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 Set first character of argv[0] to @ to indicate mdadm was launched
261 from initrd/initramfs and should not be shutdown by systemd as part of
262 the regular shutdown process. This option is normally only used by
263 the system's initscripts. Please see here for more details on how
264 systemd handled argv[0]:
266 .B http://www.freedesktop.org/wiki/Software/systemd/RootStorageDaemons
271 .BR \-f ", " \-\-force
272 Be more forceful about certain operations. See the various modes for
273 the exact meaning of this option in different contexts.
276 .BR \-c ", " \-\-config=
277 Specify the config file. Default is to use
278 .BR /etc/mdadm.conf ,
279 or if that is missing then
280 .BR /etc/mdadm/mdadm.conf .
281 If the config file given is
283 then nothing will be read, but
285 will act as though the config file contained exactly
286 .B "DEVICE partitions containers"
289 to find a list of devices to scan, and
291 to find a list of containers to examine.
294 is given for the config file, then
296 will act as though the config file were empty.
299 .BR \-s ", " \-\-scan
302 for missing information.
303 In general, this option gives
305 permission to get any missing information (like component devices,
306 array devices, array identities, and alert destination) from the
307 configuration file (see previous option);
308 one exception is MISC mode when using
314 says to get a list of array devices from
318 .BR \-e ", " \-\-metadata=
319 Declare the style of RAID metadata (superblock) to be used. The
320 default is {DEFAULT_METADATA} for
322 and to guess for other operations.
323 The default can be overridden by setting the
332 .ie '{DEFAULT_METADATA}'0.90'
333 .IP "0, 0.90, default"
336 Use the original 0.90 format superblock. This format limits arrays to
337 28 component devices and limits component devices of levels 1 and
338 greater to 2 terabytes. It is also possible for there to be confusion
339 about whether the superblock applies to a whole device or just the
340 last partition, if that partition starts on a 64K boundary.
341 .ie '{DEFAULT_METADATA}'0.90'
342 .IP "1, 1.0, 1.1, 1.2"
344 .IP "1, 1.0, 1.1, 1.2 default"
345 Use the new version-1 format superblock. This has fewer restrictions.
346 It can easily be moved between hosts with different endian-ness, and a
347 recovery operation can be checkpointed and restarted. The different
348 sub-versions store the superblock at different locations on the
349 device, either at the end (for 1.0), at the start (for 1.1) or 4K from
350 the start (for 1.2). "1" is equivalent to "1.2" (the commonly
351 preferred 1.x format).
352 'if '{DEFAULT_METADATA}'1.2' "default" is equivalent to "1.2".
354 Use the "Industry Standard" DDF (Disk Data Format) format defined by
356 When creating a DDF array a
358 will be created, and normal arrays can be created in that container.
360 Use the Intel(R) Matrix Storage Manager metadata format. This creates a
362 which is managed in a similar manner to DDF, and is supported by an
363 option-rom on some platforms:
365 .B http://www.intel.com/design/chipsets/matrixstorage_sb.htm
371 This will override any
373 setting in the config file and provides the identity of the host which
374 should be considered the home for any arrays.
376 When creating an array, the
378 will be recorded in the metadata. For version-1 superblocks, it will
379 be prefixed to the array name. For version-0.90 superblocks, part of
380 the SHA1 hash of the hostname will be stored in the later half of the
383 When reporting information about an array, any array which is tagged
384 for the given homehost will be reported as such.
386 When using Auto-Assemble, only arrays tagged for the given homehost
387 will be allowed to use 'local' names (i.e. not ending in '_' followed
388 by a digit string). See below under
389 .BR "Auto Assembly" .
395 needs to print the name for a device it normally finds the name in
397 which refers to the device and is shortest. When a path component is
401 will prefer a longer name if it contains that component. For example
402 .B \-\-prefer=by-uuid
403 will prefer a name in a subdirectory of
408 This functionality is currently only provided by
413 .SH For create, build, or grow:
416 .BR \-n ", " \-\-raid\-devices=
417 Specify the number of active devices in the array. This, plus the
418 number of spare devices (see below) must equal the number of
420 (including "\fBmissing\fP" devices)
421 that are listed on the command line for
423 Setting a value of 1 is probably
424 a mistake and so requires that
426 be specified first. A value of 1 will then be allowed for linear,
427 multipath, RAID0 and RAID1. It is never allowed for RAID4, RAID5 or RAID6.
429 This number can only be changed using
431 for RAID1, RAID4, RAID5 and RAID6 arrays, and only on kernels which provide
432 the necessary support.
435 .BR \-x ", " \-\-spare\-devices=
436 Specify the number of spare (eXtra) devices in the initial array.
437 Spares can also be added
438 and removed later. The number of component devices listed
439 on the command line must equal the number of RAID devices plus the
440 number of spare devices.
443 .BR \-z ", " \-\-size=
444 Amount (in Kibibytes) of space to use from each drive in RAID levels 1/4/5/6.
445 This must be a multiple of the chunk size, and must leave about 128Kb
446 of space at the end of the drive for the RAID superblock.
447 If this is not specified
448 (as it normally is not) the smallest drive (or partition) sets the
449 size, though if there is a variance among the drives of greater than 1%, a warning is
452 A suffix of 'M' or 'G' can be given to indicate Megabytes or
453 Gigabytes respectively.
455 Sometimes a replacement drive can be a little smaller than the
456 original drives though this should be minimised by IDEMA standards.
457 Such a replacement drive will be rejected by
459 To guard against this it can be useful to set the initial size
460 slightly smaller than the smaller device with the aim that it will
461 still be larger than any replacement.
463 This value can be set with
465 for RAID level 1/4/5/6 though
467 based arrays such as those with IMSM metadata may not be able to
469 If the array was created with a size smaller than the currently
470 active drives, the extra space can be accessed using
472 The size can be given as
474 which means to choose the largest size that fits on all current drives.
476 Before reducing the size of the array (with
477 .BR "\-\-grow \-\-size=" )
478 you should make sure that space isn't needed. If the device holds a
479 filesystem, you would need to resize the filesystem to use less space.
481 After reducing the array size you should check that the data stored in
482 the device is still available. If the device holds a filesystem, then
483 an 'fsck' of the filesystem is a minimum requirement. If there are
484 problems the array can be made bigger again with no loss with another
485 .B "\-\-grow \-\-size="
488 This value cannot be used when creating a
490 such as with DDF and IMSM metadata, though it perfectly valid when
491 creating an array inside a container.
494 .BR \-Z ", " \-\-array\-size=
495 This is only meaningful with
497 and its effect is not persistent: when the array is stopped and
498 restarted the default array size will be restored.
500 Setting the array-size causes the array to appear smaller to programs
501 that access the data. This is particularly needed before reshaping an
502 array so that it will be smaller. As the reshape is not reversible,
503 but setting the size with
505 is, it is required that the array size is reduced as appropriate
506 before the number of devices in the array is reduced.
508 Before reducing the size of the array you should make sure that space
509 isn't needed. If the device holds a filesystem, you would need to
510 resize the filesystem to use less space.
512 After reducing the array size you should check that the data stored in
513 the device is still available. If the device holds a filesystem, then
514 an 'fsck' of the filesystem is a minimum requirement. If there are
515 problems the array can be made bigger again with no loss with another
516 .B "\-\-grow \-\-array\-size="
519 A suffix of 'M' or 'G' can be given to indicate Megabytes or
520 Gigabytes respectively.
523 restores the apparent size of the array to be whatever the real
524 amount of available space is.
527 .BR \-c ", " \-\-chunk=
528 Specify chunk size of kibibytes. The default when creating an
529 array is 512KB. To ensure compatibility with earlier versions, the
530 default when Building and array with no persistent metadata is 64KB.
531 This is only meaningful for RAID0, RAID4, RAID5, RAID6, and RAID10.
533 RAID4, RAID5, RAID6, and RAID10 require the chunk size to be a power
534 of 2. In any case it must be a multiple of 4KB.
536 A suffix of 'M' or 'G' can be given to indicate Megabytes or
537 Gigabytes respectively.
541 Specify rounding factor for a Linear array. The size of each
542 component will be rounded down to a multiple of this size.
543 This is a synonym for
545 but highlights the different meaning for Linear as compared to other
546 RAID levels. The default is 64K if a kernel earlier than 2.6.16 is in
547 use, and is 0K (i.e. no rounding) in later kernels.
550 .BR \-l ", " \-\-level=
551 Set RAID level. When used with
553 options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
554 raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty, container.
555 Obviously some of these are synonymous.
559 metadata type is requested, only the
561 level is permitted, and it does not need to be explicitly given.
565 only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid.
569 to change the RAID level in some cases. See LEVEL CHANGES below.
572 .BR \-p ", " \-\-layout=
573 This option configures the fine details of data layout for RAID5, RAID6,
574 and RAID10 arrays, and controls the failure modes for
577 The layout of the RAID5 parity block can be one of
578 .BR left\-asymmetric ,
579 .BR left\-symmetric ,
580 .BR right\-asymmetric ,
581 .BR right\-symmetric ,
582 .BR la ", " ra ", " ls ", " rs .
584 .BR left\-symmetric .
586 It is also possible to cause RAID5 to use a RAID4-like layout by
592 Finally for RAID5 there are DDF\-compatible layouts,
593 .BR ddf\-zero\-restart ,
594 .BR ddf\-N\-restart ,
596 .BR ddf\-N\-continue .
598 These same layouts are available for RAID6. There are also 4 layouts
599 that will provide an intermediate stage for converting between RAID5
600 and RAID6. These provide a layout which is identical to the
601 corresponding RAID5 layout on the first N\-1 devices, and has the 'Q'
602 syndrome (the second 'parity' block used by RAID6) on the last device.
604 .BR left\-symmetric\-6 ,
605 .BR right\-symmetric\-6 ,
606 .BR left\-asymmetric\-6 ,
607 .BR right\-asymmetric\-6 ,
609 .BR parity\-first\-6 .
611 When setting the failure mode for level
614 .BR write\-transient ", " wt ,
615 .BR read\-transient ", " rt ,
616 .BR write\-persistent ", " wp ,
617 .BR read\-persistent ", " rp ,
619 .BR read\-fixable ", " rf ,
620 .BR clear ", " flush ", " none .
622 Each failure mode can be followed by a number, which is used as a period
623 between fault generation. Without a number, the fault is generated
624 once on the first relevant request. With a number, the fault will be
625 generated after that many requests, and will continue to be generated
626 every time the period elapses.
628 Multiple failure modes can be current simultaneously by using the
630 option to set subsequent failure modes.
632 "clear" or "none" will remove any pending or periodic failure modes,
633 and "flush" will clear any persistent faults.
635 Finally, the layout options for RAID10 are one of 'n', 'o' or 'f' followed
636 by a small number. The default is 'n2'. The supported options are:
639 signals 'near' copies. Multiple copies of one data block are at
640 similar offsets in different devices.
643 signals 'offset' copies. Rather than the chunks being duplicated
644 within a stripe, whole stripes are duplicated but are rotated by one
645 device so duplicate blocks are on different devices. Thus subsequent
646 copies of a block are in the next drive, and are one chunk further
651 (multiple copies have very different offsets).
652 See md(4) for more detail about 'near', 'offset', and 'far'.
654 The number is the number of copies of each datablock. 2 is normal, 3
655 can be useful. This number can be at most equal to the number of
656 devices in the array. It does not need to divide evenly into that
657 number (e.g. it is perfectly legal to have an 'n2' layout for an array
658 with an odd number of devices).
660 When an array is converted between RAID5 and RAID6 an intermediate
661 RAID6 layout is used in which the second parity block (Q) is always on
662 the last device. To convert a RAID5 to RAID6 and leave it in this new
663 layout (which does not require re-striping) use
664 .BR \-\-layout=preserve .
665 This will try to avoid any restriping.
667 The converse of this is
668 .B \-\-layout=normalise
669 which will change a non-standard RAID6 layout into a more standard
676 (thus explaining the p of
680 .BR \-b ", " \-\-bitmap=
681 Specify a file to store a write-intent bitmap in. The file should not
684 is also given. The same file should be provided
685 when assembling the array. If the word
687 is given, then the bitmap is stored with the metadata on the array,
688 and so is replicated on all devices. If the word
692 mode, then any bitmap that is present is removed.
694 To help catch typing errors, the filename must contain at least one
695 slash ('/') if it is a real file (not 'internal' or 'none').
697 Note: external bitmaps are only known to work on ext2 and ext3.
698 Storing bitmap files on other filesystems may result in serious problems.
701 .BR \-\-bitmap\-chunk=
702 Set the chunksize of the bitmap. Each bit corresponds to that many
703 Kilobytes of storage.
704 When using a file based bitmap, the default is to use the smallest
705 size that is at-least 4 and requires no more than 2^21 chunks.
708 bitmap, the chunksize defaults to 64Meg, or larger if necessary to
709 fit the bitmap into the available space.
711 A suffix of 'M' or 'G' can be given to indicate Megabytes or
712 Gigabytes respectively.
715 .BR \-W ", " \-\-write\-mostly
716 subsequent devices listed in a
721 command will be flagged as 'write-mostly'. This is valid for RAID1
722 only and means that the 'md' driver will avoid reading from these
723 devices if at all possible. This can be useful if mirroring over a
727 .BR \-\-write\-behind=
728 Specify that write-behind mode should be enabled (valid for RAID1
729 only). If an argument is specified, it will set the maximum number
730 of outstanding writes allowed. The default value is 256.
731 A write-intent bitmap is required in order to use write-behind
732 mode, and write-behind is only attempted on drives marked as
736 .BR \-\-assume\-clean
739 that the array pre-existed and is known to be clean. It can be useful
740 when trying to recover from a major failure as you can be sure that no
741 data will be affected unless you actually write to the array. It can
742 also be used when creating a RAID1 or RAID10 if you want to avoid the
743 initial resync, however this practice \(em while normally safe \(em is not
744 recommended. Use this only if you really know what you are doing.
746 When the devices that will be part of a new array were filled
747 with zeros before creation the operator knows the array is
748 actually clean. If that is the case, such as after running
749 badblocks, this argument can be used to tell mdadm the
750 facts the operator knows.
752 When an array is resized to a larger size with
753 .B "\-\-grow \-\-size="
754 the new space is normally resynced in that same way that the whole
755 array is resynced at creation. From Linux version 3.0,
757 can be used with that command to avoid the automatic resync.
760 .BR \-\-backup\-file=
763 is used to increase the number of raid-devices in a RAID5 or RAID6 if
764 there are no spare devices available, or to shrink, change RAID level
765 or layout. See the GROW MODE section below on RAID\-DEVICES CHANGES.
766 The file must be stored on a separate device, not on the RAID array
771 Arrays with 1.x metadata can leave a gap between the start of the
772 device and the start of array data. This gap can be used for various
773 metadata. The start of data is known as the
775 Normally an appropriate data offset is computed automatically.
776 However it can be useful to set it explicitly such as when re-creating
777 an array which was originally created using a different version of
779 which computed a different offset.
781 Setting the offset explicitly over-rides the default. The value given
782 is in Kilobytes unless an 'M' or 'G' suffix is given.
786 can also be used with
788 for some RAID levels (initially on RAID10). This allows the
789 data\-offset to be changed as part of the reshape process. When the
790 data offset is changed, no backup file is required as the difference
791 in offsets is used to provide the same functionality.
793 When the new offset is earlier than the old offset, the number of
794 devices in the array cannot shrink. When it is after the old offset,
795 the number of devices in the array cannot increase.
797 When creating an array,
801 In the case each member device is expected to have a offset appended
802 to the name, separated by a colon. This makes it possible to recreate
803 exactly an array which has varying data offsets (as can happen when
804 different versions of
806 are used to add different devices).
810 This option is complementary to the
811 .B \-\-freeze-reshape
812 option for assembly. It is needed when
814 operation is interrupted and it is not restarted automatically due to
815 .B \-\-freeze-reshape
816 usage during array assembly. This option is used together with
820 ) command and device for a pending reshape to be continued.
821 All parameters required for reshape continuation will be read from array metadata.
825 .BR \-\-backup\-file=
826 option to be set, continuation option will require to have exactly the same
827 backup file given as well.
829 Any other parameter passed together with
831 option will be ignored.
834 .BR \-N ", " \-\-name=
837 for the array. This is currently only effective when creating an
838 array with a version-1 superblock, or an array in a DDF container.
839 The name is a simple textual string that can be used to identify array
840 components when assembling. If name is needed but not specified, it
841 is taken from the basename of the device that is being created.
853 run the array, even if some of the components
854 appear to be active in another array or filesystem. Normally
856 will ask for confirmation before including such components in an
857 array. This option causes that question to be suppressed.
860 .BR \-f ", " \-\-force
863 accept the geometry and layout specified without question. Normally
865 will not allow creation of an array with only one device, and will try
866 to create a RAID5 array with one missing drive (as this makes the
867 initial resync work faster). With
870 will not try to be so clever.
873 .BR \-o ", " \-\-readonly
876 rather than read-write as normal. No writes will be allowed to the
877 array, and no resync, recovery, or reshape will be started.
880 .BR \-a ", " "\-\-auto{=yes,md,mdp,part,p}{NN}"
881 Instruct mdadm how to create the device file if needed, possibly allocating
882 an unused minor number. "md" causes a non-partitionable array
883 to be used (though since Linux 2.6.28, these array devices are in fact
884 partitionable). "mdp", "part" or "p" causes a partitionable array (2.6 and
885 later) to be used. "yes" requires the named md device to have
886 a 'standard' format, and the type and minor number will be determined
887 from this. With mdadm 3.0, device creation is normally left up to
889 so this option is unlikely to be needed.
890 See DEVICE NAMES below.
892 The argument can also come immediately after
897 is not given on the command line or in the config file, then
903 is also given, then any
905 entries in the config file will override the
907 instruction given on the command line.
909 For partitionable arrays,
911 will create the device file for the whole array and for the first 4
912 partitions. A different number of partitions can be specified at the
913 end of this option (e.g.
915 If the device name ends with a digit, the partition names add a 'p',
917 .IR /dev/md/home1p3 .
918 If there is no trailing digit, then the partition names just have a
920 .IR /dev/md/scratch3 .
922 If the md device name is in a 'standard' format as described in DEVICE
923 NAMES, then it will be created, if necessary, with the appropriate
924 device number based on that name. If the device name is not in one of these
925 formats, then a unused device number will be allocated. The device
926 number will be considered unused if there is no active array for that
927 number, and there is no entry in /dev for that number and with a
928 non-standard name. Names that are not in 'standard' format are only
929 allowed in "/dev/md/".
931 This is meaningful with
937 .BR \-a ", " "\-\-add"
938 This option can be used in Grow mode in two cases.
940 If the target array is a Linear array, then
942 can be used to add one or more devices to the array. They
943 are simply catenated on to the end of the array. Once added, the
944 devices cannot be removed.
948 option is being used to increase the number of devices in an array,
951 can be used to add some extra devices to be included in the array.
952 In most cases this is not needed as the extra devices can be added as
953 spares first, and then the number of raid-disks can be changed.
954 However for RAID0, it is not possible to add spares. So to increase
955 the number of devices in a RAID0, it is necessary to set the new
956 number of devices, and to add the new devices, in the same command.
961 .BR \-u ", " \-\-uuid=
962 uuid of array to assemble. Devices which don't have this uuid are
966 .BR \-m ", " \-\-super\-minor=
967 Minor number of device that array was created for. Devices which
968 don't have this minor number are excluded. If you create an array as
969 /dev/md1, then all superblocks will contain the minor number 1, even if
970 the array is later assembled as /dev/md2.
972 Giving the literal word "dev" for
976 to use the minor number of the md device that is being assembled.
979 .B \-\-super\-minor=dev
980 will look for super blocks with a minor number of 0.
983 is only relevant for v0.90 metadata, and should not normally be used.
989 .BR \-N ", " \-\-name=
990 Specify the name of the array to assemble. This must be the name
991 that was specified when creating the array. It must either match
992 the name stored in the superblock exactly, or it must match
995 prefixed to the start of the given name.
998 .BR \-f ", " \-\-force
999 Assemble the array even if the metadata on some devices appears to be
1002 cannot find enough working devices to start the array, but can find
1003 some devices that are recorded as having failed, then it will mark
1004 those devices as working so that the array can be started.
1005 An array which requires
1007 to be started may contain data corruption. Use it carefully.
1010 .BR \-R ", " \-\-run
1011 Attempt to start the array even if fewer drives were given than were
1012 present last time the array was active. Normally if not all the
1013 expected drives are found and
1015 is not used, then the array will be assembled but not started.
1018 an attempt will be made to start it anyway.
1022 This is the reverse of
1024 in that it inhibits the startup of array unless all expected drives
1025 are present. This is only needed with
1027 and can be used if the physical connections to devices are
1028 not as reliable as you would like.
1031 .BR \-a ", " "\-\-auto{=no,yes,md,mdp,part}"
1032 See this option under Create and Build options.
1035 .BR \-b ", " \-\-bitmap=
1036 Specify the bitmap file that was given when the array was created. If
1039 bitmap, there is no need to specify this when assembling the array.
1042 .BR \-\-backup\-file=
1045 was used while reshaping an array (e.g. changing number of devices or
1046 chunk size) and the system crashed during the critical section, then the same
1048 must be presented to
1050 to allow possibly corrupted data to be restored, and the reshape
1054 .BR \-\-invalid\-backup
1055 If the file needed for the above option is not available for any
1056 reason an empty file can be given together with this option to
1057 indicate that the backup file is invalid. In this case the data that
1058 was being rearranged at the time of the crash could be irrecoverably
1059 lost, but the rest of the array may still be recoverable. This option
1060 should only be used as a last resort if there is no way to recover the
1065 .BR \-U ", " \-\-update=
1066 Update the superblock on each device while assembling the array. The
1067 argument given to this flag can be one of
1084 option will adjust the superblock of an array what was created on a Sparc
1085 machine running a patched 2.2 Linux kernel. This kernel got the
1086 alignment of part of the superblock wrong. You can use the
1087 .B "\-\-examine \-\-sparc2.2"
1090 to see what effect this would have.
1094 option will update the
1095 .B "preferred minor"
1096 field on each superblock to match the minor number of the array being
1098 This can be useful if
1100 reports a different "Preferred Minor" to
1102 In some cases this update will be performed automatically
1103 by the kernel driver. In particular the update happens automatically
1104 at the first write to an array with redundancy (RAID level 1 or
1105 greater) on a 2.6 (or later) kernel.
1109 option will change the uuid of the array. If a UUID is given with the
1111 option that UUID will be used as a new UUID and will
1113 be used to help identify the devices in the array.
1116 is given, a random UUID is chosen.
1120 option will change the
1122 of the array as stored in the superblock. This is only supported for
1123 version-1 superblocks.
1127 option will change the
1129 as recorded in the superblock. For version-0 superblocks, this is the
1130 same as updating the UUID.
1131 For version-1 superblocks, this involves updating the name.
1135 option will cause the array to be marked
1137 meaning that any redundancy in the array (e.g. parity for RAID5,
1138 copies for RAID1) may be incorrect. This will cause the RAID system
1139 to perform a "resync" pass to make sure that all redundant information
1144 option allows arrays to be moved between machines with different
1146 When assembling such an array for the first time after a move, giving
1147 .B "\-\-update=byteorder"
1150 to expect superblocks to have their byteorder reversed, and will
1151 correct that order before assembling the array. This is only valid
1152 with original (Version 0.90) superblocks.
1156 option will correct the summaries in the superblock. That is the
1157 counts of total, working, active, failed, and spare devices.
1161 option will rarely be of use. It applies to version 1.1 and 1.2 metadata
1162 only (where the metadata is at the start of the device) and is only
1163 useful when the component device has changed size (typically become
1164 larger). The version 1 metadata records the amount of the device that
1165 can be used to store data, so if a device in a version 1.1 or 1.2
1166 array becomes larger, the metadata will still be visible, but the
1167 extra space will not. In this case it might be useful to assemble the
1169 .BR \-\-update=devicesize .
1172 to determine the maximum usable amount of space on each device and
1173 update the relevant field in the metadata.
1177 option can be used when an array has an internal bitmap which is
1178 corrupt in some way so that assembling the array normally fails. It
1179 will cause any internal bitmap to be ignored.
1183 option will reserve space in each device for a bad block list. This
1184 will be 4K in size and positioned near the end of any free space
1185 between the superblock and the data.
1189 option will cause any reservation of space for a bad block list to be
1190 removed. If the bad block list contains entries, this will fail, as
1191 removing the list could cause data corruption.
1194 .BR \-\-freeze\-reshape
1195 Option is intended to be used in start-up scripts during initrd boot phase.
1196 When array under reshape is assembled during initrd phase, this option
1197 stops reshape after reshape critical section is being restored. This happens
1198 before file system pivot operation and avoids loss of file system context.
1199 Losing file system context would cause reshape to be broken.
1201 Reshape can be continued later using the
1203 option for the grow command.
1205 .SH For Manage mode:
1208 .BR \-t ", " \-\-test
1209 Unless a more serious error occurred,
1211 will exit with a status of 2 if no changes were made to the array and
1212 0 if at least one change was made.
1213 This can be useful when an indirect specifier such as
1218 is used in requesting an operation on the array.
1220 will report failure if these specifiers didn't find any match.
1223 .BR \-a ", " \-\-add
1224 hot-add listed devices.
1225 If a device appears to have recently been part of the array
1226 (possibly it failed or was removed) the device is re\-added as described
1228 If that fails or the device was never part of the array, the device is
1229 added as a hot-spare.
1230 If the array is degraded, it will immediately start to rebuild data
1233 Note that this and the following options are only meaningful on array
1234 with redundancy. They don't apply to RAID0 or Linear.
1238 re\-add a device that was previous removed from an array.
1239 If the metadata on the device reports that it is a member of the
1240 array, and the slot that it used is still vacant, then the device will
1241 be added back to the array in the same position. This will normally
1242 cause the data for that device to be recovered. However based on the
1243 event count on the device, the recovery may only require sections that
1244 are flagged a write-intent bitmap to be recovered or may not require
1245 any recovery at all.
1247 When used on an array that has no metadata (i.e. it was built with
1249 it will be assumed that bitmap-based recovery is enough to make the
1250 device fully consistent with the array.
1252 When used with v1.x metadata,
1254 can be accompanied by
1255 .BR \-\-update=devicesize ,
1256 .BR \-\-update=bbl ", or"
1257 .BR \-\-update=no\-bbl .
1258 See the description of these option when used in Assemble mode for an
1259 explanation of their use.
1261 If the device name given is
1265 will try to find any device that looks like it should be
1266 part of the array but isn't and will try to re\-add all such devices.
1268 If the device name given is
1272 will find all devices in the array that are marked
1274 remove them and attempt to immediately re\-add them. This can be
1275 useful if you are certain that the reason for failure has been
1279 .BR \-r ", " \-\-remove
1280 remove listed devices. They must not be active. i.e. they should
1281 be failed or spare devices. As well as the name of a device file
1290 The first causes all failed device to be removed. The second causes
1291 any device which is no longer connected to the system (i.e an 'open'
1294 to be removed. This will only succeed for devices that are spares or
1295 have already been marked as failed.
1298 .BR \-f ", " \-\-fail
1299 Mark listed devices as faulty.
1300 As well as the name of a device file, the word
1302 can be given. This will cause any device that has been detached from
1303 the system to be marked as failed. It can then be removed.
1312 Mark listed devices as requiring replacement. As soon as a spare is
1313 available, it will be rebuilt and will replace the marked device.
1314 This is similar to marking a device as faulty, but the device remains
1315 in service during the recovery process to increase resilience against
1316 multiple failures. When the replacement process finishes, the
1317 replaced device will be marked as faulty.
1321 This can follow a list of
1323 devices. The devices listed after
1325 will be preferentially used to replace the devices listed after
1327 These device must already be spare devices in the array.
1330 .BR \-\-write\-mostly
1331 Subsequent devices that are added or re\-added will have the 'write-mostly'
1332 flag set. This is only valid for RAID1 and means that the 'md' driver
1333 will avoid reading from these devices if possible.
1336 Subsequent devices that are added or re\-added will have the 'write-mostly'
1340 Each of these options requires that the first device listed is the array
1341 to be acted upon, and the remainder are component devices to be added,
1342 removed, marked as faulty, etc. Several different operations can be
1343 specified for different devices, e.g.
1345 mdadm /dev/md0 \-\-add /dev/sda1 \-\-fail /dev/sdb1 \-\-remove /dev/sdb1
1347 Each operation applies to all devices listed until the next
1350 If an array is using a write-intent bitmap, then devices which have
1351 been removed can be re\-added in a way that avoids a full
1352 reconstruction but instead just updates the blocks that have changed
1353 since the device was removed. For arrays with persistent metadata
1354 (superblocks) this is done automatically. For arrays created with
1356 mdadm needs to be told that this device we removed recently with
1359 Devices can only be removed from an array if they are not in active
1360 use, i.e. that must be spares or failed devices. To remove an active
1361 device, it must first be marked as
1367 .BR \-Q ", " \-\-query
1368 Examine a device to see
1369 (1) if it is an md device and (2) if it is a component of an md
1371 Information about what is discovered is presented.
1374 .BR \-D ", " \-\-detail
1375 Print details of one or more md devices.
1378 .BR \-\-detail\-platform
1379 Print details of the platform's RAID capabilities (firmware / hardware
1380 topology) for a given metadata format. If used without argument, mdadm
1381 will scan all controllers looking for their capabilities. Otherwise, mdadm
1382 will only look at the controller specified by the argument in form of an
1383 absolute filepath or a link, e.g.
1384 .IR /sys/devices/pci0000:00/0000:00:1f.2 .
1387 .BR \-Y ", " \-\-export
1389 .B \-\-detail , \-\-detail-platform
1392 output will be formatted as
1394 pairs for easy import into the environment.
1397 .BR \-E ", " \-\-examine
1398 Print contents of the metadata stored on the named device(s).
1399 Note the contrast between
1404 applies to devices which are components of an array, while
1406 applies to a whole array which is currently active.
1409 If an array was created on a SPARC machine with a 2.2 Linux kernel
1410 patched with RAID support, the superblock will have been created
1411 incorrectly, or at least incompatibly with 2.4 and later kernels.
1416 will fix the superblock before displaying it. If this appears to do
1417 the right thing, then the array can be successfully assembled using
1418 .BR "\-\-assemble \-\-update=sparc2.2" .
1421 .BR \-X ", " \-\-examine\-bitmap
1422 Report information about a bitmap file.
1423 The argument is either an external bitmap file or an array component
1424 in case of an internal bitmap. Note that running this on an array
1427 does not report the bitmap for that array.
1430 .BR \-R ", " \-\-run
1431 start a partially assembled array. If
1433 did not find enough devices to fully start the array, it might leaving
1434 it partially assembled. If you wish, you can then use
1436 to start the array in degraded mode.
1439 .BR \-S ", " \-\-stop
1440 deactivate array, releasing all resources.
1443 .BR \-o ", " \-\-readonly
1444 mark array as readonly.
1447 .BR \-w ", " \-\-readwrite
1448 mark array as readwrite.
1451 .B \-\-zero\-superblock
1452 If the device contains a valid md superblock, the block is
1453 overwritten with zeros. With
1455 the block where the superblock would be is overwritten even if it
1456 doesn't appear to be valid.
1459 .B \-\-kill\-subarray=
1460 If the device is a container and the argument to \-\-kill\-subarray
1461 specifies an inactive subarray in the container, then the subarray is
1462 deleted. Deleting all subarrays will leave an 'empty-container' or
1463 spare superblock on the drives. See \-\-zero\-superblock for completely
1464 removing a superblock. Note that some formats depend on the subarray
1465 index for generating a UUID, this command will fail if it would change
1466 the UUID of an active subarray.
1469 .B \-\-update\-subarray=
1470 If the device is a container and the argument to \-\-update\-subarray
1471 specifies a subarray in the container, then attempt to update the given
1472 superblock field in the subarray. See below in
1477 .BR \-t ", " \-\-test
1482 is set to reflect the status of the device. See below in
1487 .BR \-W ", " \-\-wait
1488 For each md device given, wait for any resync, recovery, or reshape
1489 activity to finish before returning.
1491 will return with success if it actually waited for every device
1492 listed, otherwise it will return failure.
1496 For each md device given, or each device in /proc/mdstat if
1498 is given, arrange for the array to be marked clean as soon as possible.
1500 will return with success if the array uses external metadata and we
1501 successfully waited. For native arrays this returns immediately as the
1502 kernel handles dirty-clean transitions at shutdown. No action is taken
1503 if safe-mode handling is disabled.
1505 .SH For Incremental Assembly mode:
1507 .BR \-\-rebuild\-map ", " \-r
1508 Rebuild the map file
1512 uses to help track which arrays are currently being assembled.
1515 .BR \-\-run ", " \-R
1516 Run any array assembled as soon as a minimal number of devices are
1517 available, rather than waiting until all expected devices are present.
1520 .BR \-\-scan ", " \-s
1521 Only meaningful with
1525 file for arrays that are being incrementally assembled and will try to
1526 start any that are not already started. If any such array is listed
1529 as requiring an external bitmap, that bitmap will be attached first.
1532 .BR \-\-fail ", " \-f
1533 This allows the hot-plug system to remove devices that have fully disappeared
1534 from the kernel. It will first fail and then remove the device from any
1535 array it belongs to.
1536 The device name given should be a kernel device name such as "sda",
1542 Only used with \-\-fail. The 'path' given will be recorded so that if
1543 a new device appears at the same location it can be automatically
1544 added to the same array. This allows the failed device to be
1545 automatically replaced by a new device without metadata if it appears
1546 at specified path. This option is normally only set by a
1550 .SH For Monitor mode:
1552 .BR \-m ", " \-\-mail
1553 Give a mail address to send alerts to.
1556 .BR \-p ", " \-\-program ", " \-\-alert
1557 Give a program to be run whenever an event is detected.
1560 .BR \-y ", " \-\-syslog
1561 Cause all events to be reported through 'syslog'. The messages have
1562 facility of 'daemon' and varying priorities.
1565 .BR \-d ", " \-\-delay
1566 Give a delay in seconds.
1568 polls the md arrays and then waits this many seconds before polling
1569 again. The default is 60 seconds. Since 2.6.16, there is no need to
1570 reduce this as the kernel alerts
1572 immediately when there is any change.
1575 .BR \-r ", " \-\-increment
1576 Give a percentage increment.
1578 will generate RebuildNN events with the given percentage increment.
1581 .BR \-f ", " \-\-daemonise
1584 to run as a background daemon if it decides to monitor anything. This
1585 causes it to fork and run in the child, and to disconnect from the
1586 terminal. The process id of the child is written to stdout.
1589 which will only continue monitoring if a mail address or alert program
1590 is found in the config file.
1593 .BR \-i ", " \-\-pid\-file
1596 is running in daemon mode, write the pid of the daemon process to
1597 the specified file, instead of printing it on standard output.
1600 .BR \-1 ", " \-\-oneshot
1601 Check arrays only once. This will generate
1603 events and more significantly
1609 .B " mdadm \-\-monitor \-\-scan \-1"
1611 from a cron script will ensure regular notification of any degraded arrays.
1614 .BR \-t ", " \-\-test
1617 alert for every array found at startup. This alert gets mailed and
1618 passed to the alert program. This can be used for testing that alert
1619 message do get through successfully.
1623 This inhibits the functionality for moving spares between arrays.
1624 Only one monitoring process started with
1626 but without this flag is allowed, otherwise the two could interfere
1633 .B mdadm \-\-assemble
1634 .I md-device options-and-component-devices...
1637 .B mdadm \-\-assemble \-\-scan
1638 .I md-devices-and-options...
1641 .B mdadm \-\-assemble \-\-scan
1645 This usage assembles one or more RAID arrays from pre-existing components.
1646 For each array, mdadm needs to know the md device, the identity of the
1647 array, and a number of component-devices. These can be found in a number of ways.
1649 In the first usage example (without the
1651 the first device given is the md device.
1652 In the second usage example, all devices listed are treated as md
1653 devices and assembly is attempted.
1654 In the third (where no devices are listed) all md devices that are
1655 listed in the configuration file are assembled. If no arrays are
1656 described by the configuration file, then any arrays that
1657 can be found on unused devices will be assembled.
1659 If precisely one device is listed, but
1665 was given and identity information is extracted from the configuration file.
1667 The identity can be given with the
1673 option, will be taken from the md-device record in the config file, or
1674 will be taken from the super block of the first component-device
1675 listed on the command line.
1677 Devices can be given on the
1679 command line or in the config file. Only devices which have an md
1680 superblock which contains the right identity will be considered for
1683 The config file is only used if explicitly named with
1685 or requested with (a possibly implicit)
1690 .B /etc/mdadm/mdadm.conf
1695 is not given, then the config file will only be used to find the
1696 identity of md arrays.
1698 Normally the array will be started after it is assembled. However if
1700 is not given and not all expected drives were listed, then the array
1701 is not started (to guard against usage errors). To insist that the
1702 array be started in this case (as may work for RAID1, 4, 5, 6, or 10),
1711 does not create any entries in
1715 It does record information in
1719 to choose the correct name.
1723 detects that udev is not configured, it will create the devices in
1727 In Linux kernels prior to version 2.6.28 there were two distinctly
1728 different types of md devices that could be created: one that could be
1729 partitioned using standard partitioning tools and one that could not.
1730 Since 2.6.28 that distinction is no longer relevant as both type of
1731 devices can be partitioned.
1733 will normally create the type that originally could not be partitioned
1734 as it has a well defined major number (9).
1736 Prior to 2.6.28, it is important that mdadm chooses the correct type
1737 of array device to use. This can be controlled with the
1739 option. In particular, a value of "mdp" or "part" or "p" tells mdadm
1740 to use a partitionable device rather than the default.
1742 In the no-udev case, the value given to
1744 can be suffixed by a number. This tells
1746 to create that number of partition devices rather than the default of 4.
1750 can also be given in the configuration file as a word starting
1752 on the ARRAY line for the relevant array.
1759 and no devices are listed,
1761 will first attempt to assemble all the arrays listed in the config
1764 If no arrays are listed in the config (other than those marked
1766 it will look through the available devices for possible arrays and
1767 will try to assemble anything that it finds. Arrays which are tagged
1768 as belonging to the given homehost will be assembled and started
1769 normally. Arrays which do not obviously belong to this host are given
1770 names that are expected not to conflict with anything local, and are
1771 started "read-auto" so that nothing is written to any device until the
1772 array is written to. i.e. automatic resync etc is delayed.
1776 finds a consistent set of devices that look like they should comprise
1777 an array, and if the superblock is tagged as belonging to the given
1778 home host, it will automatically choose a device name and try to
1779 assemble the array. If the array uses version-0.90 metadata, then the
1781 number as recorded in the superblock is used to create a name in
1785 If the array uses version-1 metadata, then the
1787 from the superblock is used to similarly create a name in
1789 (the name will have any 'host' prefix stripped first).
1791 This behaviour can be modified by the
1795 configuration file. This line can indicate that specific metadata
1796 type should, or should not, be automatically assembled. If an array
1797 is found which is not listed in
1799 and has a metadata format that is denied by the
1801 line, then it will not be assembled.
1804 line can also request that all arrays identified as being for this
1805 homehost should be assembled regardless of their metadata type.
1808 for further details.
1810 Note: Auto assembly cannot be used for assembling and activating some
1811 arrays which are undergoing reshape. In particular as the
1813 cannot be given, any reshape which requires a backup-file to continue
1814 cannot be started by auto assembly. An array which is growing to more
1815 devices and has passed the critical section can be assembled using
1826 .BI \-\-raid\-devices= Z
1830 This usage is similar to
1832 The difference is that it creates an array without a superblock. With
1833 these arrays there is no difference between initially creating the array and
1834 subsequently assembling the array, except that hopefully there is useful
1835 data there in the second case.
1837 The level may raid0, linear, raid1, raid10, multipath, or faulty, or
1838 one of their synonyms. All devices must be listed and the array will
1839 be started once complete. It will often be appropriate to use
1840 .B \-\-assume\-clean
1841 with levels raid1 or raid10.
1852 .BI \-\-raid\-devices= Z
1856 This usage will initialise a new md array, associate some devices with
1857 it, and activate the array.
1859 The named device will normally not exist when
1860 .I "mdadm \-\-create"
1861 is run, but will be created by
1863 once the array becomes active.
1865 As devices are added, they are checked to see if they contain RAID
1866 superblocks or filesystems. They are also checked to see if the variance in
1867 device size exceeds 1%.
1869 If any discrepancy is found, the array will not automatically be run, though
1872 can override this caution.
1874 To create a "degraded" array in which some devices are missing, simply
1875 give the word "\fBmissing\fP"
1876 in place of a device name. This will cause
1878 to leave the corresponding slot in the array empty.
1879 For a RAID4 or RAID5 array at most one slot can be
1880 "\fBmissing\fP"; for a RAID6 array at most two slots.
1881 For a RAID1 array, only one real device needs to be given. All of the
1885 When creating a RAID5 array,
1887 will automatically create a degraded array with an extra spare drive.
1888 This is because building the spare into a degraded array is in general
1889 faster than resyncing the parity on a non-degraded, but not clean,
1890 array. This feature can be overridden with the
1894 When creating an array with version-1 metadata a name for the array is
1896 If this is not given with the
1900 will choose a name based on the last component of the name of the
1901 device being created. So if
1903 is being created, then the name
1908 is being created, then the name
1912 When creating a partition based array, using
1914 with version-1.x metadata, the partition type should be set to
1916 (non fs-data). This type selection allows for greater precision since
1917 using any other [RAID auto-detect (0xFD) or a GNU/Linux partition (0x83)],
1918 might create problems in the event of array recovery through a live cdrom.
1920 A new array will normally get a randomly assigned 128bit UUID which is
1921 very likely to be unique. If you have a specific need, you can choose
1922 a UUID for the array by giving the
1924 option. Be warned that creating two arrays with the same UUID is a
1925 recipe for disaster. Also, using
1927 when creating a v0.90 array will silently override any
1932 .\"option is given, it is not necessary to list any component-devices in this command.
1933 .\"They can be added later, before a
1937 .\"is given, the apparent size of the smallest drive given is used.
1939 If the metadata type supports it (currently only 1.x metadata), space
1940 will be allocated to store a bad block list. This allows a modest
1941 number of bad blocks to be recorded, allowing the drive to remain in
1942 service while only partially functional.
1944 When creating an array within a
1947 can be given either the list of devices to use, or simply the name of
1948 the container. The former case gives control over which devices in
1949 the container will be used for the array. The latter case allows
1951 to automatically choose which devices to use based on how much spare
1954 The General Management options that are valid with
1959 insist on running the array even if some devices look like they might
1964 start the array readonly \(em not supported yet.
1971 .I options... devices...
1974 This usage will allow individual devices in an array to be failed,
1975 removed or added. It is possible to perform multiple operations with
1976 on command. For example:
1978 .B " mdadm /dev/md0 \-f /dev/hda1 \-r /dev/hda1 \-a /dev/hda1"
1984 and will then remove it from the array and finally add it back
1985 in as a spare. However only one md array can be affected by a single
1988 When a device is added to an active array, mdadm checks to see if it
1989 has metadata on it which suggests that it was recently a member of the
1990 array. If it does, it tries to "re\-add" the device. If there have
1991 been no changes since the device was removed, or if the array has a
1992 write-intent bitmap which has recorded whatever changes there were,
1993 then the device will immediately become a full member of the array and
1994 those differences recorded in the bitmap will be resolved.
2004 MISC mode includes a number of distinct operations that
2005 operate on distinct devices. The operations are:
2008 The device is examined to see if it is
2009 (1) an active md array, or
2010 (2) a component of an md array.
2011 The information discovered is reported.
2015 The device should be an active md device.
2017 will display a detailed description of the array.
2021 will cause the output to be less detailed and the format to be
2022 suitable for inclusion in
2026 will normally be 0 unless
2028 failed to get useful information about the device(s); however, if the
2030 option is given, then the exit status will be:
2034 The array is functioning normally.
2037 The array has at least one failed device.
2040 The array has multiple failed devices such that it is unusable.
2043 There was an error while trying to get information about the device.
2047 .B \-\-detail\-platform
2048 Print detail of the platform's RAID capabilities (firmware / hardware
2049 topology). If the metadata is specified with
2053 then the return status will be:
2057 metadata successfully enumerated its platform components on this system
2060 metadata is platform independent
2063 metadata failed to find its platform components on this system
2067 .B \-\-update\-subarray=
2068 If the device is a container and the argument to \-\-update\-subarray
2069 specifies a subarray in the container, then attempt to update the given
2070 superblock field in the subarray. Similar to updating an array in
2071 "assemble" mode, the field to update is selected by
2075 option. Currently only
2081 option updates the subarray name in the metadata, it may not affect the
2082 device node name or the device node symlink until the subarray is
2083 re\-assembled. If updating
2085 would change the UUID of an active subarray this operation is blocked,
2086 and the command will end in an error.
2090 The device should be a component of an md array.
2092 will read the md superblock of the device and display the contents.
2097 is given, then multiple devices that are components of the one array
2098 are grouped together and reported in a single entry suitable
2104 without listing any devices will cause all devices listed in the
2105 config file to be examined.
2109 The devices should be active md arrays which will be deactivated, as
2110 long as they are not currently in use.
2114 This will fully activate a partially assembled md array.
2118 This will mark an active array as read-only, providing that it is
2119 not currently being used.
2125 array back to being read/write.
2129 For all operations except
2132 will cause the operation to be applied to all arrays listed in
2137 causes all devices listed in the config file to be examined.
2140 .BR \-b ", " \-\-brief
2141 Be less verbose. This is used with
2149 gives an intermediate level of verbosity.
2155 .B mdadm \-\-monitor
2156 .I options... devices...
2161 to periodically poll a number of md arrays and to report on any events
2164 will never exit once it decides that there are arrays to be checked,
2165 so it should normally be run in the background.
2167 As well as reporting events,
2169 may move a spare drive from one array to another if they are in the
2174 and if the destination array has a failed drive but no spares.
2176 If any devices are listed on the command line,
2178 will only monitor those devices. Otherwise all arrays listed in the
2179 configuration file will be monitored. Further, if
2181 is given, then any other md devices that appear in
2183 will also be monitored.
2185 The result of monitoring the arrays is the generation of events.
2186 These events are passed to a separate program (if specified) and may
2187 be mailed to a given E-mail address.
2189 When passing events to a program, the program is run once for each event,
2190 and is given 2 or 3 command-line arguments: the first is the
2191 name of the event (see below), the second is the name of the
2192 md device which is affected, and the third is the name of a related
2193 device if relevant (such as a component device that has failed).
2197 is given, then a program or an E-mail address must be specified on the
2198 command line or in the config file. If neither are available, then
2200 will not monitor anything.
2204 will continue monitoring as long as something was found to monitor. If
2205 no program or email is given, then each event is reported to
2208 The different events are:
2212 .B DeviceDisappeared
2213 An md array which previously was configured appears to no longer be
2214 configured. (syslog priority: Critical)
2218 was told to monitor an array which is RAID0 or Linear, then it will
2220 .B DeviceDisappeared
2221 with the extra information
2223 This is because RAID0 and Linear do not support the device-failed,
2224 hot-spare and resync operations which are monitored.
2228 An md array started reconstruction. (syslog priority: Warning)
2234 is a two-digit number (ie. 05, 48). This indicates that rebuild
2235 has passed that many percent of the total. The events are generated
2236 with fixed increment since 0. Increment size may be specified with
2237 a commandline option (default is 20). (syslog priority: Warning)
2241 An md array that was rebuilding, isn't any more, either because it
2242 finished normally or was aborted. (syslog priority: Warning)
2246 An active component device of an array has been marked as
2247 faulty. (syslog priority: Critical)
2251 A spare component device which was being rebuilt to replace a faulty
2252 device has failed. (syslog priority: Critical)
2256 A spare component device which was being rebuilt to replace a faulty
2257 device has been successfully rebuilt and has been made active.
2258 (syslog priority: Info)
2262 A new md array has been detected in the
2264 file. (syslog priority: Info)
2268 A newly noticed array appears to be degraded. This message is not
2271 notices a drive failure which causes degradation, but only when
2273 notices that an array is degraded when it first sees the array.
2274 (syslog priority: Critical)
2278 A spare drive has been moved from one array in a
2282 to another to allow a failed drive to be replaced.
2283 (syslog priority: Info)
2289 has been told, via the config file, that an array should have a certain
2290 number of spare devices, and
2292 detects that it has fewer than this number when it first sees the
2293 array, it will report a
2296 (syslog priority: Warning)
2300 An array was found at startup, and the
2303 (syslog priority: Info)
2313 cause Email to be sent. All events cause the program to be run.
2314 The program is run with two or three arguments: the event
2315 name, the array device and possibly a second device.
2317 Each event has an associated array device (e.g.
2319 and possibly a second device. For
2324 the second device is the relevant component device.
2327 the second device is the array that the spare was moved from.
2331 to move spares from one array to another, the different arrays need to
2332 be labeled with the same
2334 or the spares must be allowed to migrate through matching POLICY domains
2335 in the configuration file. The
2337 name can be any string; it is only necessary that different spare
2338 groups use different names.
2342 detects that an array in a spare group has fewer active
2343 devices than necessary for the complete array, and has no spare
2344 devices, it will look for another array in the same spare group that
2345 has a full complement of working drive and a spare. It will then
2346 attempt to remove the spare from the second drive and add it to the
2348 If the removal succeeds but the adding fails, then it is added back to
2351 If the spare group for a degraded array is not defined,
2353 will look at the rules of spare migration specified by POLICY lines in
2355 and then follow similar steps as above if a matching spare is found.
2358 The GROW mode is used for changing the size or shape of an active
2360 For this to work, the kernel must support the necessary change.
2361 Various types of growth are being added during 2.6 development.
2363 Currently the supported changes include
2365 change the "size" attribute for RAID1, RAID4, RAID5 and RAID6.
2367 increase or decrease the "raid\-devices" attribute of RAID0, RAID1, RAID4,
2370 change the chunk-size and layout of RAID0, RAID4, RAID5 and RAID6.
2372 convert between RAID1 and RAID5, between RAID5 and RAID6, between
2373 RAID0, RAID4, and RAID5, and between RAID0 and RAID10 (in the near-2 mode).
2375 add a write-intent bitmap to any array which supports these bitmaps, or
2376 remove a write-intent bitmap from such an array.
2379 Using GROW on containers is currently supported only for Intel's IMSM
2380 container format. The number of devices in a container can be
2381 increased - which affects all arrays in the container - or an array
2382 in a container can be converted between levels where those levels are
2383 supported by the container, and the conversion is on of those listed
2384 above. Resizing arrays in an IMSM container with
2386 is not yet supported.
2388 Grow functionality (e.g. expand a number of raid devices) for Intel's
2389 IMSM container format has an experimental status. It is guarded by the
2390 .B MDADM_EXPERIMENTAL
2391 environment variable which must be set to '1' for a GROW command to
2393 This is for the following reasons:
2396 Intel's native IMSM check-pointing is not fully tested yet.
2397 This can causes IMSM incompatibility during the grow process: an array
2398 which is growing cannot roam between Microsoft Windows(R) and Linux
2402 Interrupting a grow operation is not recommended, because it
2403 has not been fully tested for Intel's IMSM container format yet.
2406 Note: Intel's native checkpointing doesn't use
2408 option and it is transparent for assembly feature.
2411 Normally when an array is built the "size" is taken from the smallest
2412 of the drives. If all the small drives in an arrays are, one at a
2413 time, removed and replaced with larger drives, then you could have an
2414 array of large drives with only a small amount used. In this
2415 situation, changing the "size" with "GROW" mode will allow the extra
2416 space to start being used. If the size is increased in this way, a
2417 "resync" process will start to make sure the new parts of the array
2420 Note that when an array changes size, any filesystem that may be
2421 stored in the array will not automatically grow or shrink to use or
2422 vacate the space. The
2423 filesystem will need to be explicitly told to use the extra space
2424 after growing, or to reduce its size
2426 to shrinking the array.
2428 Also the size of an array cannot be changed while it has an active
2429 bitmap. If an array has a bitmap, it must be removed before the size
2430 can be changed. Once the change is complete a new bitmap can be created.
2432 .SS RAID\-DEVICES CHANGES
2434 A RAID1 array can work with any number of devices from 1 upwards
2435 (though 1 is not very useful). There may be times which you want to
2436 increase or decrease the number of active devices. Note that this is
2437 different to hot-add or hot-remove which changes the number of
2440 When reducing the number of devices in a RAID1 array, the slots which
2441 are to be removed from the array must already be vacant. That is, the
2442 devices which were in those slots must be failed and removed.
2444 When the number of devices is increased, any hot spares that are
2445 present will be activated immediately.
2447 Changing the number of active devices in a RAID5 or RAID6 is much more
2448 effort. Every block in the array will need to be read and written
2449 back to a new location. From 2.6.17, the Linux Kernel is able to
2450 increase the number of devices in a RAID5 safely, including restarting
2451 an interrupted "reshape". From 2.6.31, the Linux Kernel is able to
2452 increase or decrease the number of devices in a RAID5 or RAID6.
2454 From 2.6.35, the Linux Kernel is able to convert a RAID0 in to a RAID4
2457 uses this functionality and the ability to add
2458 devices to a RAID4 to allow devices to be added to a RAID0. When
2459 requested to do this,
2461 will convert the RAID0 to a RAID4, add the necessary disks and make
2462 the reshape happen, and then convert the RAID4 back to RAID0.
2464 When decreasing the number of devices, the size of the array will also
2465 decrease. If there was data in the array, it could get destroyed and
2466 this is not reversible, so you should firstly shrink the filesystem on
2467 the array to fit within the new size. To help prevent accidents,
2469 requires that the size of the array be decreased first with
2470 .BR "mdadm --grow --array-size" .
2471 This is a reversible change which simply makes the end of the array
2472 inaccessible. The integrity of any data can then be checked before
2473 the non-reversible reduction in the number of devices is request.
2475 When relocating the first few stripes on a RAID5 or RAID6, it is not
2476 possible to keep the data on disk completely consistent and
2477 crash-proof. To provide the required safety, mdadm disables writes to
2478 the array while this "critical section" is reshaped, and takes a
2479 backup of the data that is in that section. For grows, this backup may be
2480 stored in any spare devices that the array has, however it can also be
2481 stored in a separate file specified with the
2483 option, and is required to be specified for shrinks, RAID level
2484 changes and layout changes. If this option is used, and the system
2485 does crash during the critical period, the same file must be passed to
2487 to restore the backup and reassemble the array. When shrinking rather
2488 than growing the array, the reshape is done from the end towards the
2489 beginning, so the "critical section" is at the end of the reshape.
2493 Changing the RAID level of any array happens instantaneously. However
2494 in the RAID5 to RAID6 case this requires a non-standard layout of the
2495 RAID6 data, and in the RAID6 to RAID5 case that non-standard layout is
2496 required before the change can be accomplished. So while the level
2497 change is instant, the accompanying layout change can take quite a
2500 is required. If the array is not simultaneously being grown or
2501 shrunk, so that the array size will remain the same - for example,
2502 reshaping a 3-drive RAID5 into a 4-drive RAID6 - the backup file will
2503 be used not just for a "cricital section" but throughout the reshape
2504 operation, as described below under LAYOUT CHANGES.
2506 .SS CHUNK-SIZE AND LAYOUT CHANGES
2508 Changing the chunk-size of layout without also changing the number of
2509 devices as the same time will involve re-writing all blocks in-place.
2510 To ensure against data loss in the case of a crash, a
2512 must be provided for these changes. Small sections of the array will
2513 be copied to the backup file while they are being rearranged. This
2514 means that all the data is copied twice, once to the backup and once
2515 to the new layout on the array, so this type of reshape will go very
2518 If the reshape is interrupted for any reason, this backup file must be
2520 .B "mdadm --assemble"
2521 so the array can be reassembled. Consequently the file cannot be
2522 stored on the device being reshaped.
2527 A write-intent bitmap can be added to, or removed from, an active
2528 array. Either internal bitmaps, or bitmaps stored in a separate file,
2529 can be added. Note that if you add a bitmap stored in a file which is
2530 in a filesystem that is on the RAID array being affected, the system
2531 will deadlock. The bitmap must be on a separate filesystem.
2533 .SH INCREMENTAL MODE
2537 .B mdadm \-\-incremental
2543 .B mdadm \-\-incremental \-\-fail
2547 .B mdadm \-\-incremental \-\-rebuild\-map
2550 .B mdadm \-\-incremental \-\-run \-\-scan
2553 This mode is designed to be used in conjunction with a device
2554 discovery system. As devices are found in a system, they can be
2556 .B "mdadm \-\-incremental"
2557 to be conditionally added to an appropriate array.
2559 Conversely, it can also be used with the
2561 flag to do just the opposite and find whatever array a particular device
2562 is part of and remove the device from that array.
2564 If the device passed is a
2566 device created by a previous call to
2568 then rather than trying to add that device to an array, all the arrays
2569 described by the metadata of the container will be started.
2572 performs a number of tests to determine if the device is part of an
2573 array, and which array it should be part of. If an appropriate array
2574 is found, or can be created,
2576 adds the device to the array and conditionally starts the array.
2580 will normally only add devices to an array which were previously working
2581 (active or spare) parts of that array. The support for automatic
2582 inclusion of a new drive as a spare in some array requires
2583 a configuration through POLICY in config file.
2587 makes are as follow:
2589 Is the device permitted by
2591 That is, is it listed in a
2593 line in that file. If
2595 is absent then the default it to allow any device. Similar if
2597 contains the special word
2599 then any device is allowed. Otherwise the device name given to
2601 must match one of the names or patterns in a
2606 Does the device have a valid md superblock? If a specific metadata
2607 version is requested with
2611 then only that style of metadata is accepted, otherwise
2613 finds any known version of metadata. If no
2615 metadata is found, the device may be still added to an array
2616 as a spare if POLICY allows.
2620 Does the metadata match an expected array?
2621 The metadata can match in two ways. Either there is an array listed
2624 which identifies the array (either by UUID, by name, by device list,
2625 or by minor-number), or the array was created with a
2631 or on the command line.
2634 is not able to positively identify the array as belonging to the
2635 current host, the device will be rejected.
2640 keeps a list of arrays that it has partially assembled in
2642 If no array exists which matches
2643 the metadata on the new device,
2645 must choose a device name and unit number. It does this based on any
2648 or any name information stored in the metadata. If this name
2649 suggests a unit number, that number will be used, otherwise a free
2650 unit number will be chosen. Normally
2652 will prefer to create a partitionable array, however if the
2656 suggests that a non-partitionable array is preferred, that will be
2659 If the array is not found in the config file and its metadata does not
2660 identify it as belonging to the "homehost", then
2662 will choose a name for the array which is certain not to conflict with
2663 any array which does belong to this host. It does this be adding an
2664 underscore and a small number to the name preferred by the metadata.
2666 Once an appropriate array is found or created and the device is added,
2668 must decide if the array is ready to be started. It will
2669 normally compare the number of available (non-spare) devices to the
2670 number of devices that the metadata suggests need to be active. If
2671 there are at least that many, the array will be started. This means
2672 that if any devices are missing the array will not be restarted.
2678 in which case the array will be run as soon as there are enough
2679 devices present for the data to be accessible. For a RAID1, that
2680 means one device will start the array. For a clean RAID5, the array
2681 will be started as soon as all but one drive is present.
2683 Note that neither of these approaches is really ideal. If it can
2684 be known that all device discovery has completed, then
2688 can be run which will try to start all arrays that are being
2689 incrementally assembled. They are started in "read-auto" mode in
2690 which they are read-only until the first write request. This means
2691 that no metadata updates are made and no attempt at resync or recovery
2692 happens. Further devices that are found before the first write can
2693 still be added safely.
2696 This section describes environment variables that affect how mdadm
2701 Setting this value to 1 will prevent mdadm from automatically launching
2702 mdmon. This variable is intended primarily for debugging mdadm/mdmon.
2708 does not create any device nodes in /dev, but leaves that task to
2712 appears not to be configured, or if this environment variable is set
2715 will create and devices that are needed.
2719 .B " mdadm \-\-query /dev/name-of-device"
2721 This will find out if a given device is a RAID array, or is part of
2722 one, and will provide brief information about the device.
2724 .B " mdadm \-\-assemble \-\-scan"
2726 This will assemble and start all arrays listed in the standard config
2727 file. This command will typically go in a system startup file.
2729 .B " mdadm \-\-stop \-\-scan"
2731 This will shut down all arrays that can be shut down (i.e. are not
2732 currently in use). This will typically go in a system shutdown script.
2734 .B " mdadm \-\-follow \-\-scan \-\-delay=120"
2736 If (and only if) there is an Email address or program given in the
2737 standard config file, then
2738 monitor the status of all arrays listed in that file by
2739 polling them ever 2 minutes.
2741 .B " mdadm \-\-create /dev/md0 \-\-level=1 \-\-raid\-devices=2 /dev/hd[ac]1"
2743 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
2746 .B " echo 'DEVICE /dev/hd*[0\-9] /dev/sd*[0\-9]' > mdadm.conf"
2748 .B " mdadm \-\-detail \-\-scan >> mdadm.conf"
2750 This will create a prototype config file that describes currently
2751 active arrays that are known to be made from partitions of IDE or SCSI drives.
2752 This file should be reviewed before being used as it may
2753 contain unwanted detail.
2755 .B " echo 'DEVICE /dev/hd[a\-z] /dev/sd*[a\-z]' > mdadm.conf"
2757 .B " mdadm \-\-examine \-\-scan \-\-config=mdadm.conf >> mdadm.conf"
2759 This will find arrays which could be assembled from existing IDE and
2760 SCSI whole drives (not partitions), and store the information in the
2761 format of a config file.
2762 This file is very likely to contain unwanted detail, particularly
2765 entries. It should be reviewed and edited before being used as an
2768 .B " mdadm \-\-examine \-\-brief \-\-scan \-\-config=partitions"
2770 .B " mdadm \-Ebsc partitions"
2772 Create a list of devices by reading
2773 .BR /proc/partitions ,
2774 scan these for RAID superblocks, and printout a brief listing of all
2777 .B " mdadm \-Ac partitions \-m 0 /dev/md0"
2779 Scan all partitions and devices listed in
2780 .BR /proc/partitions
2783 out of all such devices with a RAID superblock with a minor number of 0.
2785 .B " mdadm \-\-monitor \-\-scan \-\-daemonise > /run/mdadm/mon.pid"
2787 If config file contains a mail address or alert program, run mdadm in
2788 the background in monitor mode monitoring all md devices. Also write
2789 pid of mdadm daemon to
2790 .BR /run/mdadm/mon.pid .
2792 .B " mdadm \-Iq /dev/somedevice"
2794 Try to incorporate newly discovered device into some array as
2797 .B " mdadm \-\-incremental \-\-rebuild\-map \-\-run \-\-scan"
2799 Rebuild the array map from any current arrays, and then start any that
2802 .B " mdadm /dev/md4 --fail detached --remove detached"
2804 Any devices which are components of /dev/md4 will be marked as faulty
2805 and then remove from the array.
2807 .B " mdadm --grow /dev/md4 --level=6 --backup-file=/root/backup-md4"
2811 which is currently a RAID5 array will be converted to RAID6. There
2812 should normally already be a spare drive attached to the array as a
2813 RAID6 needs one more drive than a matching RAID5.
2815 .B " mdadm --create /dev/md/ddf --metadata=ddf --raid-disks 6 /dev/sd[a-f]"
2817 Create a DDF array over 6 devices.
2819 .B " mdadm --create /dev/md/home -n3 -l5 -z 30000000 /dev/md/ddf"
2821 Create a RAID5 array over any 3 devices in the given DDF set. Use
2822 only 30 gigabytes of each device.
2824 .B " mdadm -A /dev/md/ddf1 /dev/sd[a-f]"
2826 Assemble a pre-exist ddf array.
2828 .B " mdadm -I /dev/md/ddf1"
2830 Assemble all arrays contained in the ddf array, assigning names as
2833 .B " mdadm \-\-create \-\-help"
2835 Provide help about the Create mode.
2837 .B " mdadm \-\-config \-\-help"
2839 Provide help about the format of the config file.
2841 .B " mdadm \-\-help"
2843 Provide general help.
2853 lists all active md devices with information about them.
2855 uses this to find arrays when
2857 is given in Misc mode, and to monitor array reconstruction
2862 The config file lists which devices may be scanned to see if
2863 they contain MD super block, and gives identifying information
2864 (e.g. UUID) about known MD arrays. See
2871 mode is used, this file gets a list of arrays currently being created.
2876 understand two sorts of names for array devices.
2878 The first is the so-called 'standard' format name, which matches the
2879 names used by the kernel and which appear in
2882 The second sort can be freely chosen, but must reside in
2884 When giving a device name to
2886 to create or assemble an array, either full path name such as
2890 can be given, or just the suffix of the second sort of name, such as
2896 chooses device names during auto-assembly or incremental assembly, it
2897 will sometimes add a small sequence number to the end of the name to
2898 avoid conflicted between multiple arrays that have the same name. If
2900 can reasonably determine that the array really is meant for this host,
2901 either by a hostname in the metadata, or by the presence of the array
2904 then it will leave off the suffix if possible.
2905 Also if the homehost is specified as
2908 will only use a suffix if a different array of the same name already
2909 exists or is listed in the config file.
2911 The standard names for non-partitioned arrays (the only sort of md
2912 array available in 2.4 and earlier) are of the form
2916 where NN is a number.
2917 The standard names for partitionable arrays (as available from 2.6
2918 onwards) are of the form
2922 Partition numbers should be indicated by added "pMM" to these, thus "/dev/md/d1p2".
2924 From kernel version, 2.6.28 the "non-partitioned array" can actually
2925 be partitioned. So the "md_dNN" names are no longer needed, and
2926 partitions such as "/dev/mdNNpXX" are possible.
2930 was previously known as
2934 is completely separate from the
2936 package, and does not use the
2938 configuration file at all.
2941 For further information on mdadm usage, MD and the various levels of
2944 .B http://raid.wiki.kernel.org/
2946 (based upon Jakob \(/Ostergaard's Software\-RAID.HOWTO)
2948 .\"for new releases of the RAID driver check out:
2951 .\".UR ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2952 .\"ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2957 .\".UR http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2958 .\"http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2961 The latest version of
2963 should always be available from
2965 .B http://www.kernel.org/pub/linux/utils/raid/mdadm/
projects / thirdparty / mdadm.git / blob