2 .\" Copyright Neil Brown and others.
3 .\" This program is free software; you can redistribute it and/or modify
4 .\" it under the terms of the GNU General Public License as published by
5 .\" the Free Software Foundation; either version 2 of the License, or
6 .\" (at your option) any later version.
7 .\" See file COPYING in distribution for details.
10 mdadm \- manage MD devices
16 .BI mdadm " [mode] <raiddevice> [options] <component-devices>"
19 RAID devices are virtual devices created from two or more
20 real block devices. This allows multiple devices (typically disk
21 drives or partitions thereof) to be combined into a single device to
22 hold (for example) a single filesystem.
23 Some RAID levels include redundancy and so can survive some degree of
26 Linux Software RAID devices are implemented through the md (Multiple
27 Devices) device driver.
29 Currently, Linux supports
46 is not a Software RAID mechanism, but does involve
48 each device is a path to one common physical storage device.
49 New installations should not use md/multipath as it is not well
50 supported and has no ongoing development. Use the Device Mapper based
51 multipath-tools instead.
54 is also not true RAID, and it only involves one device. It
55 provides a layer over a true device that can be used to inject faults.
60 is a collection of devices that are
61 managed as a set. This is similar to the set of devices connected to
62 a hardware RAID controller. The set of devices may contain a number
63 of different RAID arrays each utilising some (or all) of the blocks from a
64 number of the devices in the set. For example, two devices in a 5-device set
65 might form a RAID1 using the whole devices. The remaining three might
66 have a RAID5 over the first half of each device, and a RAID0 over the
71 there is one set of metadata that describes all of
72 the arrays in the container. So when
76 device, the device just represents the metadata. Other normal arrays (RAID1
77 etc) can be created inside the container.
80 mdadm has several major modes of operation:
83 Assemble the components of a previously created
84 array into an active array. Components can be explicitly given
85 or can be searched for.
87 checks that the components
88 do form a bona fide array, and can, on request, fiddle superblock
89 information so as to assemble a faulty array.
93 Build an array that doesn't have per-device metadata (superblocks). For these
96 cannot differentiate between initial creation and subsequent assembly
97 of an array. It also cannot perform any checks that appropriate
98 components have been requested. Because of this, the
100 mode should only be used together with a complete understanding of
105 Create a new array with per-device metadata (superblocks).
106 Appropriate metadata is written to each device, and then the array
107 comprising those devices is activated. A 'resync' process is started
108 to make sure that the array is consistent (e.g. both sides of a mirror
109 contain the same data) but the content of the device is left otherwise
111 The array can be used as soon as it has been created. There is no
112 need to wait for the initial resync to finish.
115 .B "Follow or Monitor"
116 Monitor one or more md devices and act on any state changes. This is
117 only meaningful for RAID1, 4, 5, 6, 10 or multipath arrays, as
118 only these have interesting state. RAID0 or Linear never have
119 missing, spare, or failed drives, so there is nothing to monitor.
123 Grow (or shrink) an array, or otherwise reshape it in some way.
124 Currently supported growth options including changing the active size
125 of component devices and changing the number of active devices in
126 Linear and RAID levels 0/1/4/5/6,
127 changing the RAID level between 0, 1, 5, and 6, and between 0 and 10,
128 changing the chunk size and layout for RAID 0,4,5,6, as well as adding or
129 removing a write-intent bitmap.
132 .B "Incremental Assembly"
133 Add a single device to an appropriate array. If the addition of the
134 device makes the array runnable, the array will be started.
135 This provides a convenient interface to a
137 system. As each device is detected,
139 has a chance to include it in some array as appropriate.
142 flag is passed in we will remove the device from any active array
143 instead of adding it.
149 in this mode, then any arrays within that container will be assembled
154 This is for doing things to specific components of an array such as
155 adding new spares and removing faulty devices.
159 This is an 'everything else' mode that supports operations on active
160 arrays, operations on component devices such as erasing old superblocks, and
161 information gathering operations.
162 .\"This mode allows operations on independent devices such as examine MD
163 .\"superblocks, erasing old superblocks and stopping active arrays.
167 This mode does not act on a specific device or array, but rather it
168 requests the Linux Kernel to activate any auto-detected arrays.
171 .SH Options for selecting a mode are:
174 .BR \-A ", " \-\-assemble
175 Assemble a pre-existing array.
178 .BR \-B ", " \-\-build
179 Build a legacy array without superblocks.
182 .BR \-C ", " \-\-create
186 .BR \-F ", " \-\-follow ", " \-\-monitor
192 .BR \-G ", " \-\-grow
193 Change the size or shape of an active array.
196 .BR \-I ", " \-\-incremental
197 Add/remove a single device to/from an appropriate array, and possibly start the array.
201 Request that the kernel starts any auto-detected arrays. This can only
204 is compiled into the kernel \(em not if it is a module.
205 Arrays can be auto-detected by the kernel if all the components are in
206 primary MS-DOS partitions with partition type
208 and all use v0.90 metadata.
209 In-kernel autodetect is not recommended for new installations. Using
211 to detect and assemble arrays \(em possibly in an
213 \(em is substantially more flexible and should be preferred.
216 If a device is given before any options, or if the first option is
222 then the MANAGE mode is assumed.
223 Anything other than these will cause the
227 .SH Options that are not mode-specific are:
230 .BR \-h ", " \-\-help
231 Display general help message or, after one of the above options, a
232 mode-specific help message.
236 Display more detailed help about command line parsing and some commonly
240 .BR \-V ", " \-\-version
241 Print version information for mdadm.
244 .BR \-v ", " \-\-verbose
245 Be more verbose about what is happening. This can be used twice to be
247 The extra verbosity currently only affects
248 .B \-\-detail \-\-scan
250 .BR "\-\-examine \-\-scan" .
253 .BR \-q ", " \-\-quiet
254 Avoid printing purely informative messages. With this,
256 will be silent unless there is something really important to report.
260 .BR \-f ", " \-\-force
261 Be more forceful about certain operations. See the various modes for
262 the exact meaning of this option in different contexts.
265 .BR \-c ", " \-\-config=
266 Specify the config file. Default is to use
267 .BR /etc/mdadm.conf ,
268 or if that is missing then
269 .BR /etc/mdadm/mdadm.conf .
270 If the config file given is
272 then nothing will be read, but
274 will act as though the config file contained exactly
275 .B "DEVICE partitions containers"
278 to find a list of devices to scan, and
280 to find a list of containers to examine.
283 is given for the config file, then
285 will act as though the config file were empty.
288 .BR \-s ", " \-\-scan
291 for missing information.
292 In general, this option gives
294 permission to get any missing information (like component devices,
295 array devices, array identities, and alert destination) from the
296 configuration file (see previous option);
297 one exception is MISC mode when using
303 says to get a list of array devices from
307 .BR \-e ", " \-\-metadata=
308 Declare the style of RAID metadata (superblock) to be used. The
309 default is {DEFAULT_METADATA} for
311 and to guess for other operations.
312 The default can be overridden by setting the
321 .ie '{DEFAULT_METADATA}'0.90'
322 .IP "0, 0.90, default"
325 Use the original 0.90 format superblock. This format limits arrays to
326 28 component devices and limits component devices of levels 1 and
327 greater to 2 terabytes. It is also possible for there to be confusion
328 about whether the superblock applies to a whole device or just the
329 last partition, if that partition starts on a 64K boundary.
330 .ie '{DEFAULT_METADATA}'0.90'
331 .IP "1, 1.0, 1.1, 1.2"
333 .IP "1, 1.0, 1.1, 1.2 default"
334 Use the new version-1 format superblock. This has fewer restrictions.
335 It can easily be moved between hosts with different endian-ness, and a
336 recovery operation can be checkpointed and restarted. The different
337 sub-versions store the superblock at different locations on the
338 device, either at the end (for 1.0), at the start (for 1.1) or 4K from
339 the start (for 1.2). "1" is equivalent to "1.2" (the commonly
340 preferred 1.x format).
341 'if '{DEFAULT_METADATA}'1.2' "default" is equivalent to "1.2".
343 Use the "Industry Standard" DDF (Disk Data Format) format defined by
345 When creating a DDF array a
347 will be created, and normal arrays can be created in that container.
349 Use the Intel(R) Matrix Storage Manager metadata format. This creates a
351 which is managed in a similar manner to DDF, and is supported by an
352 option-rom on some platforms:
354 .B http://www.intel.com/design/chipsets/matrixstorage_sb.htm
360 This will override any
362 setting in the config file and provides the identity of the host which
363 should be considered the home for any arrays.
365 When creating an array, the
367 will be recorded in the metadata. For version-1 superblocks, it will
368 be prefixed to the array name. For version-0.90 superblocks, part of
369 the SHA1 hash of the hostname will be stored in the later half of the
372 When reporting information about an array, any array which is tagged
373 for the given homehost will be reported as such.
375 When using Auto-Assemble, only arrays tagged for the given homehost
376 will be allowed to use 'local' names (i.e. not ending in '_' followed
377 by a digit string). See below under
378 .BR "Auto Assembly" .
384 needs to print the name for a device it normally finds the name in
386 which refers to the device and is shortest. When a path component is
390 will prefer a longer name if it contains that component. For example
391 .B \-\-prefer=by-uuid
392 will prefer a name in a subdirectory of
397 This functionality is currently only provided by
402 .SH For create, build, or grow:
405 .BR \-n ", " \-\-raid\-devices=
406 Specify the number of active devices in the array. This, plus the
407 number of spare devices (see below) must equal the number of
409 (including "\fBmissing\fP" devices)
410 that are listed on the command line for
412 Setting a value of 1 is probably
413 a mistake and so requires that
415 be specified first. A value of 1 will then be allowed for linear,
416 multipath, RAID0 and RAID1. It is never allowed for RAID4, RAID5 or RAID6.
418 This number can only be changed using
420 for RAID1, RAID4, RAID5 and RAID6 arrays, and only on kernels which provide
421 the necessary support.
424 .BR \-x ", " \-\-spare\-devices=
425 Specify the number of spare (eXtra) devices in the initial array.
426 Spares can also be added
427 and removed later. The number of component devices listed
428 on the command line must equal the number of RAID devices plus the
429 number of spare devices.
432 .BR \-z ", " \-\-size=
433 Amount (in Kibibytes) of space to use from each drive in RAID levels 1/4/5/6.
434 This must be a multiple of the chunk size, and must leave about 128Kb
435 of space at the end of the drive for the RAID superblock.
436 If this is not specified
437 (as it normally is not) the smallest drive (or partition) sets the
438 size, though if there is a variance among the drives of greater than 1%, a warning is
441 A suffix of 'M' or 'G' can be given to indicate Megabytes or
442 Gigabytes respectively.
444 Sometimes a replacement drive can be a little smaller than the
445 original drives though this should be minimised by IDEMA standards.
446 Such a replacement drive will be rejected by
448 To guard against this it can be useful to set the initial size
449 slightly smaller than the smaller device with the aim that it will
450 still be larger than any replacement.
452 This value can be set with
454 for RAID level 1/4/5/6 though
456 based arrays such as those with IMSM metadata may not be able to
458 If the array was created with a size smaller than the currently
459 active drives, the extra space can be accessed using
461 The size can be given as
463 which means to choose the largest size that fits on all current drives.
465 Before reducing the size of the array (with
466 .BR "\-\-grow \-\-size=" )
467 you should make sure that space isn't needed. If the device holds a
468 filesystem, you would need to resize the filesystem to use less space.
470 After reducing the array size you should check that the data stored in
471 the device is still available. If the device holds a filesystem, then
472 an 'fsck' of the filesystem is a minimum requirement. If there are
473 problems the array can be made bigger again with no loss with another
474 .B "\-\-grow \-\-size="
477 This value cannot be used when creating a
479 such as with DDF and IMSM metadata, though it perfectly valid when
480 creating an array inside a container.
483 .BR \-Z ", " \-\-array\-size=
484 This is only meaningful with
486 and its effect is not persistent: when the array is stopped and
487 restarted the default array size will be restored.
489 Setting the array-size causes the array to appear smaller to programs
490 that access the data. This is particularly needed before reshaping an
491 array so that it will be smaller. As the reshape is not reversible,
492 but setting the size with
494 is, it is required that the array size is reduced as appropriate
495 before the number of devices in the array is reduced.
497 Before reducing the size of the array you should make sure that space
498 isn't needed. If the device holds a filesystem, you would need to
499 resize the filesystem to use less space.
501 After reducing the array size you should check that the data stored in
502 the device is still available. If the device holds a filesystem, then
503 an 'fsck' of the filesystem is a minimum requirement. If there are
504 problems the array can be made bigger again with no loss with another
505 .B "\-\-grow \-\-array\-size="
508 A suffix of 'M' or 'G' can be given to indicate Megabytes or
509 Gigabytes respectively.
512 restores the apparent size of the array to be whatever the real
513 amount of available space is.
516 .BR \-c ", " \-\-chunk=
517 Specify chunk size of kibibytes. The default when creating an
518 array is 512KB. To ensure compatibility with earlier versions, the
519 default when Building and array with no persistent metadata is 64KB.
520 This is only meaningful for RAID0, RAID4, RAID5, RAID6, and RAID10.
522 RAID4, RAID5, RAID6, and RAID10 require the chunk size to be a power
523 of 2. In any case it must be a multiple of 4KB.
525 A suffix of 'M' or 'G' can be given to indicate Megabytes or
526 Gigabytes respectively.
530 Specify rounding factor for a Linear array. The size of each
531 component will be rounded down to a multiple of this size.
532 This is a synonym for
534 but highlights the different meaning for Linear as compared to other
535 RAID levels. The default is 64K if a kernel earlier than 2.6.16 is in
536 use, and is 0K (i.e. no rounding) in later kernels.
539 .BR \-l ", " \-\-level=
540 Set RAID level. When used with
542 options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
543 raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty, container.
544 Obviously some of these are synonymous.
548 metadata type is requested, only the
550 level is permitted, and it does not need to be explicitly given.
554 only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid.
558 to change the RAID level in some cases. See LEVEL CHANGES below.
561 .BR \-p ", " \-\-layout=
562 This option configures the fine details of data layout for RAID5, RAID6,
563 and RAID10 arrays, and controls the failure modes for
566 The layout of the RAID5 parity block can be one of
567 .BR left\-asymmetric ,
568 .BR left\-symmetric ,
569 .BR right\-asymmetric ,
570 .BR right\-symmetric ,
571 .BR la ", " ra ", " ls ", " rs .
573 .BR left\-symmetric .
575 It is also possible to cause RAID5 to use a RAID4-like layout by
581 Finally for RAID5 there are DDF\-compatible layouts,
582 .BR ddf\-zero\-restart ,
583 .BR ddf\-N\-restart ,
585 .BR ddf\-N\-continue .
587 These same layouts are available for RAID6. There are also 4 layouts
588 that will provide an intermediate stage for converting between RAID5
589 and RAID6. These provide a layout which is identical to the
590 corresponding RAID5 layout on the first N\-1 devices, and has the 'Q'
591 syndrome (the second 'parity' block used by RAID6) on the last device.
593 .BR left\-symmetric\-6 ,
594 .BR right\-symmetric\-6 ,
595 .BR left\-asymmetric\-6 ,
596 .BR right\-asymmetric\-6 ,
598 .BR parity\-first\-6 .
600 When setting the failure mode for level
603 .BR write\-transient ", " wt ,
604 .BR read\-transient ", " rt ,
605 .BR write\-persistent ", " wp ,
606 .BR read\-persistent ", " rp ,
608 .BR read\-fixable ", " rf ,
609 .BR clear ", " flush ", " none .
611 Each failure mode can be followed by a number, which is used as a period
612 between fault generation. Without a number, the fault is generated
613 once on the first relevant request. With a number, the fault will be
614 generated after that many requests, and will continue to be generated
615 every time the period elapses.
617 Multiple failure modes can be current simultaneously by using the
619 option to set subsequent failure modes.
621 "clear" or "none" will remove any pending or periodic failure modes,
622 and "flush" will clear any persistent faults.
624 Finally, the layout options for RAID10 are one of 'n', 'o' or 'f' followed
625 by a small number. The default is 'n2'. The supported options are:
628 signals 'near' copies. Multiple copies of one data block are at
629 similar offsets in different devices.
632 signals 'offset' copies. Rather than the chunks being duplicated
633 within a stripe, whole stripes are duplicated but are rotated by one
634 device so duplicate blocks are on different devices. Thus subsequent
635 copies of a block are in the next drive, and are one chunk further
640 (multiple copies have very different offsets).
641 See md(4) for more detail about 'near', 'offset', and 'far'.
643 The number is the number of copies of each datablock. 2 is normal, 3
644 can be useful. This number can be at most equal to the number of
645 devices in the array. It does not need to divide evenly into that
646 number (e.g. it is perfectly legal to have an 'n2' layout for an array
647 with an odd number of devices).
649 When an array is converted between RAID5 and RAID6 an intermediate
650 RAID6 layout is used in which the second parity block (Q) is always on
651 the last device. To convert a RAID5 to RAID6 and leave it in this new
652 layout (which does not require re-striping) use
653 .BR \-\-layout=preserve .
654 This will try to avoid any restriping.
656 The converse of this is
657 .B \-\-layout=normalise
658 which will change a non-standard RAID6 layout into a more standard
665 (thus explaining the p of
669 .BR \-b ", " \-\-bitmap=
670 Specify a file to store a write-intent bitmap in. The file should not
673 is also given. The same file should be provided
674 when assembling the array. If the word
676 is given, then the bitmap is stored with the metadata on the array,
677 and so is replicated on all devices. If the word
681 mode, then any bitmap that is present is removed.
683 To help catch typing errors, the filename must contain at least one
684 slash ('/') if it is a real file (not 'internal' or 'none').
686 Note: external bitmaps are only known to work on ext2 and ext3.
687 Storing bitmap files on other filesystems may result in serious problems.
689 When creating an array on devices which are 100G or larger,
691 automatically adds an internal bitmap as it will usually be
692 beneficial. This can be suppressed with
693 .B "\-\-bitmap=none".
696 .BR \-\-bitmap\-chunk=
697 Set the chunksize of the bitmap. Each bit corresponds to that many
698 Kilobytes of storage.
699 When using a file based bitmap, the default is to use the smallest
700 size that is at-least 4 and requires no more than 2^21 chunks.
703 bitmap, the chunksize defaults to 64Meg, or larger if necessary to
704 fit the bitmap into the available space.
706 A suffix of 'M' or 'G' can be given to indicate Megabytes or
707 Gigabytes respectively.
710 .BR \-W ", " \-\-write\-mostly
711 subsequent devices listed in a
716 command will be flagged as 'write-mostly'. This is valid for RAID1
717 only and means that the 'md' driver will avoid reading from these
718 devices if at all possible. This can be useful if mirroring over a
722 .BR \-\-write\-behind=
723 Specify that write-behind mode should be enabled (valid for RAID1
724 only). If an argument is specified, it will set the maximum number
725 of outstanding writes allowed. The default value is 256.
726 A write-intent bitmap is required in order to use write-behind
727 mode, and write-behind is only attempted on drives marked as
731 .BR \-\-assume\-clean
734 that the array pre-existed and is known to be clean. It can be useful
735 when trying to recover from a major failure as you can be sure that no
736 data will be affected unless you actually write to the array. It can
737 also be used when creating a RAID1 or RAID10 if you want to avoid the
738 initial resync, however this practice \(em while normally safe \(em is not
739 recommended. Use this only if you really know what you are doing.
741 When the devices that will be part of a new array were filled
742 with zeros before creation the operator knows the array is
743 actually clean. If that is the case, such as after running
744 badblocks, this argument can be used to tell mdadm the
745 facts the operator knows.
747 When an array is resized to a larger size with
748 .B "\-\-grow \-\-size="
749 the new space is normally resynced in that same way that the whole
750 array is resynced at creation. From Linux version 3.0,
752 can be used with that command to avoid the automatic resync.
755 .BR \-\-backup\-file=
758 is used to increase the number of raid-devices in a RAID5 or RAID6 if
759 there are no spare devices available, or to shrink, change RAID level
760 or layout. See the GROW MODE section below on RAID\-DEVICES CHANGES.
761 The file must be stored on a separate device, not on the RAID array
766 Arrays with 1.x metadata can leave a gap between the start of the
767 device and the start of array data. This gap can be used for various
768 metadata. The start of data is known as the
770 Normally an appropriate data offset is computed automatically.
771 However it can be useful to set it explicitly such as when re-creating
772 an array which was originally created using a different version of
774 which computed a different offset.
776 Setting the offset explicitly over-rides the default. The value given
777 is in Kilobytes unless an 'M' or 'G' suffix is given.
781 can also be used with
783 for some RAID levels (initially on RAID10). This allows the
784 data\-offset to be changed as part of the reshape process. When the
785 data offset is changed, no backup file is required as the difference
786 in offsets is used to provide the same functionality.
788 When the new offset is earlier than the old offset, the number of
789 devices in the array cannot shrink. When it is after the old offset,
790 the number of devices in the array cannot increase.
792 When creating an array,
796 In the case each member device is expected to have a offset appended
797 to the name, separated by a colon. This makes it possible to recreate
798 exactly an array which has varying data offsets (as can happen when
799 different versions of
801 are used to add different devices).
805 This option is complementary to the
806 .B \-\-freeze-reshape
807 option for assembly. It is needed when
809 operation is interrupted and it is not restarted automatically due to
810 .B \-\-freeze-reshape
811 usage during array assembly. This option is used together with
815 ) command and device for a pending reshape to be continued.
816 All parameters required for reshape continuation will be read from array metadata.
820 .BR \-\-backup\-file=
821 option to be set, continuation option will require to have exactly the same
822 backup file given as well.
824 Any other parameter passed together with
826 option will be ignored.
829 .BR \-N ", " \-\-name=
832 for the array. This is currently only effective when creating an
833 array with a version-1 superblock, or an array in a DDF container.
834 The name is a simple textual string that can be used to identify array
835 components when assembling. If name is needed but not specified, it
836 is taken from the basename of the device that is being created.
848 run the array, even if some of the components
849 appear to be active in another array or filesystem. Normally
851 will ask for confirmation before including such components in an
852 array. This option causes that question to be suppressed.
855 .BR \-f ", " \-\-force
858 accept the geometry and layout specified without question. Normally
860 will not allow creation of an array with only one device, and will try
861 to create a RAID5 array with one missing drive (as this makes the
862 initial resync work faster). With
865 will not try to be so clever.
868 .BR \-o ", " \-\-readonly
871 rather than read-write as normal. No writes will be allowed to the
872 array, and no resync, recovery, or reshape will be started.
875 .BR \-a ", " "\-\-auto{=yes,md,mdp,part,p}{NN}"
876 Instruct mdadm how to create the device file if needed, possibly allocating
877 an unused minor number. "md" causes a non-partitionable array
878 to be used (though since Linux 2.6.28, these array devices are in fact
879 partitionable). "mdp", "part" or "p" causes a partitionable array (2.6 and
880 later) to be used. "yes" requires the named md device to have
881 a 'standard' format, and the type and minor number will be determined
882 from this. With mdadm 3.0, device creation is normally left up to
884 so this option is unlikely to be needed.
885 See DEVICE NAMES below.
887 The argument can also come immediately after
892 is not given on the command line or in the config file, then
898 is also given, then any
900 entries in the config file will override the
902 instruction given on the command line.
904 For partitionable arrays,
906 will create the device file for the whole array and for the first 4
907 partitions. A different number of partitions can be specified at the
908 end of this option (e.g.
910 If the device name ends with a digit, the partition names add a 'p',
912 .IR /dev/md/home1p3 .
913 If there is no trailing digit, then the partition names just have a
915 .IR /dev/md/scratch3 .
917 If the md device name is in a 'standard' format as described in DEVICE
918 NAMES, then it will be created, if necessary, with the appropriate
919 device number based on that name. If the device name is not in one of these
920 formats, then a unused device number will be allocated. The device
921 number will be considered unused if there is no active array for that
922 number, and there is no entry in /dev for that number and with a
923 non-standard name. Names that are not in 'standard' format are only
924 allowed in "/dev/md/".
926 This is meaningful with
932 .BR \-a ", " "\-\-add"
933 This option can be used in Grow mode in two cases.
935 If the target array is a Linear array, then
937 can be used to add one or more devices to the array. They
938 are simply catenated on to the end of the array. Once added, the
939 devices cannot be removed.
943 option is being used to increase the number of devices in an array,
946 can be used to add some extra devices to be included in the array.
947 In most cases this is not needed as the extra devices can be added as
948 spares first, and then the number of raid-disks can be changed.
949 However for RAID0, it is not possible to add spares. So to increase
950 the number of devices in a RAID0, it is necessary to set the new
951 number of devices, and to add the new devices, in the same command.
956 .BR \-u ", " \-\-uuid=
957 uuid of array to assemble. Devices which don't have this uuid are
961 .BR \-m ", " \-\-super\-minor=
962 Minor number of device that array was created for. Devices which
963 don't have this minor number are excluded. If you create an array as
964 /dev/md1, then all superblocks will contain the minor number 1, even if
965 the array is later assembled as /dev/md2.
967 Giving the literal word "dev" for
971 to use the minor number of the md device that is being assembled.
974 .B \-\-super\-minor=dev
975 will look for super blocks with a minor number of 0.
978 is only relevant for v0.90 metadata, and should not normally be used.
984 .BR \-N ", " \-\-name=
985 Specify the name of the array to assemble. This must be the name
986 that was specified when creating the array. It must either match
987 the name stored in the superblock exactly, or it must match
990 prefixed to the start of the given name.
993 .BR \-f ", " \-\-force
994 Assemble the array even if the metadata on some devices appears to be
997 cannot find enough working devices to start the array, but can find
998 some devices that are recorded as having failed, then it will mark
999 those devices as working so that the array can be started.
1000 An array which requires
1002 to be started may contain data corruption. Use it carefully.
1005 .BR \-R ", " \-\-run
1006 Attempt to start the array even if fewer drives were given than were
1007 present last time the array was active. Normally if not all the
1008 expected drives are found and
1010 is not used, then the array will be assembled but not started.
1013 an attempt will be made to start it anyway.
1017 This is the reverse of
1019 in that it inhibits the startup of array unless all expected drives
1020 are present. This is only needed with
1022 and can be used if the physical connections to devices are
1023 not as reliable as you would like.
1026 .BR \-a ", " "\-\-auto{=no,yes,md,mdp,part}"
1027 See this option under Create and Build options.
1030 .BR \-b ", " \-\-bitmap=
1031 Specify the bitmap file that was given when the array was created. If
1034 bitmap, there is no need to specify this when assembling the array.
1037 .BR \-\-backup\-file=
1040 was used while reshaping an array (e.g. changing number of devices or
1041 chunk size) and the system crashed during the critical section, then the same
1043 must be presented to
1045 to allow possibly corrupted data to be restored, and the reshape
1049 .BR \-\-invalid\-backup
1050 If the file needed for the above option is not available for any
1051 reason an empty file can be given together with this option to
1052 indicate that the backup file is invalid. In this case the data that
1053 was being rearranged at the time of the crash could be irrecoverably
1054 lost, but the rest of the array may still be recoverable. This option
1055 should only be used as a last resort if there is no way to recover the
1060 .BR \-U ", " \-\-update=
1061 Update the superblock on each device while assembling the array. The
1062 argument given to this flag can be one of
1079 option will adjust the superblock of an array what was created on a Sparc
1080 machine running a patched 2.2 Linux kernel. This kernel got the
1081 alignment of part of the superblock wrong. You can use the
1082 .B "\-\-examine \-\-sparc2.2"
1085 to see what effect this would have.
1089 option will update the
1090 .B "preferred minor"
1091 field on each superblock to match the minor number of the array being
1093 This can be useful if
1095 reports a different "Preferred Minor" to
1097 In some cases this update will be performed automatically
1098 by the kernel driver. In particular the update happens automatically
1099 at the first write to an array with redundancy (RAID level 1 or
1100 greater) on a 2.6 (or later) kernel.
1104 option will change the uuid of the array. If a UUID is given with the
1106 option that UUID will be used as a new UUID and will
1108 be used to help identify the devices in the array.
1111 is given, a random UUID is chosen.
1115 option will change the
1117 of the array as stored in the superblock. This is only supported for
1118 version-1 superblocks.
1122 option will change the
1124 as recorded in the superblock. For version-0 superblocks, this is the
1125 same as updating the UUID.
1126 For version-1 superblocks, this involves updating the name.
1130 option will cause the array to be marked
1132 meaning that any redundancy in the array (e.g. parity for RAID5,
1133 copies for RAID1) may be incorrect. This will cause the RAID system
1134 to perform a "resync" pass to make sure that all redundant information
1139 option allows arrays to be moved between machines with different
1141 When assembling such an array for the first time after a move, giving
1142 .B "\-\-update=byteorder"
1145 to expect superblocks to have their byteorder reversed, and will
1146 correct that order before assembling the array. This is only valid
1147 with original (Version 0.90) superblocks.
1151 option will correct the summaries in the superblock. That is the
1152 counts of total, working, active, failed, and spare devices.
1156 option will rarely be of use. It applies to version 1.1 and 1.2 metadata
1157 only (where the metadata is at the start of the device) and is only
1158 useful when the component device has changed size (typically become
1159 larger). The version 1 metadata records the amount of the device that
1160 can be used to store data, so if a device in a version 1.1 or 1.2
1161 array becomes larger, the metadata will still be visible, but the
1162 extra space will not. In this case it might be useful to assemble the
1164 .BR \-\-update=devicesize .
1167 to determine the maximum usable amount of space on each device and
1168 update the relevant field in the metadata.
1172 option can be used when an array has an internal bitmap which is
1173 corrupt in some way so that assembling the array normally fails. It
1174 will cause any internal bitmap to be ignored.
1178 option will reserve space in each device for a bad block list. This
1179 will be 4K in size and positioned near the end of any free space
1180 between the superblock and the data.
1184 option will cause any reservation of space for a bad block list to be
1185 removed. If the bad block list contains entries, this will fail, as
1186 removing the list could cause data corruption.
1189 .BR \-\-freeze\-reshape
1190 Option is intended to be used in start-up scripts during initrd boot phase.
1191 When array under reshape is assembled during initrd phase, this option
1192 stops reshape after reshape critical section is being restored. This happens
1193 before file system pivot operation and avoids loss of file system context.
1194 Losing file system context would cause reshape to be broken.
1196 Reshape can be continued later using the
1198 option for the grow command.
1200 .SH For Manage mode:
1203 .BR \-t ", " \-\-test
1204 Unless a more serious error occurred,
1206 will exit with a status of 2 if no changes were made to the array and
1207 0 if at least one change was made.
1208 This can be useful when an indirect specifier such as
1213 is used in requesting an operation on the array.
1215 will report failure if these specifiers didn't find any match.
1218 .BR \-a ", " \-\-add
1219 hot-add listed devices.
1220 If a device appears to have recently been part of the array
1221 (possibly it failed or was removed) the device is re\-added as described
1223 If that fails or the device was never part of the array, the device is
1224 added as a hot-spare.
1225 If the array is degraded, it will immediately start to rebuild data
1228 Note that this and the following options are only meaningful on array
1229 with redundancy. They don't apply to RAID0 or Linear.
1233 re\-add a device that was previous removed from an array.
1234 If the metadata on the device reports that it is a member of the
1235 array, and the slot that it used is still vacant, then the device will
1236 be added back to the array in the same position. This will normally
1237 cause the data for that device to be recovered. However based on the
1238 event count on the device, the recovery may only require sections that
1239 are flagged a write-intent bitmap to be recovered or may not require
1240 any recovery at all.
1242 When used on an array that has no metadata (i.e. it was built with
1244 it will be assumed that bitmap-based recovery is enough to make the
1245 device fully consistent with the array.
1247 When used with v1.x metadata,
1249 can be accompanied by
1250 .BR \-\-update=devicesize ,
1251 .BR \-\-update=bbl ", or"
1252 .BR \-\-update=no\-bbl .
1253 See the description of these option when used in Assemble mode for an
1254 explanation of their use.
1256 If the device name given is
1260 will try to find any device that looks like it should be
1261 part of the array but isn't and will try to re\-add all such devices.
1263 If the device name given is
1267 will find all devices in the array that are marked
1269 remove them and attempt to immediately re\-add them. This can be
1270 useful if you are certain that the reason for failure has been
1274 .BR \-r ", " \-\-remove
1275 remove listed devices. They must not be active. i.e. they should
1276 be failed or spare devices.
1278 As well as the name of a device file
1288 The first causes all failed device to be removed. The second causes
1289 any device which is no longer connected to the system (i.e an 'open'
1293 The third will remove a set as describe below under
1297 .BR \-f ", " \-\-fail
1298 Mark listed devices as faulty.
1299 As well as the name of a device file, the word
1303 can be given. The former will cause any device that has been detached from
1304 the system to be marked as failed. It can then be removed.
1306 For RAID10 arrays where the number of copies evenly divides the number
1307 of devices, the devices can be conceptually divided into sets where
1308 each set contains a single complete copy of the data on the array.
1309 Sometimes a RAID10 array will be configured so that these sets are on
1310 separate controllers. In this case all the devices in one set can be
1311 failed by giving a name like
1317 The appropriate set names are reported by
1327 Mark listed devices as requiring replacement. As soon as a spare is
1328 available, it will be rebuilt and will replace the marked device.
1329 This is similar to marking a device as faulty, but the device remains
1330 in service during the recovery process to increase resilience against
1331 multiple failures. When the replacement process finishes, the
1332 replaced device will be marked as faulty.
1336 This can follow a list of
1338 devices. The devices listed after
1340 will be preferentially used to replace the devices listed after
1342 These device must already be spare devices in the array.
1345 .BR \-\-write\-mostly
1346 Subsequent devices that are added or re\-added will have the 'write-mostly'
1347 flag set. This is only valid for RAID1 and means that the 'md' driver
1348 will avoid reading from these devices if possible.
1351 Subsequent devices that are added or re\-added will have the 'write-mostly'
1355 Each of these options requires that the first device listed is the array
1356 to be acted upon, and the remainder are component devices to be added,
1357 removed, marked as faulty, etc. Several different operations can be
1358 specified for different devices, e.g.
1360 mdadm /dev/md0 \-\-add /dev/sda1 \-\-fail /dev/sdb1 \-\-remove /dev/sdb1
1362 Each operation applies to all devices listed until the next
1365 If an array is using a write-intent bitmap, then devices which have
1366 been removed can be re\-added in a way that avoids a full
1367 reconstruction but instead just updates the blocks that have changed
1368 since the device was removed. For arrays with persistent metadata
1369 (superblocks) this is done automatically. For arrays created with
1371 mdadm needs to be told that this device we removed recently with
1374 Devices can only be removed from an array if they are not in active
1375 use, i.e. that must be spares or failed devices. To remove an active
1376 device, it must first be marked as
1382 .BR \-Q ", " \-\-query
1383 Examine a device to see
1384 (1) if it is an md device and (2) if it is a component of an md
1386 Information about what is discovered is presented.
1389 .BR \-D ", " \-\-detail
1390 Print details of one or more md devices.
1393 .BR \-\-detail\-platform
1394 Print details of the platform's RAID capabilities (firmware / hardware
1395 topology) for a given metadata format. If used without argument, mdadm
1396 will scan all controllers looking for their capabilities. Otherwise, mdadm
1397 will only look at the controller specified by the argument in form of an
1398 absolute filepath or a link, e.g.
1399 .IR /sys/devices/pci0000:00/0000:00:1f.2 .
1402 .BR \-Y ", " \-\-export
1404 .B \-\-detail , \-\-detail-platform
1407 output will be formatted as
1409 pairs for easy import into the environment.
1412 .BR \-E ", " \-\-examine
1413 Print contents of the metadata stored on the named device(s).
1414 Note the contrast between
1419 applies to devices which are components of an array, while
1421 applies to a whole array which is currently active.
1424 If an array was created on a SPARC machine with a 2.2 Linux kernel
1425 patched with RAID support, the superblock will have been created
1426 incorrectly, or at least incompatibly with 2.4 and later kernels.
1431 will fix the superblock before displaying it. If this appears to do
1432 the right thing, then the array can be successfully assembled using
1433 .BR "\-\-assemble \-\-update=sparc2.2" .
1436 .BR \-X ", " \-\-examine\-bitmap
1437 Report information about a bitmap file.
1438 The argument is either an external bitmap file or an array component
1439 in case of an internal bitmap. Note that running this on an array
1442 does not report the bitmap for that array.
1445 .B \-\-examine\-badblocks
1446 List the bad-blocks recorded for the device, if a bad-blocks list has
1447 been configured. Currently only
1449 metadata supports bad-blocks lists.
1452 .BR \-R ", " \-\-run
1453 start a partially assembled array. If
1455 did not find enough devices to fully start the array, it might leaving
1456 it partially assembled. If you wish, you can then use
1458 to start the array in degraded mode.
1461 .BR \-S ", " \-\-stop
1462 deactivate array, releasing all resources.
1465 .BR \-o ", " \-\-readonly
1466 mark array as readonly.
1469 .BR \-w ", " \-\-readwrite
1470 mark array as readwrite.
1473 .B \-\-zero\-superblock
1474 If the device contains a valid md superblock, the block is
1475 overwritten with zeros. With
1477 the block where the superblock would be is overwritten even if it
1478 doesn't appear to be valid.
1481 .B \-\-kill\-subarray=
1482 If the device is a container and the argument to \-\-kill\-subarray
1483 specifies an inactive subarray in the container, then the subarray is
1484 deleted. Deleting all subarrays will leave an 'empty-container' or
1485 spare superblock on the drives. See \-\-zero\-superblock for completely
1486 removing a superblock. Note that some formats depend on the subarray
1487 index for generating a UUID, this command will fail if it would change
1488 the UUID of an active subarray.
1491 .B \-\-update\-subarray=
1492 If the device is a container and the argument to \-\-update\-subarray
1493 specifies a subarray in the container, then attempt to update the given
1494 superblock field in the subarray. See below in
1499 .BR \-t ", " \-\-test
1504 is set to reflect the status of the device. See below in
1509 .BR \-W ", " \-\-wait
1510 For each md device given, wait for any resync, recovery, or reshape
1511 activity to finish before returning.
1513 will return with success if it actually waited for every device
1514 listed, otherwise it will return failure.
1518 For each md device given, or each device in /proc/mdstat if
1520 is given, arrange for the array to be marked clean as soon as possible.
1522 will return with success if the array uses external metadata and we
1523 successfully waited. For native arrays this returns immediately as the
1524 kernel handles dirty-clean transitions at shutdown. No action is taken
1525 if safe-mode handling is disabled.
1527 .SH For Incremental Assembly mode:
1529 .BR \-\-rebuild\-map ", " \-r
1530 Rebuild the map file
1534 uses to help track which arrays are currently being assembled.
1537 .BR \-\-run ", " \-R
1538 Run any array assembled as soon as a minimal number of devices are
1539 available, rather than waiting until all expected devices are present.
1542 .BR \-\-scan ", " \-s
1543 Only meaningful with
1547 file for arrays that are being incrementally assembled and will try to
1548 start any that are not already started. If any such array is listed
1551 as requiring an external bitmap, that bitmap will be attached first.
1554 .BR \-\-fail ", " \-f
1555 This allows the hot-plug system to remove devices that have fully disappeared
1556 from the kernel. It will first fail and then remove the device from any
1557 array it belongs to.
1558 The device name given should be a kernel device name such as "sda",
1564 Only used with \-\-fail. The 'path' given will be recorded so that if
1565 a new device appears at the same location it can be automatically
1566 added to the same array. This allows the failed device to be
1567 automatically replaced by a new device without metadata if it appears
1568 at specified path. This option is normally only set by a
1572 .SH For Monitor mode:
1574 .BR \-m ", " \-\-mail
1575 Give a mail address to send alerts to.
1578 .BR \-p ", " \-\-program ", " \-\-alert
1579 Give a program to be run whenever an event is detected.
1582 .BR \-y ", " \-\-syslog
1583 Cause all events to be reported through 'syslog'. The messages have
1584 facility of 'daemon' and varying priorities.
1587 .BR \-d ", " \-\-delay
1588 Give a delay in seconds.
1590 polls the md arrays and then waits this many seconds before polling
1591 again. The default is 60 seconds. Since 2.6.16, there is no need to
1592 reduce this as the kernel alerts
1594 immediately when there is any change.
1597 .BR \-r ", " \-\-increment
1598 Give a percentage increment.
1600 will generate RebuildNN events with the given percentage increment.
1603 .BR \-f ", " \-\-daemonise
1606 to run as a background daemon if it decides to monitor anything. This
1607 causes it to fork and run in the child, and to disconnect from the
1608 terminal. The process id of the child is written to stdout.
1611 which will only continue monitoring if a mail address or alert program
1612 is found in the config file.
1615 .BR \-i ", " \-\-pid\-file
1618 is running in daemon mode, write the pid of the daemon process to
1619 the specified file, instead of printing it on standard output.
1622 .BR \-1 ", " \-\-oneshot
1623 Check arrays only once. This will generate
1625 events and more significantly
1631 .B " mdadm \-\-monitor \-\-scan \-1"
1633 from a cron script will ensure regular notification of any degraded arrays.
1636 .BR \-t ", " \-\-test
1639 alert for every array found at startup. This alert gets mailed and
1640 passed to the alert program. This can be used for testing that alert
1641 message do get through successfully.
1645 This inhibits the functionality for moving spares between arrays.
1646 Only one monitoring process started with
1648 but without this flag is allowed, otherwise the two could interfere
1655 .B mdadm \-\-assemble
1656 .I md-device options-and-component-devices...
1659 .B mdadm \-\-assemble \-\-scan
1660 .I md-devices-and-options...
1663 .B mdadm \-\-assemble \-\-scan
1667 This usage assembles one or more RAID arrays from pre-existing components.
1668 For each array, mdadm needs to know the md device, the identity of the
1669 array, and a number of component-devices. These can be found in a number of ways.
1671 In the first usage example (without the
1673 the first device given is the md device.
1674 In the second usage example, all devices listed are treated as md
1675 devices and assembly is attempted.
1676 In the third (where no devices are listed) all md devices that are
1677 listed in the configuration file are assembled. If no arrays are
1678 described by the configuration file, then any arrays that
1679 can be found on unused devices will be assembled.
1681 If precisely one device is listed, but
1687 was given and identity information is extracted from the configuration file.
1689 The identity can be given with the
1695 option, will be taken from the md-device record in the config file, or
1696 will be taken from the super block of the first component-device
1697 listed on the command line.
1699 Devices can be given on the
1701 command line or in the config file. Only devices which have an md
1702 superblock which contains the right identity will be considered for
1705 The config file is only used if explicitly named with
1707 or requested with (a possibly implicit)
1712 .B /etc/mdadm/mdadm.conf
1717 is not given, then the config file will only be used to find the
1718 identity of md arrays.
1720 Normally the array will be started after it is assembled. However if
1722 is not given and not all expected drives were listed, then the array
1723 is not started (to guard against usage errors). To insist that the
1724 array be started in this case (as may work for RAID1, 4, 5, 6, or 10),
1733 does not create any entries in
1737 It does record information in
1741 to choose the correct name.
1745 detects that udev is not configured, it will create the devices in
1749 In Linux kernels prior to version 2.6.28 there were two distinctly
1750 different types of md devices that could be created: one that could be
1751 partitioned using standard partitioning tools and one that could not.
1752 Since 2.6.28 that distinction is no longer relevant as both type of
1753 devices can be partitioned.
1755 will normally create the type that originally could not be partitioned
1756 as it has a well defined major number (9).
1758 Prior to 2.6.28, it is important that mdadm chooses the correct type
1759 of array device to use. This can be controlled with the
1761 option. In particular, a value of "mdp" or "part" or "p" tells mdadm
1762 to use a partitionable device rather than the default.
1764 In the no-udev case, the value given to
1766 can be suffixed by a number. This tells
1768 to create that number of partition devices rather than the default of 4.
1772 can also be given in the configuration file as a word starting
1774 on the ARRAY line for the relevant array.
1781 and no devices are listed,
1783 will first attempt to assemble all the arrays listed in the config
1786 If no arrays are listed in the config (other than those marked
1788 it will look through the available devices for possible arrays and
1789 will try to assemble anything that it finds. Arrays which are tagged
1790 as belonging to the given homehost will be assembled and started
1791 normally. Arrays which do not obviously belong to this host are given
1792 names that are expected not to conflict with anything local, and are
1793 started "read-auto" so that nothing is written to any device until the
1794 array is written to. i.e. automatic resync etc is delayed.
1798 finds a consistent set of devices that look like they should comprise
1799 an array, and if the superblock is tagged as belonging to the given
1800 home host, it will automatically choose a device name and try to
1801 assemble the array. If the array uses version-0.90 metadata, then the
1803 number as recorded in the superblock is used to create a name in
1807 If the array uses version-1 metadata, then the
1809 from the superblock is used to similarly create a name in
1811 (the name will have any 'host' prefix stripped first).
1813 This behaviour can be modified by the
1817 configuration file. This line can indicate that specific metadata
1818 type should, or should not, be automatically assembled. If an array
1819 is found which is not listed in
1821 and has a metadata format that is denied by the
1823 line, then it will not be assembled.
1826 line can also request that all arrays identified as being for this
1827 homehost should be assembled regardless of their metadata type.
1830 for further details.
1832 Note: Auto assembly cannot be used for assembling and activating some
1833 arrays which are undergoing reshape. In particular as the
1835 cannot be given, any reshape which requires a backup-file to continue
1836 cannot be started by auto assembly. An array which is growing to more
1837 devices and has passed the critical section can be assembled using
1848 .BI \-\-raid\-devices= Z
1852 This usage is similar to
1854 The difference is that it creates an array without a superblock. With
1855 these arrays there is no difference between initially creating the array and
1856 subsequently assembling the array, except that hopefully there is useful
1857 data there in the second case.
1859 The level may raid0, linear, raid1, raid10, multipath, or faulty, or
1860 one of their synonyms. All devices must be listed and the array will
1861 be started once complete. It will often be appropriate to use
1862 .B \-\-assume\-clean
1863 with levels raid1 or raid10.
1874 .BI \-\-raid\-devices= Z
1878 This usage will initialise a new md array, associate some devices with
1879 it, and activate the array.
1881 The named device will normally not exist when
1882 .I "mdadm \-\-create"
1883 is run, but will be created by
1885 once the array becomes active.
1887 As devices are added, they are checked to see if they contain RAID
1888 superblocks or filesystems. They are also checked to see if the variance in
1889 device size exceeds 1%.
1891 If any discrepancy is found, the array will not automatically be run, though
1894 can override this caution.
1896 To create a "degraded" array in which some devices are missing, simply
1897 give the word "\fBmissing\fP"
1898 in place of a device name. This will cause
1900 to leave the corresponding slot in the array empty.
1901 For a RAID4 or RAID5 array at most one slot can be
1902 "\fBmissing\fP"; for a RAID6 array at most two slots.
1903 For a RAID1 array, only one real device needs to be given. All of the
1907 When creating a RAID5 array,
1909 will automatically create a degraded array with an extra spare drive.
1910 This is because building the spare into a degraded array is in general
1911 faster than resyncing the parity on a non-degraded, but not clean,
1912 array. This feature can be overridden with the
1916 When creating an array with version-1 metadata a name for the array is
1918 If this is not given with the
1922 will choose a name based on the last component of the name of the
1923 device being created. So if
1925 is being created, then the name
1930 is being created, then the name
1934 When creating a partition based array, using
1936 with version-1.x metadata, the partition type should be set to
1938 (non fs-data). This type selection allows for greater precision since
1939 using any other [RAID auto-detect (0xFD) or a GNU/Linux partition (0x83)],
1940 might create problems in the event of array recovery through a live cdrom.
1942 A new array will normally get a randomly assigned 128bit UUID which is
1943 very likely to be unique. If you have a specific need, you can choose
1944 a UUID for the array by giving the
1946 option. Be warned that creating two arrays with the same UUID is a
1947 recipe for disaster. Also, using
1949 when creating a v0.90 array will silently override any
1954 .\"option is given, it is not necessary to list any component-devices in this command.
1955 .\"They can be added later, before a
1959 .\"is given, the apparent size of the smallest drive given is used.
1961 If the array type supports a write-intent bitmap, and if the devices
1962 in the array exceed 100G is size, an internal write-intent bitmap
1963 will automatically be added unless some other option is explicitly
1966 option. In any case space for a bitmap will be reserved so that one
1967 can be added layer with
1968 .BR "\-\-grow \-\-bitmap=internal" .
1970 If the metadata type supports it (currently only 1.x metadata), space
1971 will be allocated to store a bad block list. This allows a modest
1972 number of bad blocks to be recorded, allowing the drive to remain in
1973 service while only partially functional.
1975 When creating an array within a
1978 can be given either the list of devices to use, or simply the name of
1979 the container. The former case gives control over which devices in
1980 the container will be used for the array. The latter case allows
1982 to automatically choose which devices to use based on how much spare
1985 The General Management options that are valid with
1990 insist on running the array even if some devices look like they might
1995 start the array readonly \(em not supported yet.
2002 .I options... devices...
2005 This usage will allow individual devices in an array to be failed,
2006 removed or added. It is possible to perform multiple operations with
2007 on command. For example:
2009 .B " mdadm /dev/md0 \-f /dev/hda1 \-r /dev/hda1 \-a /dev/hda1"
2015 and will then remove it from the array and finally add it back
2016 in as a spare. However only one md array can be affected by a single
2019 When a device is added to an active array, mdadm checks to see if it
2020 has metadata on it which suggests that it was recently a member of the
2021 array. If it does, it tries to "re\-add" the device. If there have
2022 been no changes since the device was removed, or if the array has a
2023 write-intent bitmap which has recorded whatever changes there were,
2024 then the device will immediately become a full member of the array and
2025 those differences recorded in the bitmap will be resolved.
2035 MISC mode includes a number of distinct operations that
2036 operate on distinct devices. The operations are:
2039 The device is examined to see if it is
2040 (1) an active md array, or
2041 (2) a component of an md array.
2042 The information discovered is reported.
2046 The device should be an active md device.
2048 will display a detailed description of the array.
2052 will cause the output to be less detailed and the format to be
2053 suitable for inclusion in
2057 will normally be 0 unless
2059 failed to get useful information about the device(s); however, if the
2061 option is given, then the exit status will be:
2065 The array is functioning normally.
2068 The array has at least one failed device.
2071 The array has multiple failed devices such that it is unusable.
2074 There was an error while trying to get information about the device.
2078 .B \-\-detail\-platform
2079 Print detail of the platform's RAID capabilities (firmware / hardware
2080 topology). If the metadata is specified with
2084 then the return status will be:
2088 metadata successfully enumerated its platform components on this system
2091 metadata is platform independent
2094 metadata failed to find its platform components on this system
2098 .B \-\-update\-subarray=
2099 If the device is a container and the argument to \-\-update\-subarray
2100 specifies a subarray in the container, then attempt to update the given
2101 superblock field in the subarray. Similar to updating an array in
2102 "assemble" mode, the field to update is selected by
2106 option. Currently only
2112 option updates the subarray name in the metadata, it may not affect the
2113 device node name or the device node symlink until the subarray is
2114 re\-assembled. If updating
2116 would change the UUID of an active subarray this operation is blocked,
2117 and the command will end in an error.
2121 The device should be a component of an md array.
2123 will read the md superblock of the device and display the contents.
2128 is given, then multiple devices that are components of the one array
2129 are grouped together and reported in a single entry suitable
2135 without listing any devices will cause all devices listed in the
2136 config file to be examined.
2140 The devices should be active md arrays which will be deactivated, as
2141 long as they are not currently in use.
2145 This will fully activate a partially assembled md array.
2149 This will mark an active array as read-only, providing that it is
2150 not currently being used.
2156 array back to being read/write.
2160 For all operations except
2163 will cause the operation to be applied to all arrays listed in
2168 causes all devices listed in the config file to be examined.
2171 .BR \-b ", " \-\-brief
2172 Be less verbose. This is used with
2180 gives an intermediate level of verbosity.
2186 .B mdadm \-\-monitor
2187 .I options... devices...
2192 to periodically poll a number of md arrays and to report on any events
2195 will never exit once it decides that there are arrays to be checked,
2196 so it should normally be run in the background.
2198 As well as reporting events,
2200 may move a spare drive from one array to another if they are in the
2205 and if the destination array has a failed drive but no spares.
2207 If any devices are listed on the command line,
2209 will only monitor those devices. Otherwise all arrays listed in the
2210 configuration file will be monitored. Further, if
2212 is given, then any other md devices that appear in
2214 will also be monitored.
2216 The result of monitoring the arrays is the generation of events.
2217 These events are passed to a separate program (if specified) and may
2218 be mailed to a given E-mail address.
2220 When passing events to a program, the program is run once for each event,
2221 and is given 2 or 3 command-line arguments: the first is the
2222 name of the event (see below), the second is the name of the
2223 md device which is affected, and the third is the name of a related
2224 device if relevant (such as a component device that has failed).
2228 is given, then a program or an E-mail address must be specified on the
2229 command line or in the config file. If neither are available, then
2231 will not monitor anything.
2235 will continue monitoring as long as something was found to monitor. If
2236 no program or email is given, then each event is reported to
2239 The different events are:
2243 .B DeviceDisappeared
2244 An md array which previously was configured appears to no longer be
2245 configured. (syslog priority: Critical)
2249 was told to monitor an array which is RAID0 or Linear, then it will
2251 .B DeviceDisappeared
2252 with the extra information
2254 This is because RAID0 and Linear do not support the device-failed,
2255 hot-spare and resync operations which are monitored.
2259 An md array started reconstruction. (syslog priority: Warning)
2265 is a two-digit number (ie. 05, 48). This indicates that rebuild
2266 has passed that many percent of the total. The events are generated
2267 with fixed increment since 0. Increment size may be specified with
2268 a commandline option (default is 20). (syslog priority: Warning)
2272 An md array that was rebuilding, isn't any more, either because it
2273 finished normally or was aborted. (syslog priority: Warning)
2277 An active component device of an array has been marked as
2278 faulty. (syslog priority: Critical)
2282 A spare component device which was being rebuilt to replace a faulty
2283 device has failed. (syslog priority: Critical)
2287 A spare component device which was being rebuilt to replace a faulty
2288 device has been successfully rebuilt and has been made active.
2289 (syslog priority: Info)
2293 A new md array has been detected in the
2295 file. (syslog priority: Info)
2299 A newly noticed array appears to be degraded. This message is not
2302 notices a drive failure which causes degradation, but only when
2304 notices that an array is degraded when it first sees the array.
2305 (syslog priority: Critical)
2309 A spare drive has been moved from one array in a
2313 to another to allow a failed drive to be replaced.
2314 (syslog priority: Info)
2320 has been told, via the config file, that an array should have a certain
2321 number of spare devices, and
2323 detects that it has fewer than this number when it first sees the
2324 array, it will report a
2327 (syslog priority: Warning)
2331 An array was found at startup, and the
2334 (syslog priority: Info)
2344 cause Email to be sent. All events cause the program to be run.
2345 The program is run with two or three arguments: the event
2346 name, the array device and possibly a second device.
2348 Each event has an associated array device (e.g.
2350 and possibly a second device. For
2355 the second device is the relevant component device.
2358 the second device is the array that the spare was moved from.
2362 to move spares from one array to another, the different arrays need to
2363 be labeled with the same
2365 or the spares must be allowed to migrate through matching POLICY domains
2366 in the configuration file. The
2368 name can be any string; it is only necessary that different spare
2369 groups use different names.
2373 detects that an array in a spare group has fewer active
2374 devices than necessary for the complete array, and has no spare
2375 devices, it will look for another array in the same spare group that
2376 has a full complement of working drive and a spare. It will then
2377 attempt to remove the spare from the second drive and add it to the
2379 If the removal succeeds but the adding fails, then it is added back to
2382 If the spare group for a degraded array is not defined,
2384 will look at the rules of spare migration specified by POLICY lines in
2386 and then follow similar steps as above if a matching spare is found.
2389 The GROW mode is used for changing the size or shape of an active
2391 For this to work, the kernel must support the necessary change.
2392 Various types of growth are being added during 2.6 development.
2394 Currently the supported changes include
2396 change the "size" attribute for RAID1, RAID4, RAID5 and RAID6.
2398 increase or decrease the "raid\-devices" attribute of RAID0, RAID1, RAID4,
2401 change the chunk-size and layout of RAID0, RAID4, RAID5 and RAID6.
2403 convert between RAID1 and RAID5, between RAID5 and RAID6, between
2404 RAID0, RAID4, and RAID5, and between RAID0 and RAID10 (in the near-2 mode).
2406 add a write-intent bitmap to any array which supports these bitmaps, or
2407 remove a write-intent bitmap from such an array.
2410 Using GROW on containers is currently supported only for Intel's IMSM
2411 container format. The number of devices in a container can be
2412 increased - which affects all arrays in the container - or an array
2413 in a container can be converted between levels where those levels are
2414 supported by the container, and the conversion is on of those listed
2415 above. Resizing arrays in an IMSM container with
2417 is not yet supported.
2419 Grow functionality (e.g. expand a number of raid devices) for Intel's
2420 IMSM container format has an experimental status. It is guarded by the
2421 .B MDADM_EXPERIMENTAL
2422 environment variable which must be set to '1' for a GROW command to
2424 This is for the following reasons:
2427 Intel's native IMSM check-pointing is not fully tested yet.
2428 This can causes IMSM incompatibility during the grow process: an array
2429 which is growing cannot roam between Microsoft Windows(R) and Linux
2433 Interrupting a grow operation is not recommended, because it
2434 has not been fully tested for Intel's IMSM container format yet.
2437 Note: Intel's native checkpointing doesn't use
2439 option and it is transparent for assembly feature.
2442 Normally when an array is built the "size" is taken from the smallest
2443 of the drives. If all the small drives in an arrays are, one at a
2444 time, removed and replaced with larger drives, then you could have an
2445 array of large drives with only a small amount used. In this
2446 situation, changing the "size" with "GROW" mode will allow the extra
2447 space to start being used. If the size is increased in this way, a
2448 "resync" process will start to make sure the new parts of the array
2451 Note that when an array changes size, any filesystem that may be
2452 stored in the array will not automatically grow or shrink to use or
2453 vacate the space. The
2454 filesystem will need to be explicitly told to use the extra space
2455 after growing, or to reduce its size
2457 to shrinking the array.
2459 Also the size of an array cannot be changed while it has an active
2460 bitmap. If an array has a bitmap, it must be removed before the size
2461 can be changed. Once the change is complete a new bitmap can be created.
2463 .SS RAID\-DEVICES CHANGES
2465 A RAID1 array can work with any number of devices from 1 upwards
2466 (though 1 is not very useful). There may be times which you want to
2467 increase or decrease the number of active devices. Note that this is
2468 different to hot-add or hot-remove which changes the number of
2471 When reducing the number of devices in a RAID1 array, the slots which
2472 are to be removed from the array must already be vacant. That is, the
2473 devices which were in those slots must be failed and removed.
2475 When the number of devices is increased, any hot spares that are
2476 present will be activated immediately.
2478 Changing the number of active devices in a RAID5 or RAID6 is much more
2479 effort. Every block in the array will need to be read and written
2480 back to a new location. From 2.6.17, the Linux Kernel is able to
2481 increase the number of devices in a RAID5 safely, including restarting
2482 an interrupted "reshape". From 2.6.31, the Linux Kernel is able to
2483 increase or decrease the number of devices in a RAID5 or RAID6.
2485 From 2.6.35, the Linux Kernel is able to convert a RAID0 in to a RAID4
2488 uses this functionality and the ability to add
2489 devices to a RAID4 to allow devices to be added to a RAID0. When
2490 requested to do this,
2492 will convert the RAID0 to a RAID4, add the necessary disks and make
2493 the reshape happen, and then convert the RAID4 back to RAID0.
2495 When decreasing the number of devices, the size of the array will also
2496 decrease. If there was data in the array, it could get destroyed and
2497 this is not reversible, so you should firstly shrink the filesystem on
2498 the array to fit within the new size. To help prevent accidents,
2500 requires that the size of the array be decreased first with
2501 .BR "mdadm --grow --array-size" .
2502 This is a reversible change which simply makes the end of the array
2503 inaccessible. The integrity of any data can then be checked before
2504 the non-reversible reduction in the number of devices is request.
2506 When relocating the first few stripes on a RAID5 or RAID6, it is not
2507 possible to keep the data on disk completely consistent and
2508 crash-proof. To provide the required safety, mdadm disables writes to
2509 the array while this "critical section" is reshaped, and takes a
2510 backup of the data that is in that section. For grows, this backup may be
2511 stored in any spare devices that the array has, however it can also be
2512 stored in a separate file specified with the
2514 option, and is required to be specified for shrinks, RAID level
2515 changes and layout changes. If this option is used, and the system
2516 does crash during the critical period, the same file must be passed to
2518 to restore the backup and reassemble the array. When shrinking rather
2519 than growing the array, the reshape is done from the end towards the
2520 beginning, so the "critical section" is at the end of the reshape.
2524 Changing the RAID level of any array happens instantaneously. However
2525 in the RAID5 to RAID6 case this requires a non-standard layout of the
2526 RAID6 data, and in the RAID6 to RAID5 case that non-standard layout is
2527 required before the change can be accomplished. So while the level
2528 change is instant, the accompanying layout change can take quite a
2531 is required. If the array is not simultaneously being grown or
2532 shrunk, so that the array size will remain the same - for example,
2533 reshaping a 3-drive RAID5 into a 4-drive RAID6 - the backup file will
2534 be used not just for a "cricital section" but throughout the reshape
2535 operation, as described below under LAYOUT CHANGES.
2537 .SS CHUNK-SIZE AND LAYOUT CHANGES
2539 Changing the chunk-size of layout without also changing the number of
2540 devices as the same time will involve re-writing all blocks in-place.
2541 To ensure against data loss in the case of a crash, a
2543 must be provided for these changes. Small sections of the array will
2544 be copied to the backup file while they are being rearranged. This
2545 means that all the data is copied twice, once to the backup and once
2546 to the new layout on the array, so this type of reshape will go very
2549 If the reshape is interrupted for any reason, this backup file must be
2551 .B "mdadm --assemble"
2552 so the array can be reassembled. Consequently the file cannot be
2553 stored on the device being reshaped.
2558 A write-intent bitmap can be added to, or removed from, an active
2559 array. Either internal bitmaps, or bitmaps stored in a separate file,
2560 can be added. Note that if you add a bitmap stored in a file which is
2561 in a filesystem that is on the RAID array being affected, the system
2562 will deadlock. The bitmap must be on a separate filesystem.
2564 .SH INCREMENTAL MODE
2568 .B mdadm \-\-incremental
2574 .B mdadm \-\-incremental \-\-fail
2578 .B mdadm \-\-incremental \-\-rebuild\-map
2581 .B mdadm \-\-incremental \-\-run \-\-scan
2584 This mode is designed to be used in conjunction with a device
2585 discovery system. As devices are found in a system, they can be
2587 .B "mdadm \-\-incremental"
2588 to be conditionally added to an appropriate array.
2590 Conversely, it can also be used with the
2592 flag to do just the opposite and find whatever array a particular device
2593 is part of and remove the device from that array.
2595 If the device passed is a
2597 device created by a previous call to
2599 then rather than trying to add that device to an array, all the arrays
2600 described by the metadata of the container will be started.
2603 performs a number of tests to determine if the device is part of an
2604 array, and which array it should be part of. If an appropriate array
2605 is found, or can be created,
2607 adds the device to the array and conditionally starts the array.
2611 will normally only add devices to an array which were previously working
2612 (active or spare) parts of that array. The support for automatic
2613 inclusion of a new drive as a spare in some array requires
2614 a configuration through POLICY in config file.
2618 makes are as follow:
2620 Is the device permitted by
2622 That is, is it listed in a
2624 line in that file. If
2626 is absent then the default it to allow any device. Similar if
2628 contains the special word
2630 then any device is allowed. Otherwise the device name given to
2632 must match one of the names or patterns in a
2637 Does the device have a valid md superblock? If a specific metadata
2638 version is requested with
2642 then only that style of metadata is accepted, otherwise
2644 finds any known version of metadata. If no
2646 metadata is found, the device may be still added to an array
2647 as a spare if POLICY allows.
2651 Does the metadata match an expected array?
2652 The metadata can match in two ways. Either there is an array listed
2655 which identifies the array (either by UUID, by name, by device list,
2656 or by minor-number), or the array was created with a
2662 or on the command line.
2665 is not able to positively identify the array as belonging to the
2666 current host, the device will be rejected.
2671 keeps a list of arrays that it has partially assembled in
2673 If no array exists which matches
2674 the metadata on the new device,
2676 must choose a device name and unit number. It does this based on any
2679 or any name information stored in the metadata. If this name
2680 suggests a unit number, that number will be used, otherwise a free
2681 unit number will be chosen. Normally
2683 will prefer to create a partitionable array, however if the
2687 suggests that a non-partitionable array is preferred, that will be
2690 If the array is not found in the config file and its metadata does not
2691 identify it as belonging to the "homehost", then
2693 will choose a name for the array which is certain not to conflict with
2694 any array which does belong to this host. It does this be adding an
2695 underscore and a small number to the name preferred by the metadata.
2697 Once an appropriate array is found or created and the device is added,
2699 must decide if the array is ready to be started. It will
2700 normally compare the number of available (non-spare) devices to the
2701 number of devices that the metadata suggests need to be active. If
2702 there are at least that many, the array will be started. This means
2703 that if any devices are missing the array will not be restarted.
2709 in which case the array will be run as soon as there are enough
2710 devices present for the data to be accessible. For a RAID1, that
2711 means one device will start the array. For a clean RAID5, the array
2712 will be started as soon as all but one drive is present.
2714 Note that neither of these approaches is really ideal. If it can
2715 be known that all device discovery has completed, then
2719 can be run which will try to start all arrays that are being
2720 incrementally assembled. They are started in "read-auto" mode in
2721 which they are read-only until the first write request. This means
2722 that no metadata updates are made and no attempt at resync or recovery
2723 happens. Further devices that are found before the first write can
2724 still be added safely.
2727 This section describes environment variables that affect how mdadm
2732 Setting this value to 1 will prevent mdadm from automatically launching
2733 mdmon. This variable is intended primarily for debugging mdadm/mdmon.
2739 does not create any device nodes in /dev, but leaves that task to
2743 appears not to be configured, or if this environment variable is set
2746 will create and devices that are needed.
2750 A key value of IMSM metadata is that it allows interoperability with
2751 boot ROMs on Intel platforms, and with other major operating systems.
2754 will only allow an IMSM array to be created or modified if detects
2755 that it is running on an Intel platform which supports IMSM, and
2756 supports the particular configuration of IMSM that is being requested
2757 (some functionality requires newer OROM support).
2759 These checks can be suppressed by setting IMSM_NO_PLATFORM=1 in the
2760 environment. This can be useful for testing or for disaster
2761 recovery. You should be aware that interoperability may be
2762 compromised by setting this value.
2765 .B " mdadm \-\-query /dev/name-of-device"
2767 This will find out if a given device is a RAID array, or is part of
2768 one, and will provide brief information about the device.
2770 .B " mdadm \-\-assemble \-\-scan"
2772 This will assemble and start all arrays listed in the standard config
2773 file. This command will typically go in a system startup file.
2775 .B " mdadm \-\-stop \-\-scan"
2777 This will shut down all arrays that can be shut down (i.e. are not
2778 currently in use). This will typically go in a system shutdown script.
2780 .B " mdadm \-\-follow \-\-scan \-\-delay=120"
2782 If (and only if) there is an Email address or program given in the
2783 standard config file, then
2784 monitor the status of all arrays listed in that file by
2785 polling them ever 2 minutes.
2787 .B " mdadm \-\-create /dev/md0 \-\-level=1 \-\-raid\-devices=2 /dev/hd[ac]1"
2789 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
2792 .B " echo 'DEVICE /dev/hd*[0\-9] /dev/sd*[0\-9]' > mdadm.conf"
2794 .B " mdadm \-\-detail \-\-scan >> mdadm.conf"
2796 This will create a prototype config file that describes currently
2797 active arrays that are known to be made from partitions of IDE or SCSI drives.
2798 This file should be reviewed before being used as it may
2799 contain unwanted detail.
2801 .B " echo 'DEVICE /dev/hd[a\-z] /dev/sd*[a\-z]' > mdadm.conf"
2803 .B " mdadm \-\-examine \-\-scan \-\-config=mdadm.conf >> mdadm.conf"
2805 This will find arrays which could be assembled from existing IDE and
2806 SCSI whole drives (not partitions), and store the information in the
2807 format of a config file.
2808 This file is very likely to contain unwanted detail, particularly
2811 entries. It should be reviewed and edited before being used as an
2814 .B " mdadm \-\-examine \-\-brief \-\-scan \-\-config=partitions"
2816 .B " mdadm \-Ebsc partitions"
2818 Create a list of devices by reading
2819 .BR /proc/partitions ,
2820 scan these for RAID superblocks, and printout a brief listing of all
2823 .B " mdadm \-Ac partitions \-m 0 /dev/md0"
2825 Scan all partitions and devices listed in
2826 .BR /proc/partitions
2829 out of all such devices with a RAID superblock with a minor number of 0.
2831 .B " mdadm \-\-monitor \-\-scan \-\-daemonise > /run/mdadm/mon.pid"
2833 If config file contains a mail address or alert program, run mdadm in
2834 the background in monitor mode monitoring all md devices. Also write
2835 pid of mdadm daemon to
2836 .BR /run/mdadm/mon.pid .
2838 .B " mdadm \-Iq /dev/somedevice"
2840 Try to incorporate newly discovered device into some array as
2843 .B " mdadm \-\-incremental \-\-rebuild\-map \-\-run \-\-scan"
2845 Rebuild the array map from any current arrays, and then start any that
2848 .B " mdadm /dev/md4 --fail detached --remove detached"
2850 Any devices which are components of /dev/md4 will be marked as faulty
2851 and then remove from the array.
2853 .B " mdadm --grow /dev/md4 --level=6 --backup-file=/root/backup-md4"
2857 which is currently a RAID5 array will be converted to RAID6. There
2858 should normally already be a spare drive attached to the array as a
2859 RAID6 needs one more drive than a matching RAID5.
2861 .B " mdadm --create /dev/md/ddf --metadata=ddf --raid-disks 6 /dev/sd[a-f]"
2863 Create a DDF array over 6 devices.
2865 .B " mdadm --create /dev/md/home -n3 -l5 -z 30000000 /dev/md/ddf"
2867 Create a RAID5 array over any 3 devices in the given DDF set. Use
2868 only 30 gigabytes of each device.
2870 .B " mdadm -A /dev/md/ddf1 /dev/sd[a-f]"
2872 Assemble a pre-exist ddf array.
2874 .B " mdadm -I /dev/md/ddf1"
2876 Assemble all arrays contained in the ddf array, assigning names as
2879 .B " mdadm \-\-create \-\-help"
2881 Provide help about the Create mode.
2883 .B " mdadm \-\-config \-\-help"
2885 Provide help about the format of the config file.
2887 .B " mdadm \-\-help"
2889 Provide general help.
2899 lists all active md devices with information about them.
2901 uses this to find arrays when
2903 is given in Misc mode, and to monitor array reconstruction
2908 The config file lists which devices may be scanned to see if
2909 they contain MD super block, and gives identifying information
2910 (e.g. UUID) about known MD arrays. See
2917 mode is used, this file gets a list of arrays currently being created.
2922 understand two sorts of names for array devices.
2924 The first is the so-called 'standard' format name, which matches the
2925 names used by the kernel and which appear in
2928 The second sort can be freely chosen, but must reside in
2930 When giving a device name to
2932 to create or assemble an array, either full path name such as
2936 can be given, or just the suffix of the second sort of name, such as
2942 chooses device names during auto-assembly or incremental assembly, it
2943 will sometimes add a small sequence number to the end of the name to
2944 avoid conflicted between multiple arrays that have the same name. If
2946 can reasonably determine that the array really is meant for this host,
2947 either by a hostname in the metadata, or by the presence of the array
2950 then it will leave off the suffix if possible.
2951 Also if the homehost is specified as
2954 will only use a suffix if a different array of the same name already
2955 exists or is listed in the config file.
2957 The standard names for non-partitioned arrays (the only sort of md
2958 array available in 2.4 and earlier) are of the form
2962 where NN is a number.
2963 The standard names for partitionable arrays (as available from 2.6
2964 onwards) are of the form:
2968 Partition numbers should be indicated by adding "pMM" to these, thus "/dev/md/d1p2".
2970 From kernel version 2.6.28 the "non-partitioned array" can actually
2971 be partitioned. So the "md_d\fBNN\fP"
2972 names are no longer needed, and
2973 partitions such as "/dev/md\fBNN\fPp\fBXX\fp"
2976 From kernel version 2.6.29 standard names can be non-numeric following
2983 is any string. These names are supported by
2985 since version 3.3 provided they are enabled in
2990 was previously known as
2994 For further information on mdadm usage, MD and the various levels of
2997 .B http://raid.wiki.kernel.org/
2999 (based upon Jakob \(/Ostergaard's Software\-RAID.HOWTO)
3001 The latest version of
3003 should always be available from
3005 .B http://www.kernel.org/pub/linux/utils/raid/mdadm/