array will not automatically grow to use the space. The
1222 filesystem will need to be explicitly told to use the extra space.
1223 RAID-DEVICES CHANGES
1226 A RAID1 array can work with any number of devices from 1 upwards
1227 (though 1 is not very useful). There may be times which you want to
1228 increase or decrease the number of active devices. Note that this is
1229 different to hot-add or hot-remove which changes the number of inactive
1230 devices.
1231 When reducing the number of devices in a RAID1 array, the slots which
1233 are to be removed from the array must already be vacant. That is, the
1234 devices that which were in those slots must be failed and removed.
1235 When the number of devices is increased, any hot spares that are
1237 present will be activated immediately.
1238 Increasing the number of active devices in a RAID5 is much more effort.
1240 Every block in the array will need to be read and written back to a new
1241 location. From 2.6.17, the Linux Kernel is able to do this safely,
1242 including restart and interrupted "reshape".
1243 When relocating the first few stripes on a raid5, it is not possible to
1245 keep the data on disk completely consistent and crash-proof. To pro‐
1246 vide the required safety, mdadm disables writes to the array while this
1247 "critical section" is reshaped, and takes a backup of the data that is
1248 in that section. This backup is normally stored in any spare devices
1249 that the array has, however it can also be stored in a separate file
1250 specified with the --backup-file option. If this option is used, and
1251 the system does crash during the critical period, the same file must be
1252 passed to --assemble to restore the backup and reassemble the array.
1253 BITMAP CHANGES
1256 A write-intent bitmap can be added to, or removed from, an active
1257 array. Either internal bitmaps, or bitmaps stored in a separate file
1258 can be added. Note that if you add a bitmap stored in a file which is
1259 in a filesystem that is on the raid array being affected, the system
1260 will deadlock. The bitmap must be on a separate filesystem.
1261
1264 Usage: mdadm --incremental [--run] [--quiet] component-device
1265 Usage: mdadm --incremental --rebuild
1267 Usage: mdadm --incremental --run --scan
1269 This mode is designed to be used in conjunction with a device discovery
1273 system. As devices are found in a system, they can be passed to mdadm
1274 --incremental to be conditionally added to an appropriate array.
1275 mdadm performs a number of tests to determine if the device is part of
1277 an array, and which array is should be part of. If an appropriate
1278 array is found, or can be created, mdadm adds the device to the array
1279 and conditionally starts the array.
1280 Note that mdadm will only add devices to an array which were previously
1282 working (active or spare) parts of that array. It does not currently
1283 support automatic inclusion of a new drive as a spare in some array.
1284 mdadm --incremental requires a bug present in all kernels through
1286 2.6.19, to be fixed. Hopefully this will be fixed in 2.6.20. Alter‐
1287 nately apply the patch which is included with the mdadm source distri‐
1288 bution. If mdadm detects that this bug is present, it will abort any
1289 attempt to use --incremental.
1290 The tests that mdadm makes are as follow:
1292 + Is the device permitted by mdadm.conf? That is, is it listed in
1294 a DEVICES line in that file. If DEVICES is absent then the
1295 default it to allow any device. Similar if DEVICES contains the
1296 special word partitions then any device is allowed. Otherwise
1297 the device name given to mdadm must match one of the names or
1298 patterns in a DEVICES line.
1299 + Does the device have a valid md superblock. If a specific meta‐
1302 data version is request with --metadata or -e then only that
1303 style of metadata is accepted, otherwise mdadm finds any known
1304 version of metadata. If no md metadata is found, the device is
1305 rejected.
1306 + Does the metadata match an expected array? The metadata can
1309 match in two ways. Either there is an array listed in
1310 mdadm.conf which identifies the array (either by UUID, by name,
1311 by device list, or by minor-number), the array was created with
1312 a homehost specified, and that homehost matches that which is
1313 given in mdadm.conf or on the command line. If mdadm is not
1314 able to positively identify the array as belonging to the cur‐
1315 rent host, the device will be rejected.
1316 + mdadm keeps a list of arrays that is has partly assembled in
1319 /var/run/mdadm/map (or /var/run/mdadm.map if the directory
1320 doesn't exist). If no array exists which matches the metadata
1321 on the new device, mdadm must choose a device name and unit num‐
1322 ber. It does this based on any name given in mdadm.conf or any
1323 name information stored in the metadata. If this name suggests
1324 a unit number, that number will be used, otherwise a free unit
1325 number will be chosen. Normally mdadm will prefer to create a
1326 partitionable array, however if the CREATE line in mdadm.conf
1327 suggests that a non-partitionable array is preferred, that will
1328 be honoured.
1329 + Once an appropriate array is found or created and the device is
1332 added, mdadm must decide if the array is ready to be started.
1333 It will normally compare the number of available (non-spare)
1334 devices to the number of devices that the metadata suggests need
1335 to be active. If there are at least that many, the array will
1336 be started. This means that if any devices are missing the
1337 array will not be restarted.
1338 As an alternative, --run may be passed to mdadm in which case
1340 the array will be run as soon as there are enough devices
1341 present for the data to be accessible. For a raid1, that means
1342 one device will start the array. For a clean raid5, the array
1343 will be started as soon as all but one drive is present.
1344 Note that neither of these approaches is really ideal. If it is
1346 can be known that all device discovery has completed, then
1347 mdadm -IRs
1348 can be run which will try to start all arrays that are being
1349 incrementally assembled. They are started in "read-auto" mode
1350 in which they are read-only until the first write request. This
1351 means that no metadata updates are made and no attempt at resync
1352 or recovery happens. Further devices that are found before the
1353 first write can still be added safely.
1354
1357 mdadm --query /dev/name-of-device
1358 This will find out if a given device is a raid array, or is part of
1359 one, and will provide brief information about the device.
1360 mdadm --assemble --scan
1362 This will assemble and start all arrays listed in the standard config
1363 file file. This command will typically go in a system startup file.
1364 mdadm --stop --scan
1366 This will shut down all array that can be shut down (i.e. are not cur‐
1367 rently in use). This will typically go in a system shutdown script.
1368 mdadm --follow --scan --delay=120
1370 If (and only if) there is an Email address or program given in the
1371 standard config file, then monitor the status of all arrays listed in
1372 that file by polling them ever 2 minutes.
1373 mdadm --create /dev/md0 --level=1 --raid-devices=2 /dev/hd[ac]1
1375 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
1376 echo 'DEVICE /dev/hd*[0-9] /dev/sd*[0-9]' > mdadm.conf
1378 mdadm --detail --scan >> mdadm.conf
1379 This will create a prototype config file that describes currently
1380 active arrays that are known to be made from partitions of IDE or SCSI
1381 drives. This file should be reviewed before being used as it may con‐
1382 tain unwanted detail.
1383 echo 'DEVICE /dev/hd[a-z] /dev/sd*[a-z]' > mdadm.conf
1385 mdadm --examine --scan --config=mdadm.conf >> mdadm.conf This will
1386 find what arrays could be assembled from existing IDE and SCSI whole
1387 drives (not partitions) and store the information is the format of a
1388 config file. This file is very likely to contain unwanted detail, par‐
1389 ticularly the devices= entries. It should be reviewed and edited
1390 before being used as an actual config file.
1391 mdadm --examine --brief --scan --config=partitions
1393 mdadm -Ebsc partitions
1394 Create a list of devices by reading /proc/partitions, scan these for
1395 RAID superblocks, and printout a brief listing of all that was found.
1396 mdadm -Ac partitions -m 0 /dev/md0
1398 Scan all partitions and devices listed in /proc/partitions and assemble
1399 /dev/md0 out of all such devices with a RAID superblock with a minor
1400 number of 0.
1401 mdadm --monitor --scan --daemonise > /var/run/mdadm
1403 If config file contains a mail address or alert program, run mdadm in
1404 the background in monitor mode monitoring all md devices. Also write
1405 pid of mdadm daemon to /var/run/mdadm.
1406 mdadm -Iq /dev/somedevice
1408 Try to incorporate newly discovered device into some array as appropri‐
1409 ate.
1410 mdadm --incremental --rebuild --run --scan
1412 Rebuild the array map from any current arrays, and then start any that
1413 can be started.
1414 mdadm /dev/md4 --fail detached --remove detached
1416 Any devices which are components of /dev/md4 will be marked as faulty
1417 and then remove from the array.
1418 mdadm --create --help
1420 Provide help about the Create mode.
1421 mdadm --config --help
1423 Provide help about the format of the config file.
1424 mdadm --help
1426 Provide general help.
1427
1431 /proc/mdstat
1432 If you're using the /proc filesystem, /proc/mdstat lists all active md
1433 devices with information about them. mdadm uses this to find arrays
1434 when --scan is given in Misc mode, and to monitor array reconstruction
1435 on Monitor mode.
1436 /etc/mdadm.conf
1440 The config file lists which devices may be scanned to see if they con‐
1441 tain MD super block, and gives identifying information (e.g. UUID)
1442 about known MD arrays. See mdadm.conf(5) for more details.
1443 /var/run/mdadm/map
1446 When --incremental mode is used. this file gets a list of arrays cur‐
1447 rently being created. If /var/run/mdadm does not exist as a directory,
1448 then /var/run/mdadm.map is used instead.
1449
1452 While entries in the /dev directory can have any format you like, mdadm
1453 has an understanding of 'standard' formats which it uses to guide its
1454 behaviour when creating device files via the --auto option.
1455 The standard names for non-partitioned arrays (the only sort of md
1457 array available in 2.4 and earlier) either of
1458 /dev/mdNN
1460 /dev/md/NN
1461 where NN is a number. The standard names for partitionable arrays (as
1463 available from 2.6 onwards) is one of
1464 /dev/md/dNN
1466 /dev/md_dNN
1467 Partition numbers should be indicated by added "pMM" to these, thus
1469 "/dev/md/d1p2".
1470
1473 mdadm was previously known as mdctl.
1474 mdadm is completely separate from the raidtools package, and does not
1476 use the /etc/raidtab configuration file at all.
1477
1480 For information on the various levels of RAID, check out:
1481 http://ostenfeld.dk/~jakob/Software-RAID.HOWTO/
1484 ⟨http://ostenfeld.dk/~jakob/Software-RAID.HOWTO/⟩
1485 The latest version of mdadm should always be available from
1487 http://www.kernel.org/pub/linux/utils/raid/mdadm/
1489 ⟨http://www.kernel.org/pub/linux/utils/raid/mdadm/⟩
1490 mdadm.conf(5), md(4).
1492 raidtab(5), raid0run(8), raidstop(8), mkraid(8).
1494v2.6.2 MDADM(8)