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
221 then the MANAGE mode is assumed.
222 Anything other than these will cause the
226 .SH Options that are not mode-specific are:
229 .BR \-h ", " \-\-help
230 Display general help message or, after one of the above options, a
231 mode-specific help message.
235 Display more detailed help about command line parsing and some commonly
239 .BR \-V ", " \-\-version
240 Print version information for mdadm.
243 .BR \-v ", " \-\-verbose
244 Be more verbose about what is happening. This can be used twice to be
246 The extra verbosity currently only affects
247 .B \-\-detail \-\-scan
249 .BR "\-\-examine \-\-scan" .
252 .BR \-q ", " \-\-quiet
253 Avoid printing purely informative messages. With this,
255 will be silent unless there is something really important to report.
259 Set first character of argv[0] to @ to indicate mdadm was launched
260 from initrd/initramfs and should not be shutdown by systemd as part of
261 the regular shutdown process. This option is normally only used by
262 the system's initscripts. Please see here for more details on how
263 systemd handled argv[0]:
265 .B http://www.freedesktop.org/wiki/Software/systemd/RootStorageDaemons
270 .BR \-f ", " \-\-force
271 Be more forceful about certain operations. See the various modes for
272 the exact meaning of this option in different contexts.
275 .BR \-c ", " \-\-config=
276 Specify the config file. Default is to use
277 .BR /etc/mdadm.conf ,
278 or if that is missing then
279 .BR /etc/mdadm/mdadm.conf .
280 If the config file given is
282 then nothing will be read, but
284 will act as though the config file contained exactly
285 .B "DEVICE partitions containers"
288 to find a list of devices to scan, and
290 to find a list of containers to examine.
293 is given for the config file, then
295 will act as though the config file were empty.
298 .BR \-s ", " \-\-scan
301 for missing information.
302 In general, this option gives
304 permission to get any missing information (like component devices,
305 array devices, array identities, and alert destination) from the
306 configuration file (see previous option);
307 one exception is MISC mode when using
313 says to get a list of array devices from
317 .BR \-e ", " \-\-metadata=
318 Declare the style of RAID metadata (superblock) to be used. The
319 default is {DEFAULT_METADATA} for
321 and to guess for other operations.
322 The default can be overridden by setting the
331 .ie '{DEFAULT_METADATA}'0.90'
332 .IP "0, 0.90, default"
335 Use the original 0.90 format superblock. This format limits arrays to
336 28 component devices and limits component devices of levels 1 and
337 greater to 2 terabytes. It is also possible for there to be confusion
338 about whether the superblock applies to a whole device or just the
339 last partition, if that partition starts on a 64K boundary.
340 .ie '{DEFAULT_METADATA}'0.90'
341 .IP "1, 1.0, 1.1, 1.2"
343 .IP "1, 1.0, 1.1, 1.2 default"
344 Use the new version-1 format superblock. This has fewer restrictions.
345 It can easily be moved between hosts with different endian-ness, and a
346 recovery operation can be checkpointed and restarted. The different
347 sub-versions store the superblock at different locations on the
348 device, either at the end (for 1.0), at the start (for 1.1) or 4K from
349 the start (for 1.2). "1" is equivalent to "1.2" (the commonly
350 preferred 1.x format).
351 'if '{DEFAULT_METADATA}'1.2' "default" is equivalent to "1.2".
353 Use the "Industry Standard" DDF (Disk Data Format) format defined by
355 When creating a DDF array a
357 will be created, and normal arrays can be created in that container.
359 Use the Intel(R) Matrix Storage Manager metadata format. This creates a
361 which is managed in a similar manner to DDF, and is supported by an
362 option-rom on some platforms:
364 .B http://www.intel.com/design/chipsets/matrixstorage_sb.htm
370 This will override any
372 setting in the config file and provides the identity of the host which
373 should be considered the home for any arrays.
375 When creating an array, the
377 will be recorded in the metadata. For version-1 superblocks, it will
378 be prefixed to the array name. For version-0.90 superblocks, part of
379 the SHA1 hash of the hostname will be stored in the later half of the
382 When reporting information about an array, any array which is tagged
383 for the given homehost will be reported as such.
385 When using Auto-Assemble, only arrays tagged for the given homehost
386 will be allowed to use 'local' names (i.e. not ending in '_' followed
387 by a digit string). See below under
388 .BR "Auto Assembly" .
394 needs to print the name for a device it normally finds the name in
396 which refers to the device and is shortest. When a path component is
400 will prefer a longer name if it contains that component. For example
401 .B \-\-prefer=by-uuid
402 will prefer a name in a subdirectory of
407 This functionality is currently only provided by
412 .SH For create, build, or grow:
415 .BR \-n ", " \-\-raid\-devices=
416 Specify the number of active devices in the array. This, plus the
417 number of spare devices (see below) must equal the number of
419 (including "\fBmissing\fP" devices)
420 that are listed on the command line for
422 Setting a value of 1 is probably
423 a mistake and so requires that
425 be specified first. A value of 1 will then be allowed for linear,
426 multipath, RAID0 and RAID1. It is never allowed for RAID4, RAID5 or RAID6.
428 This number can only be changed using
430 for RAID1, RAID4, RAID5 and RAID6 arrays, and only on kernels which provide
431 the necessary support.
434 .BR \-x ", " \-\-spare\-devices=
435 Specify the number of spare (eXtra) devices in the initial array.
436 Spares can also be added
437 and removed later. The number of component devices listed
438 on the command line must equal the number of RAID devices plus the
439 number of spare devices.
442 .BR \-z ", " \-\-size=
443 Amount (in Kibibytes) of space to use from each drive in RAID levels 1/4/5/6.
444 This must be a multiple of the chunk size, and must leave about 128Kb
445 of space at the end of the drive for the RAID superblock.
446 If this is not specified
447 (as it normally is not) the smallest drive (or partition) sets the
448 size, though if there is a variance among the drives of greater than 1%, a warning is
451 A suffix of 'M' or 'G' can be given to indicate Megabytes or
452 Gigabytes respectively.
454 Sometimes a replacement drive can be a little smaller than the
455 original drives though this should be minimised by IDEMA standards.
456 Such a replacement drive will be rejected by
458 To guard against this it can be useful to set the initial size
459 slightly smaller than the smaller device with the aim that it will
460 still be larger than any replacement.
462 This value can be set with
464 for RAID level 1/4/5/6 though
466 based arrays such as those with IMSM metadata may not be able to
468 If the array was created with a size smaller than the currently
469 active drives, the extra space can be accessed using
471 The size can be given as
473 which means to choose the largest size that fits on all current drives.
475 Before reducing the size of the array (with
476 .BR "\-\-grow \-\-size=" )
477 you should make sure that space isn't needed. If the device holds a
478 filesystem, you would need to resize the filesystem to use less space.
480 After reducing the array size you should check that the data stored in
481 the device is still available. If the device holds a filesystem, then
482 an 'fsck' of the filesystem is a minimum requirement. If there are
483 problems the array can be made bigger again with no loss with another
484 .B "\-\-grow \-\-size="
487 This value cannot be used when creating a
489 such as with DDF and IMSM metadata, though it perfectly valid when
490 creating an array inside a container.
493 .BR \-Z ", " \-\-array\-size=
494 This is only meaningful with
496 and its effect is not persistent: when the array is stopped and
497 restarted the default array size will be restored.
499 Setting the array-size causes the array to appear smaller to programs
500 that access the data. This is particularly needed before reshaping an
501 array so that it will be smaller. As the reshape is not reversible,
502 but setting the size with
504 is, it is required that the array size is reduced as appropriate
505 before the number of devices in the array is reduced.
507 Before reducing the size of the array you should make sure that space
508 isn't needed. If the device holds a filesystem, you would need to
509 resize the filesystem to use less space.
511 After reducing the array size you should check that the data stored in
512 the device is still available. If the device holds a filesystem, then
513 an 'fsck' of the filesystem is a minimum requirement. If there are
514 problems the array can be made bigger again with no loss with another
515 .B "\-\-grow \-\-array\-size="
518 A suffix of 'M' or 'G' can be given to indicate Megabytes or
519 Gigabytes respectively.
522 restores the apparent size of the array to be whatever the real
523 amount of available space is.
526 .BR \-c ", " \-\-chunk=
527 Specify chunk size of kibibytes. The default when creating an
528 array is 512KB. To ensure compatibility with earlier versions, the
529 default when Building and array with no persistent metadata is 64KB.
530 This is only meaningful for RAID0, RAID4, RAID5, RAID6, and RAID10.
532 RAID4, RAID5, RAID6, and RAID10 require the chunk size to be a power
533 of 2. In any case it must be a multiple of 4KB.
535 A suffix of 'M' or 'G' can be given to indicate Megabytes or
536 Gigabytes respectively.
540 Specify rounding factor for a Linear array. The size of each
541 component will be rounded down to a multiple of this size.
542 This is a synonym for
544 but highlights the different meaning for Linear as compared to other
545 RAID levels. The default is 64K if a kernel earlier than 2.6.16 is in
546 use, and is 0K (i.e. no rounding) in later kernels.
549 .BR \-l ", " \-\-level=
550 Set RAID level. When used with
552 options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
553 raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty, container.
554 Obviously some of these are synonymous.
558 metadata type is requested, only the
560 level is permitted, and it does not need to be explicitly given.
564 only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid.
568 to change the RAID level in some cases. See LEVEL CHANGES below.
571 .BR \-p ", " \-\-layout=
572 This option configures the fine details of data layout for RAID5, RAID6,
573 and RAID10 arrays, and controls the failure modes for
576 The layout of the RAID5 parity block can be one of
577 .BR left\-asymmetric ,
578 .BR left\-symmetric ,
579 .BR right\-asymmetric ,
580 .BR right\-symmetric ,
581 .BR la ", " ra ", " ls ", " rs .
583 .BR left\-symmetric .
585 It is also possible to cause RAID5 to use a RAID4-like layout by
591 Finally for RAID5 there are DDF\-compatible layouts,
592 .BR ddf\-zero\-restart ,
593 .BR ddf\-N\-restart ,
595 .BR ddf\-N\-continue .
597 These same layouts are available for RAID6. There are also 4 layouts
598 that will provide an intermediate stage for converting between RAID5
599 and RAID6. These provide a layout which is identical to the
600 corresponding RAID5 layout on the first N\-1 devices, and has the 'Q'
601 syndrome (the second 'parity' block used by RAID6) on the last device.
603 .BR left\-symmetric\-6 ,
604 .BR right\-symmetric\-6 ,
605 .BR left\-asymmetric\-6 ,
606 .BR right\-asymmetric\-6 ,
608 .BR parity\-first\-6 .
610 When setting the failure mode for level
613 .BR write\-transient ", " wt ,
614 .BR read\-transient ", " rt ,
615 .BR write\-persistent ", " wp ,
616 .BR read\-persistent ", " rp ,
618 .BR read\-fixable ", " rf ,
619 .BR clear ", " flush ", " none .
621 Each failure mode can be followed by a number, which is used as a period
622 between fault generation. Without a number, the fault is generated
623 once on the first relevant request. With a number, the fault will be
624 generated after that many requests, and will continue to be generated
625 every time the period elapses.
627 Multiple failure modes can be current simultaneously by using the
629 option to set subsequent failure modes.
631 "clear" or "none" will remove any pending or periodic failure modes,
632 and "flush" will clear any persistent faults.
634 Finally, the layout options for RAID10 are one of 'n', 'o' or 'f' followed
635 by a small number. The default is 'n2'. The supported options are:
638 signals 'near' copies. Multiple copies of one data block are at
639 similar offsets in different devices.
642 signals 'offset' copies. Rather than the chunks being duplicated
643 within a stripe, whole stripes are duplicated but are rotated by one
644 device so duplicate blocks are on different devices. Thus subsequent
645 copies of a block are in the next drive, and are one chunk further
650 (multiple copies have very different offsets).
651 See md(4) for more detail about 'near', 'offset', and 'far'.
653 The number is the number of copies of each datablock. 2 is normal, 3
654 can be useful. This number can be at most equal to the number of
655 devices in the array. It does not need to divide evenly into that
656 number (e.g. it is perfectly legal to have an 'n2' layout for an array
657 with an odd number of devices).
659 When an array is converted between RAID5 and RAID6 an intermediate
660 RAID6 layout is used in which the second parity block (Q) is always on
661 the last device. To convert a RAID5 to RAID6 and leave it in this new
662 layout (which does not require re-striping) use
663 .BR \-\-layout=preserve .
664 This will try to avoid any restriping.
666 The converse of this is
667 .B \-\-layout=normalise
668 which will change a non-standard RAID6 layout into a more standard
675 (thus explaining the p of
679 .BR \-b ", " \-\-bitmap=
680 Specify a file to store a write-intent bitmap in. The file should not
683 is also given. The same file should be provided
684 when assembling the array. If the word
686 is given, then the bitmap is stored with the metadata on the array,
687 and so is replicated on all devices. If the word
691 mode, then any bitmap that is present is removed.
693 To help catch typing errors, the filename must contain at least one
694 slash ('/') if it is a real file (not 'internal' or 'none').
696 Note: external bitmaps are only known to work on ext2 and ext3.
697 Storing bitmap files on other filesystems may result in serious problems.
700 .BR \-\-bitmap\-chunk=
701 Set the chunksize of the bitmap. Each bit corresponds to that many
702 Kilobytes of storage.
703 When using a file based bitmap, the default is to use the smallest
704 size that is at-least 4 and requires no more than 2^21 chunks.
707 bitmap, the chunksize defaults to 64Meg, or larger if necessary to
708 fit the bitmap into the available space.
710 A suffix of 'M' or 'G' can be given to indicate Megabytes or
711 Gigabytes respectively.
714 .BR \-W ", " \-\-write\-mostly
715 subsequent devices listed in a
720 command will be flagged as 'write-mostly'. This is valid for RAID1
721 only and means that the 'md' driver will avoid reading from these
722 devices if at all possible. This can be useful if mirroring over a
726 .BR \-\-write\-behind=
727 Specify that write-behind mode should be enabled (valid for RAID1
728 only). If an argument is specified, it will set the maximum number
729 of outstanding writes allowed. The default value is 256.
730 A write-intent bitmap is required in order to use write-behind
731 mode, and write-behind is only attempted on drives marked as
735 .BR \-\-assume\-clean
738 that the array pre-existed and is known to be clean. It can be useful
739 when trying to recover from a major failure as you can be sure that no
740 data will be affected unless you actually write to the array. It can
741 also be used when creating a RAID1 or RAID10 if you want to avoid the
742 initial resync, however this practice \(em while normally safe \(em is not
743 recommended. Use this only if you really know what you are doing.
745 When the devices that will be part of a new array were filled
746 with zeros before creation the operator knows the array is
747 actually clean. If that is the case, such as after running
748 badblocks, this argument can be used to tell mdadm the
749 facts the operator knows.
751 When an array is resized to a larger size with
752 .B "\-\-grow \-\-size="
753 the new space is normally resynced in that same way that the whole
754 array is resynced at creation. From Linux version 3.0,
756 can be used with that command to avoid the automatic resync.
759 .BR \-\-backup\-file=
762 is used to increase the number of raid-devices in a RAID5 or RAID6 if
763 there are no spare devices available, or to shrink, change RAID level
764 or layout. See the GROW MODE section below on RAID\-DEVICES CHANGES.
765 The file must be stored on a separate device, not on the RAID array
770 This option is complementary to the
771 .B \-\-freeze-reshape
772 option for assembly. It is needed when
774 operation is interrupted and it is not restarted automatically due to
775 .B \-\-freeze-reshape
776 usage during array assembly. This option is used together with
780 ) command and device for a pending reshape to be continued.
781 All parameters required for reshape continuation will be read from array metadata.
785 .BR \-\-backup\-file=
786 option to be set, continuation option will require to have exactly the same
787 backup file given as well.
789 Any other parameter passed together with
791 option will be ignored.
794 .BR \-N ", " \-\-name=
797 for the array. This is currently only effective when creating an
798 array with a version-1 superblock, or an array in a DDF container.
799 The name is a simple textual string that can be used to identify array
800 components when assembling. If name is needed but not specified, it
801 is taken from the basename of the device that is being created.
813 run the array, even if some of the components
814 appear to be active in another array or filesystem. Normally
816 will ask for confirmation before including such components in an
817 array. This option causes that question to be suppressed.
820 .BR \-f ", " \-\-force
823 accept the geometry and layout specified without question. Normally
825 will not allow creation of an array with only one device, and will try
826 to create a RAID5 array with one missing drive (as this makes the
827 initial resync work faster). With
830 will not try to be so clever.
833 .BR \-a ", " "\-\-auto{=yes,md,mdp,part,p}{NN}"
834 Instruct mdadm how to create the device file if needed, possibly allocating
835 an unused minor number. "md" causes a non-partitionable array
836 to be used (though since Linux 2.6.28, these array devices are in fact
837 partitionable). "mdp", "part" or "p" causes a partitionable array (2.6 and
838 later) to be used. "yes" requires the named md device to have
839 a 'standard' format, and the type and minor number will be determined
840 from this. With mdadm 3.0, device creation is normally left up to
842 so this option is unlikely to be needed.
843 See DEVICE NAMES below.
845 The argument can also come immediately after
850 is not given on the command line or in the config file, then
856 is also given, then any
858 entries in the config file will override the
860 instruction given on the command line.
862 For partitionable arrays,
864 will create the device file for the whole array and for the first 4
865 partitions. A different number of partitions can be specified at the
866 end of this option (e.g.
868 If the device name ends with a digit, the partition names add a 'p',
870 .IR /dev/md/home1p3 .
871 If there is no trailing digit, then the partition names just have a
873 .IR /dev/md/scratch3 .
875 If the md device name is in a 'standard' format as described in DEVICE
876 NAMES, then it will be created, if necessary, with the appropriate
877 device number based on that name. If the device name is not in one of these
878 formats, then a unused device number will be allocated. The device
879 number will be considered unused if there is no active array for that
880 number, and there is no entry in /dev for that number and with a
881 non-standard name. Names that are not in 'standard' format are only
882 allowed in "/dev/md/".
884 This is meaningful with
890 .BR \-a ", " "\-\-add"
891 This option can be used in Grow mode in two cases.
893 If the target array is a Linear array, then
895 can be used to add one or more devices to the array. They
896 are simply catenated on to the end of the array. Once added, the
897 devices cannot be removed.
901 option is being used to increase the number of devices in an array,
904 can be used to add some extra devices to be included in the array.
905 In most cases this is not needed as the extra devices can be added as
906 spares first, and then the number of raid-disks can be changed.
907 However for RAID0, it is not possible to add spares. So to increase
908 the number of devices in a RAID0, it is necessary to set the new
909 number of devices, and to add the new devices, in the same command.
914 .BR \-u ", " \-\-uuid=
915 uuid of array to assemble. Devices which don't have this uuid are
919 .BR \-m ", " \-\-super\-minor=
920 Minor number of device that array was created for. Devices which
921 don't have this minor number are excluded. If you create an array as
922 /dev/md1, then all superblocks will contain the minor number 1, even if
923 the array is later assembled as /dev/md2.
925 Giving the literal word "dev" for
929 to use the minor number of the md device that is being assembled.
932 .B \-\-super\-minor=dev
933 will look for super blocks with a minor number of 0.
936 is only relevant for v0.90 metadata, and should not normally be used.
942 .BR \-N ", " \-\-name=
943 Specify the name of the array to assemble. This must be the name
944 that was specified when creating the array. It must either match
945 the name stored in the superblock exactly, or it must match
948 prefixed to the start of the given name.
951 .BR \-f ", " \-\-force
952 Assemble the array even if the metadata on some devices appears to be
955 cannot find enough working devices to start the array, but can find
956 some devices that are recorded as having failed, then it will mark
957 those devices as working so that the array can be started.
958 An array which requires
960 to be started may contain data corruption. Use it carefully.
964 Attempt to start the array even if fewer drives were given than were
965 present last time the array was active. Normally if not all the
966 expected drives are found and
968 is not used, then the array will be assembled but not started.
971 an attempt will be made to start it anyway.
975 This is the reverse of
977 in that it inhibits the startup of array unless all expected drives
978 are present. This is only needed with
980 and can be used if the physical connections to devices are
981 not as reliable as you would like.
984 .BR \-a ", " "\-\-auto{=no,yes,md,mdp,part}"
985 See this option under Create and Build options.
988 .BR \-b ", " \-\-bitmap=
989 Specify the bitmap file that was given when the array was created. If
992 bitmap, there is no need to specify this when assembling the array.
995 .BR \-\-backup\-file=
998 was used while reshaping an array (e.g. changing number of devices or
999 chunk size) and the system crashed during the critical section, then the same
1001 must be presented to
1003 to allow possibly corrupted data to be restored, and the reshape
1007 .BR \-\-invalid\-backup
1008 If the file needed for the above option is not available for any
1009 reason an empty file can be given together with this option to
1010 indicate that the backup file is invalid. In this case the data that
1011 was being rearranged at the time of the crash could be irrecoverably
1012 lost, but the rest of the array may still be recoverable. This option
1013 should only be used as a last resort if there is no way to recover the
1018 .BR \-U ", " \-\-update=
1019 Update the superblock on each device while assembling the array. The
1020 argument given to this flag can be one of
1035 option will adjust the superblock of an array what was created on a Sparc
1036 machine running a patched 2.2 Linux kernel. This kernel got the
1037 alignment of part of the superblock wrong. You can use the
1038 .B "\-\-examine \-\-sparc2.2"
1041 to see what effect this would have.
1045 option will update the
1046 .B "preferred minor"
1047 field on each superblock to match the minor number of the array being
1049 This can be useful if
1051 reports a different "Preferred Minor" to
1053 In some cases this update will be performed automatically
1054 by the kernel driver. In particular the update happens automatically
1055 at the first write to an array with redundancy (RAID level 1 or
1056 greater) on a 2.6 (or later) kernel.
1060 option will change the uuid of the array. If a UUID is given with the
1062 option that UUID will be used as a new UUID and will
1064 be used to help identify the devices in the array.
1067 is given, a random UUID is chosen.
1071 option will change the
1073 of the array as stored in the superblock. This is only supported for
1074 version-1 superblocks.
1078 option will change the
1080 as recorded in the superblock. For version-0 superblocks, this is the
1081 same as updating the UUID.
1082 For version-1 superblocks, this involves updating the name.
1086 option will cause the array to be marked
1088 meaning that any redundancy in the array (e.g. parity for RAID5,
1089 copies for RAID1) may be incorrect. This will cause the RAID system
1090 to perform a "resync" pass to make sure that all redundant information
1095 option allows arrays to be moved between machines with different
1097 When assembling such an array for the first time after a move, giving
1098 .B "\-\-update=byteorder"
1101 to expect superblocks to have their byteorder reversed, and will
1102 correct that order before assembling the array. This is only valid
1103 with original (Version 0.90) superblocks.
1107 option will correct the summaries in the superblock. That is the
1108 counts of total, working, active, failed, and spare devices.
1112 option will rarely be of use. It applies to version 1.1 and 1.2 metadata
1113 only (where the metadata is at the start of the device) and is only
1114 useful when the component device has changed size (typically become
1115 larger). The version 1 metadata records the amount of the device that
1116 can be used to store data, so if a device in a version 1.1 or 1.2
1117 array becomes larger, the metadata will still be visible, but the
1118 extra space will not. In this case it might be useful to assemble the
1120 .BR \-\-update=devicesize .
1123 to determine the maximum usable amount of space on each device and
1124 update the relevant field in the metadata.
1128 option can be used when an array has an internal bitmap which is
1129 corrupt in some way so that assembling the array normally fails. It
1130 will cause any internal bitmap to be ignored.
1133 .BR \-\-freeze\-reshape
1134 Option is intended to be used in start-up scripts during initrd boot phase.
1135 When array under reshape is assembled during initrd phase, this option
1136 stops reshape after reshape critical section is being restored. This happens
1137 before file system pivot operation and avoids loss of file system context.
1138 Losing file system context would cause reshape to be broken.
1140 Reshape can be continued later using the
1142 option for the grow command.
1144 .SH For Manage mode:
1147 .BR \-t ", " \-\-test
1148 Unless a more serious error occurred,
1150 will exit with a status of 2 if no changes were made to the array and
1151 0 if at least one change was made.
1152 This can be useful when an indirect specifier such as
1157 is used in requesting an operation on the array.
1159 will report failure if these specifiers didn't find any match.
1162 .BR \-a ", " \-\-add
1163 hot-add listed devices.
1164 If a device appears to have recently been part of the array
1165 (possibly it failed or was removed) the device is re\-added as described
1167 If that fails or the device was never part of the array, the device is
1168 added as a hot-spare.
1169 If the array is degraded, it will immediately start to rebuild data
1172 Note that this and the following options are only meaningful on array
1173 with redundancy. They don't apply to RAID0 or Linear.
1177 re\-add a device that was previous removed from an array.
1178 If the metadata on the device reports that it is a member of the
1179 array, and the slot that it used is still vacant, then the device will
1180 be added back to the array in the same position. This will normally
1181 cause the data for that device to be recovered. However based on the
1182 event count on the device, the recovery may only require sections that
1183 are flagged a write-intent bitmap to be recovered or may not require
1184 any recovery at all.
1186 When used on an array that has no metadata (i.e. it was built with
1188 it will be assumed that bitmap-based recovery is enough to make the
1189 device fully consistent with the array.
1193 can be accompanied by
1194 .BR \-\-update=devicesize .
1195 See the description of this option when used in Assemble mode for an
1196 explanation of its use.
1198 If the device name given is
1200 then mdadm will try to find any device that looks like it should be
1201 part of the array but isn't and will try to re\-add all such devices.
1204 .BR \-r ", " \-\-remove
1205 remove listed devices. They must not be active. i.e. they should
1206 be failed or spare devices. As well as the name of a device file
1215 The first causes all failed device to be removed. The second causes
1216 any device which is no longer connected to the system (i.e an 'open'
1219 to be removed. This will only succeed for devices that are spares or
1220 have already been marked as failed.
1223 .BR \-f ", " \-\-fail
1224 mark listed devices as faulty.
1225 As well as the name of a device file, the word
1227 can be given. This will cause any device that has been detached from
1228 the system to be marked as failed. It can then be removed.
1236 .BR \-\-write\-mostly
1237 Subsequent devices that are added or re\-added will have the 'write-mostly'
1238 flag set. This is only valid for RAID1 and means that the 'md' driver
1239 will avoid reading from these devices if possible.
1242 Subsequent devices that are added or re\-added will have the 'write-mostly'
1246 Each of these options requires that the first device listed is the array
1247 to be acted upon, and the remainder are component devices to be added,
1248 removed, marked as faulty, etc. Several different operations can be
1249 specified for different devices, e.g.
1251 mdadm /dev/md0 \-\-add /dev/sda1 \-\-fail /dev/sdb1 \-\-remove /dev/sdb1
1253 Each operation applies to all devices listed until the next
1256 If an array is using a write-intent bitmap, then devices which have
1257 been removed can be re\-added in a way that avoids a full
1258 reconstruction but instead just updates the blocks that have changed
1259 since the device was removed. For arrays with persistent metadata
1260 (superblocks) this is done automatically. For arrays created with
1262 mdadm needs to be told that this device we removed recently with
1265 Devices can only be removed from an array if they are not in active
1266 use, i.e. that must be spares or failed devices. To remove an active
1267 device, it must first be marked as
1273 .BR \-Q ", " \-\-query
1274 Examine a device to see
1275 (1) if it is an md device and (2) if it is a component of an md
1277 Information about what is discovered is presented.
1280 .BR \-D ", " \-\-detail
1281 Print details of one or more md devices.
1284 .BR \-\-detail\-platform
1285 Print details of the platform's RAID capabilities (firmware / hardware
1286 topology) for a given metadata format.
1289 .BR \-Y ", " \-\-export
1294 output will be formatted as
1296 pairs for easy import into the environment.
1299 .BR \-E ", " \-\-examine
1300 Print contents of the metadata stored on the named device(s).
1301 Note the contrast between
1306 applies to devices which are components of an array, while
1308 applies to a whole array which is currently active.
1311 If an array was created on a SPARC machine with a 2.2 Linux kernel
1312 patched with RAID support, the superblock will have been created
1313 incorrectly, or at least incompatibly with 2.4 and later kernels.
1318 will fix the superblock before displaying it. If this appears to do
1319 the right thing, then the array can be successfully assembled using
1320 .BR "\-\-assemble \-\-update=sparc2.2" .
1323 .BR \-X ", " \-\-examine\-bitmap
1324 Report information about a bitmap file.
1325 The argument is either an external bitmap file or an array component
1326 in case of an internal bitmap. Note that running this on an array
1329 does not report the bitmap for that array.
1332 .BR \-R ", " \-\-run
1333 start a partially assembled array. If
1335 did not find enough devices to fully start the array, it might leaving
1336 it partially assembled. If you wish, you can then use
1338 to start the array in degraded mode.
1341 .BR \-S ", " \-\-stop
1342 deactivate array, releasing all resources.
1345 .BR \-o ", " \-\-readonly
1346 mark array as readonly.
1349 .BR \-w ", " \-\-readwrite
1350 mark array as readwrite.
1353 .B \-\-zero\-superblock
1354 If the device contains a valid md superblock, the block is
1355 overwritten with zeros. With
1357 the block where the superblock would be is overwritten even if it
1358 doesn't appear to be valid.
1361 .B \-\-kill\-subarray=
1362 If the device is a container and the argument to \-\-kill\-subarray
1363 specifies an inactive subarray in the container, then the subarray is
1364 deleted. Deleting all subarrays will leave an 'empty-container' or
1365 spare superblock on the drives. See \-\-zero\-superblock for completely
1366 removing a superblock. Note that some formats depend on the subarray
1367 index for generating a UUID, this command will fail if it would change
1368 the UUID of an active subarray.
1371 .B \-\-update\-subarray=
1372 If the device is a container and the argument to \-\-update\-subarray
1373 specifies a subarray in the container, then attempt to update the given
1374 superblock field in the subarray. See below in
1379 .BR \-t ", " \-\-test
1384 is set to reflect the status of the device. See below in
1389 .BR \-W ", " \-\-wait
1390 For each md device given, wait for any resync, recovery, or reshape
1391 activity to finish before returning.
1393 will return with success if it actually waited for every device
1394 listed, otherwise it will return failure.
1398 For each md device given, or each device in /proc/mdstat if
1400 is given, arrange for the array to be marked clean as soon as possible.
1402 will return with success if the array uses external metadata and we
1403 successfully waited. For native arrays this returns immediately as the
1404 kernel handles dirty-clean transitions at shutdown. No action is taken
1405 if safe-mode handling is disabled.
1407 .SH For Incremental Assembly mode:
1409 .BR \-\-rebuild\-map ", " \-r
1410 Rebuild the map file
1414 uses to help track which arrays are currently being assembled.
1417 .BR \-\-run ", " \-R
1418 Run any array assembled as soon as a minimal number of devices are
1419 available, rather than waiting until all expected devices are present.
1422 .BR \-\-scan ", " \-s
1423 Only meaningful with
1427 file for arrays that are being incrementally assembled and will try to
1428 start any that are not already started. If any such array is listed
1431 as requiring an external bitmap, that bitmap will be attached first.
1434 .BR \-\-fail ", " \-f
1435 This allows the hot-plug system to remove devices that have fully disappeared
1436 from the kernel. It will first fail and then remove the device from any
1437 array it belongs to.
1438 The device name given should be a kernel device name such as "sda",
1444 Only used with \-\-fail. The 'path' given will be recorded so that if
1445 a new device appears at the same location it can be automatically
1446 added to the same array. This allows the failed device to be
1447 automatically replaced by a new device without metadata if it appears
1448 at specified path. This option is normally only set by a
1452 .SH For Monitor mode:
1454 .BR \-m ", " \-\-mail
1455 Give a mail address to send alerts to.
1458 .BR \-p ", " \-\-program ", " \-\-alert
1459 Give a program to be run whenever an event is detected.
1462 .BR \-y ", " \-\-syslog
1463 Cause all events to be reported through 'syslog'. The messages have
1464 facility of 'daemon' and varying priorities.
1467 .BR \-d ", " \-\-delay
1468 Give a delay in seconds.
1470 polls the md arrays and then waits this many seconds before polling
1471 again. The default is 60 seconds. Since 2.6.16, there is no need to
1472 reduce this as the kernel alerts
1474 immediately when there is any change.
1477 .BR \-r ", " \-\-increment
1478 Give a percentage increment.
1480 will generate RebuildNN events with the given percentage increment.
1483 .BR \-f ", " \-\-daemonise
1486 to run as a background daemon if it decides to monitor anything. This
1487 causes it to fork and run in the child, and to disconnect from the
1488 terminal. The process id of the child is written to stdout.
1491 which will only continue monitoring if a mail address or alert program
1492 is found in the config file.
1495 .BR \-i ", " \-\-pid\-file
1498 is running in daemon mode, write the pid of the daemon process to
1499 the specified file, instead of printing it on standard output.
1502 .BR \-1 ", " \-\-oneshot
1503 Check arrays only once. This will generate
1505 events and more significantly
1511 .B " mdadm \-\-monitor \-\-scan \-1"
1513 from a cron script will ensure regular notification of any degraded arrays.
1516 .BR \-t ", " \-\-test
1519 alert for every array found at startup. This alert gets mailed and
1520 passed to the alert program. This can be used for testing that alert
1521 message do get through successfully.
1525 This inhibits the functionality for moving spares between arrays.
1526 Only one monitoring process started with
1528 but without this flag is allowed, otherwise the two could interfere
1535 .B mdadm \-\-assemble
1536 .I md-device options-and-component-devices...
1539 .B mdadm \-\-assemble \-\-scan
1540 .I md-devices-and-options...
1543 .B mdadm \-\-assemble \-\-scan
1547 This usage assembles one or more RAID arrays from pre-existing components.
1548 For each array, mdadm needs to know the md device, the identity of the
1549 array, and a number of component-devices. These can be found in a number of ways.
1551 In the first usage example (without the
1553 the first device given is the md device.
1554 In the second usage example, all devices listed are treated as md
1555 devices and assembly is attempted.
1556 In the third (where no devices are listed) all md devices that are
1557 listed in the configuration file are assembled. If no arrays are
1558 described by the configuration file, then any arrays that
1559 can be found on unused devices will be assembled.
1561 If precisely one device is listed, but
1567 was given and identity information is extracted from the configuration file.
1569 The identity can be given with the
1575 option, will be taken from the md-device record in the config file, or
1576 will be taken from the super block of the first component-device
1577 listed on the command line.
1579 Devices can be given on the
1581 command line or in the config file. Only devices which have an md
1582 superblock which contains the right identity will be considered for
1585 The config file is only used if explicitly named with
1587 or requested with (a possibly implicit)
1592 .B /etc/mdadm/mdadm.conf
1597 is not given, then the config file will only be used to find the
1598 identity of md arrays.
1600 Normally the array will be started after it is assembled. However if
1602 is not given and not all expected drives were listed, then the array
1603 is not started (to guard against usage errors). To insist that the
1604 array be started in this case (as may work for RAID1, 4, 5, 6, or 10),
1613 does not create any entries in
1617 It does record information in
1621 to choose the correct name.
1625 detects that udev is not configured, it will create the devices in
1629 In Linux kernels prior to version 2.6.28 there were two distinctly
1630 different types of md devices that could be created: one that could be
1631 partitioned using standard partitioning tools and one that could not.
1632 Since 2.6.28 that distinction is no longer relevant as both type of
1633 devices can be partitioned.
1635 will normally create the type that originally could not be partitioned
1636 as it has a well defined major number (9).
1638 Prior to 2.6.28, it is important that mdadm chooses the correct type
1639 of array device to use. This can be controlled with the
1641 option. In particular, a value of "mdp" or "part" or "p" tells mdadm
1642 to use a partitionable device rather than the default.
1644 In the no-udev case, the value given to
1646 can be suffixed by a number. This tells
1648 to create that number of partition devices rather than the default of 4.
1652 can also be given in the configuration file as a word starting
1654 on the ARRAY line for the relevant array.
1661 and no devices are listed,
1663 will first attempt to assemble all the arrays listed in the config
1666 If no arrays are listed in the config (other than those marked
1668 it will look through the available devices for possible arrays and
1669 will try to assemble anything that it finds. Arrays which are tagged
1670 as belonging to the given homehost will be assembled and started
1671 normally. Arrays which do not obviously belong to this host are given
1672 names that are expected not to conflict with anything local, and are
1673 started "read-auto" so that nothing is written to any device until the
1674 array is written to. i.e. automatic resync etc is delayed.
1678 finds a consistent set of devices that look like they should comprise
1679 an array, and if the superblock is tagged as belonging to the given
1680 home host, it will automatically choose a device name and try to
1681 assemble the array. If the array uses version-0.90 metadata, then the
1683 number as recorded in the superblock is used to create a name in
1687 If the array uses version-1 metadata, then the
1689 from the superblock is used to similarly create a name in
1691 (the name will have any 'host' prefix stripped first).
1693 This behaviour can be modified by the
1697 configuration file. This line can indicate that specific metadata
1698 type should, or should not, be automatically assembled. If an array
1699 is found which is not listed in
1701 and has a metadata format that is denied by the
1703 line, then it will not be assembled.
1706 line can also request that all arrays identified as being for this
1707 homehost should be assembled regardless of their metadata type.
1710 for further details.
1712 Note: Auto assembly cannot be used for assembling and activating some
1713 arrays which are undergoing reshape. In particular as the
1715 cannot be given, any reshape which requires a backup-file to continue
1716 cannot be started by auto assembly. An array which is growing to more
1717 devices and has passed the critical section can be assembled using
1728 .BI \-\-raid\-devices= Z
1732 This usage is similar to
1734 The difference is that it creates an array without a superblock. With
1735 these arrays there is no difference between initially creating the array and
1736 subsequently assembling the array, except that hopefully there is useful
1737 data there in the second case.
1739 The level may raid0, linear, raid1, raid10, multipath, or faulty, or
1740 one of their synonyms. All devices must be listed and the array will
1741 be started once complete. It will often be appropriate to use
1742 .B \-\-assume\-clean
1743 with levels raid1 or raid10.
1754 .BI \-\-raid\-devices= Z
1758 This usage will initialise a new md array, associate some devices with
1759 it, and activate the array.
1761 The named device will normally not exist when
1762 .I "mdadm \-\-create"
1763 is run, but will be created by
1765 once the array becomes active.
1767 As devices are added, they are checked to see if they contain RAID
1768 superblocks or filesystems. They are also checked to see if the variance in
1769 device size exceeds 1%.
1771 If any discrepancy is found, the array will not automatically be run, though
1774 can override this caution.
1776 To create a "degraded" array in which some devices are missing, simply
1777 give the word "\fBmissing\fP"
1778 in place of a device name. This will cause
1780 to leave the corresponding slot in the array empty.
1781 For a RAID4 or RAID5 array at most one slot can be
1782 "\fBmissing\fP"; for a RAID6 array at most two slots.
1783 For a RAID1 array, only one real device needs to be given. All of the
1787 When creating a RAID5 array,
1789 will automatically create a degraded array with an extra spare drive.
1790 This is because building the spare into a degraded array is in general
1791 faster than resyncing the parity on a non-degraded, but not clean,
1792 array. This feature can be overridden with the
1796 When creating an array with version-1 metadata a name for the array is
1798 If this is not given with the
1802 will choose a name based on the last component of the name of the
1803 device being created. So if
1805 is being created, then the name
1810 is being created, then the name
1814 When creating a partition based array, using
1816 with version-1.x metadata, the partition type should be set to
1818 (non fs-data). This type selection allows for greater precision since
1819 using any other [RAID auto-detect (0xFD) or a GNU/Linux partition (0x83)],
1820 might create problems in the event of array recovery through a live cdrom.
1822 A new array will normally get a randomly assigned 128bit UUID which is
1823 very likely to be unique. If you have a specific need, you can choose
1824 a UUID for the array by giving the
1826 option. Be warned that creating two arrays with the same UUID is a
1827 recipe for disaster. Also, using
1829 when creating a v0.90 array will silently override any
1834 .\"option is given, it is not necessary to list any component-devices in this command.
1835 .\"They can be added later, before a
1839 .\"is given, the apparent size of the smallest drive given is used.
1841 When creating an array within a
1844 can be given either the list of devices to use, or simply the name of
1845 the container. The former case gives control over which devices in
1846 the container will be used for the array. The latter case allows
1848 to automatically choose which devices to use based on how much spare
1851 The General Management options that are valid with
1856 insist on running the array even if some devices look like they might
1861 start the array readonly \(em not supported yet.
1868 .I options... devices...
1871 This usage will allow individual devices in an array to be failed,
1872 removed or added. It is possible to perform multiple operations with
1873 on command. For example:
1875 .B " mdadm /dev/md0 \-f /dev/hda1 \-r /dev/hda1 \-a /dev/hda1"
1881 and will then remove it from the array and finally add it back
1882 in as a spare. However only one md array can be affected by a single
1885 When a device is added to an active array, mdadm checks to see if it
1886 has metadata on it which suggests that it was recently a member of the
1887 array. If it does, it tries to "re\-add" the device. If there have
1888 been no changes since the device was removed, or if the array has a
1889 write-intent bitmap which has recorded whatever changes there were,
1890 then the device will immediately become a full member of the array and
1891 those differences recorded in the bitmap will be resolved.
1901 MISC mode includes a number of distinct operations that
1902 operate on distinct devices. The operations are:
1905 The device is examined to see if it is
1906 (1) an active md array, or
1907 (2) a component of an md array.
1908 The information discovered is reported.
1912 The device should be an active md device.
1914 will display a detailed description of the array.
1918 will cause the output to be less detailed and the format to be
1919 suitable for inclusion in
1923 will normally be 0 unless
1925 failed to get useful information about the device(s); however, if the
1927 option is given, then the exit status will be:
1931 The array is functioning normally.
1934 The array has at least one failed device.
1937 The array has multiple failed devices such that it is unusable.
1940 There was an error while trying to get information about the device.
1944 .B \-\-detail\-platform
1945 Print detail of the platform's RAID capabilities (firmware / hardware
1946 topology). If the metadata is specified with
1950 then the return status will be:
1954 metadata successfully enumerated its platform components on this system
1957 metadata is platform independent
1960 metadata failed to find its platform components on this system
1964 .B \-\-update\-subarray=
1965 If the device is a container and the argument to \-\-update\-subarray
1966 specifies a subarray in the container, then attempt to update the given
1967 superblock field in the subarray. Similar to updating an array in
1968 "assemble" mode, the field to update is selected by
1972 option. Currently only
1978 option updates the subarray name in the metadata, it may not affect the
1979 device node name or the device node symlink until the subarray is
1980 re\-assembled. If updating
1982 would change the UUID of an active subarray this operation is blocked,
1983 and the command will end in an error.
1987 The device should be a component of an md array.
1989 will read the md superblock of the device and display the contents.
1994 is given, then multiple devices that are components of the one array
1995 are grouped together and reported in a single entry suitable
2001 without listing any devices will cause all devices listed in the
2002 config file to be examined.
2006 The devices should be active md arrays which will be deactivated, as
2007 long as they are not currently in use.
2011 This will fully activate a partially assembled md array.
2015 This will mark an active array as read-only, providing that it is
2016 not currently being used.
2022 array back to being read/write.
2026 For all operations except
2029 will cause the operation to be applied to all arrays listed in
2034 causes all devices listed in the config file to be examined.
2037 .BR \-b ", " \-\-brief
2038 Be less verbose. This is used with
2046 gives an intermediate level of verbosity.
2052 .B mdadm \-\-monitor
2053 .I options... devices...
2058 to periodically poll a number of md arrays and to report on any events
2061 will never exit once it decides that there are arrays to be checked,
2062 so it should normally be run in the background.
2064 As well as reporting events,
2066 may move a spare drive from one array to another if they are in the
2071 and if the destination array has a failed drive but no spares.
2073 If any devices are listed on the command line,
2075 will only monitor those devices. Otherwise all arrays listed in the
2076 configuration file will be monitored. Further, if
2078 is given, then any other md devices that appear in
2080 will also be monitored.
2082 The result of monitoring the arrays is the generation of events.
2083 These events are passed to a separate program (if specified) and may
2084 be mailed to a given E-mail address.
2086 When passing events to a program, the program is run once for each event,
2087 and is given 2 or 3 command-line arguments: the first is the
2088 name of the event (see below), the second is the name of the
2089 md device which is affected, and the third is the name of a related
2090 device if relevant (such as a component device that has failed).
2094 is given, then a program or an E-mail address must be specified on the
2095 command line or in the config file. If neither are available, then
2097 will not monitor anything.
2101 will continue monitoring as long as something was found to monitor. If
2102 no program or email is given, then each event is reported to
2105 The different events are:
2109 .B DeviceDisappeared
2110 An md array which previously was configured appears to no longer be
2111 configured. (syslog priority: Critical)
2115 was told to monitor an array which is RAID0 or Linear, then it will
2117 .B DeviceDisappeared
2118 with the extra information
2120 This is because RAID0 and Linear do not support the device-failed,
2121 hot-spare and resync operations which are monitored.
2125 An md array started reconstruction. (syslog priority: Warning)
2131 is a two-digit number (ie. 05, 48). This indicates that rebuild
2132 has passed that many percent of the total. The events are generated
2133 with fixed increment since 0. Increment size may be specified with
2134 a commandline option (default is 20). (syslog priority: Warning)
2138 An md array that was rebuilding, isn't any more, either because it
2139 finished normally or was aborted. (syslog priority: Warning)
2143 An active component device of an array has been marked as
2144 faulty. (syslog priority: Critical)
2148 A spare component device which was being rebuilt to replace a faulty
2149 device has failed. (syslog priority: Critical)
2153 A spare component device which was being rebuilt to replace a faulty
2154 device has been successfully rebuilt and has been made active.
2155 (syslog priority: Info)
2159 A new md array has been detected in the
2161 file. (syslog priority: Info)
2165 A newly noticed array appears to be degraded. This message is not
2168 notices a drive failure which causes degradation, but only when
2170 notices that an array is degraded when it first sees the array.
2171 (syslog priority: Critical)
2175 A spare drive has been moved from one array in a
2179 to another to allow a failed drive to be replaced.
2180 (syslog priority: Info)
2186 has been told, via the config file, that an array should have a certain
2187 number of spare devices, and
2189 detects that it has fewer than this number when it first sees the
2190 array, it will report a
2193 (syslog priority: Warning)
2197 An array was found at startup, and the
2200 (syslog priority: Info)
2210 cause Email to be sent. All events cause the program to be run.
2211 The program is run with two or three arguments: the event
2212 name, the array device and possibly a second device.
2214 Each event has an associated array device (e.g.
2216 and possibly a second device. For
2221 the second device is the relevant component device.
2224 the second device is the array that the spare was moved from.
2228 to move spares from one array to another, the different arrays need to
2229 be labeled with the same
2231 or the spares must be allowed to migrate through matching POLICY domains
2232 in the configuration file. The
2234 name can be any string; it is only necessary that different spare
2235 groups use different names.
2239 detects that an array in a spare group has fewer active
2240 devices than necessary for the complete array, and has no spare
2241 devices, it will look for another array in the same spare group that
2242 has a full complement of working drive and a spare. It will then
2243 attempt to remove the spare from the second drive and add it to the
2245 If the removal succeeds but the adding fails, then it is added back to
2248 If the spare group for a degraded array is not defined,
2250 will look at the rules of spare migration specified by POLICY lines in
2252 and then follow similar steps as above if a matching spare is found.
2255 The GROW mode is used for changing the size or shape of an active
2257 For this to work, the kernel must support the necessary change.
2258 Various types of growth are being added during 2.6 development.
2260 Currently the supported changes include
2262 change the "size" attribute for RAID1, RAID4, RAID5 and RAID6.
2264 increase or decrease the "raid\-devices" attribute of RAID0, RAID1, RAID4,
2267 change the chunk-size and layout of RAID0, RAID4, RAID5 and RAID6.
2269 convert between RAID1 and RAID5, between RAID5 and RAID6, between
2270 RAID0, RAID4, and RAID5, and between RAID0 and RAID10 (in the near-2 mode).
2272 add a write-intent bitmap to any array which supports these bitmaps, or
2273 remove a write-intent bitmap from such an array.
2276 Using GROW on containers is currently supported only for Intel's IMSM
2277 container format. The number of devices in a container can be
2278 increased - which affects all arrays in the container - or an array
2279 in a container can be converted between levels where those levels are
2280 supported by the container, and the conversion is on of those listed
2281 above. Resizing arrays in an IMSM container with
2283 is not yet supported.
2285 Grow functionality (e.g. expand a number of raid devices) for Intel's
2286 IMSM container format has an experimental status. It is guarded by the
2287 .B MDADM_EXPERIMENTAL
2288 environment variable which must be set to '1' for a GROW command to
2290 This is for the following reasons:
2293 Intel's native IMSM check-pointing is not fully tested yet.
2294 This can causes IMSM incompatibility during the grow process: an array
2295 which is growing cannot roam between Microsoft Windows(R) and Linux
2299 Interrupting a grow operation is not recommended, because it
2300 has not been fully tested for Intel's IMSM container format yet.
2303 Note: Intel's native checkpointing doesn't use
2305 option and it is transparent for assembly feature.
2308 Normally when an array is built the "size" is taken from the smallest
2309 of the drives. If all the small drives in an arrays are, one at a
2310 time, removed and replaced with larger drives, then you could have an
2311 array of large drives with only a small amount used. In this
2312 situation, changing the "size" with "GROW" mode will allow the extra
2313 space to start being used. If the size is increased in this way, a
2314 "resync" process will start to make sure the new parts of the array
2317 Note that when an array changes size, any filesystem that may be
2318 stored in the array will not automatically grow or shrink to use or
2319 vacate the space. The
2320 filesystem will need to be explicitly told to use the extra space
2321 after growing, or to reduce its size
2323 to shrinking the array.
2325 Also the size of an array cannot be changed while it has an active
2326 bitmap. If an array has a bitmap, it must be removed before the size
2327 can be changed. Once the change is complete a new bitmap can be created.
2329 .SS RAID\-DEVICES CHANGES
2331 A RAID1 array can work with any number of devices from 1 upwards
2332 (though 1 is not very useful). There may be times which you want to
2333 increase or decrease the number of active devices. Note that this is
2334 different to hot-add or hot-remove which changes the number of
2337 When reducing the number of devices in a RAID1 array, the slots which
2338 are to be removed from the array must already be vacant. That is, the
2339 devices which were in those slots must be failed and removed.
2341 When the number of devices is increased, any hot spares that are
2342 present will be activated immediately.
2344 Changing the number of active devices in a RAID5 or RAID6 is much more
2345 effort. Every block in the array will need to be read and written
2346 back to a new location. From 2.6.17, the Linux Kernel is able to
2347 increase the number of devices in a RAID5 safely, including restarting
2348 an interrupted "reshape". From 2.6.31, the Linux Kernel is able to
2349 increase or decrease the number of devices in a RAID5 or RAID6.
2351 From 2.6.35, the Linux Kernel is able to convert a RAID0 in to a RAID4
2354 uses this functionality and the ability to add
2355 devices to a RAID4 to allow devices to be added to a RAID0. When
2356 requested to do this,
2358 will convert the RAID0 to a RAID4, add the necessary disks and make
2359 the reshape happen, and then convert the RAID4 back to RAID0.
2361 When decreasing the number of devices, the size of the array will also
2362 decrease. If there was data in the array, it could get destroyed and
2363 this is not reversible, so you should firstly shrink the filesystem on
2364 the array to fit within the new size. To help prevent accidents,
2366 requires that the size of the array be decreased first with
2367 .BR "mdadm --grow --array-size" .
2368 This is a reversible change which simply makes the end of the array
2369 inaccessible. The integrity of any data can then be checked before
2370 the non-reversible reduction in the number of devices is request.
2372 When relocating the first few stripes on a RAID5 or RAID6, it is not
2373 possible to keep the data on disk completely consistent and
2374 crash-proof. To provide the required safety, mdadm disables writes to
2375 the array while this "critical section" is reshaped, and takes a
2376 backup of the data that is in that section. For grows, this backup may be
2377 stored in any spare devices that the array has, however it can also be
2378 stored in a separate file specified with the
2380 option, and is required to be specified for shrinks, RAID level
2381 changes and layout changes. If this option is used, and the system
2382 does crash during the critical period, the same file must be passed to
2384 to restore the backup and reassemble the array. When shrinking rather
2385 than growing the array, the reshape is done from the end towards the
2386 beginning, so the "critical section" is at the end of the reshape.
2390 Changing the RAID level of any array happens instantaneously. However
2391 in the RAID5 to RAID6 case this requires a non-standard layout of the
2392 RAID6 data, and in the RAID6 to RAID5 case that non-standard layout is
2393 required before the change can be accomplished. So while the level
2394 change is instant, the accompanying layout change can take quite a
2397 is required. If the array is not simultaneously being grown or
2398 shrunk, so that the array size will remain the same - for example,
2399 reshaping a 3-drive RAID5 into a 4-drive RAID6 - the backup file will
2400 be used not just for a "cricital section" but throughout the reshape
2401 operation, as described below under LAYOUT CHANGES.
2403 .SS CHUNK-SIZE AND LAYOUT CHANGES
2405 Changing the chunk-size of layout without also changing the number of
2406 devices as the same time will involve re-writing all blocks in-place.
2407 To ensure against data loss in the case of a crash, a
2409 must be provided for these changes. Small sections of the array will
2410 be copied to the backup file while they are being rearranged. This
2411 means that all the data is copied twice, once to the backup and once
2412 to the new layout on the array, so this type of reshape will go very
2415 If the reshape is interrupted for any reason, this backup file must be
2417 .B "mdadm --assemble"
2418 so the array can be reassembled. Consequently the file cannot be
2419 stored on the device being reshaped.
2424 A write-intent bitmap can be added to, or removed from, an active
2425 array. Either internal bitmaps, or bitmaps stored in a separate file,
2426 can be added. Note that if you add a bitmap stored in a file which is
2427 in a filesystem that is on the RAID array being affected, the system
2428 will deadlock. The bitmap must be on a separate filesystem.
2430 .SH INCREMENTAL MODE
2434 .B mdadm \-\-incremental
2440 .B mdadm \-\-incremental \-\-fail
2444 .B mdadm \-\-incremental \-\-rebuild\-map
2447 .B mdadm \-\-incremental \-\-run \-\-scan
2450 This mode is designed to be used in conjunction with a device
2451 discovery system. As devices are found in a system, they can be
2453 .B "mdadm \-\-incremental"
2454 to be conditionally added to an appropriate array.
2456 Conversely, it can also be used with the
2458 flag to do just the opposite and find whatever array a particular device
2459 is part of and remove the device from that array.
2461 If the device passed is a
2463 device created by a previous call to
2465 then rather than trying to add that device to an array, all the arrays
2466 described by the metadata of the container will be started.
2469 performs a number of tests to determine if the device is part of an
2470 array, and which array it should be part of. If an appropriate array
2471 is found, or can be created,
2473 adds the device to the array and conditionally starts the array.
2477 will normally only add devices to an array which were previously working
2478 (active or spare) parts of that array. The support for automatic
2479 inclusion of a new drive as a spare in some array requires
2480 a configuration through POLICY in config file.
2484 makes are as follow:
2486 Is the device permitted by
2488 That is, is it listed in a
2490 line in that file. If
2492 is absent then the default it to allow any device. Similar if
2494 contains the special word
2496 then any device is allowed. Otherwise the device name given to
2498 must match one of the names or patterns in a
2503 Does the device have a valid md superblock? If a specific metadata
2504 version is requested with
2508 then only that style of metadata is accepted, otherwise
2510 finds any known version of metadata. If no
2512 metadata is found, the device may be still added to an array
2513 as a spare if POLICY allows.
2517 Does the metadata match an expected array?
2518 The metadata can match in two ways. Either there is an array listed
2521 which identifies the array (either by UUID, by name, by device list,
2522 or by minor-number), or the array was created with a
2528 or on the command line.
2531 is not able to positively identify the array as belonging to the
2532 current host, the device will be rejected.
2537 keeps a list of arrays that it has partially assembled in
2539 If no array exists which matches
2540 the metadata on the new device,
2542 must choose a device name and unit number. It does this based on any
2545 or any name information stored in the metadata. If this name
2546 suggests a unit number, that number will be used, otherwise a free
2547 unit number will be chosen. Normally
2549 will prefer to create a partitionable array, however if the
2553 suggests that a non-partitionable array is preferred, that will be
2556 If the array is not found in the config file and its metadata does not
2557 identify it as belonging to the "homehost", then
2559 will choose a name for the array which is certain not to conflict with
2560 any array which does belong to this host. It does this be adding an
2561 underscore and a small number to the name preferred by the metadata.
2563 Once an appropriate array is found or created and the device is added,
2565 must decide if the array is ready to be started. It will
2566 normally compare the number of available (non-spare) devices to the
2567 number of devices that the metadata suggests need to be active. If
2568 there are at least that many, the array will be started. This means
2569 that if any devices are missing the array will not be restarted.
2575 in which case the array will be run as soon as there are enough
2576 devices present for the data to be accessible. For a RAID1, that
2577 means one device will start the array. For a clean RAID5, the array
2578 will be started as soon as all but one drive is present.
2580 Note that neither of these approaches is really ideal. If it can
2581 be known that all device discovery has completed, then
2585 can be run which will try to start all arrays that are being
2586 incrementally assembled. They are started in "read-auto" mode in
2587 which they are read-only until the first write request. This means
2588 that no metadata updates are made and no attempt at resync or recovery
2589 happens. Further devices that are found before the first write can
2590 still be added safely.
2593 This section describes environment variables that affect how mdadm
2598 Setting this value to 1 will prevent mdadm from automatically launching
2599 mdmon. This variable is intended primarily for debugging mdadm/mdmon.
2605 does not create any device nodes in /dev, but leaves that task to
2609 appears not to be configured, or if this environment variable is set
2612 will create and devices that are needed.
2616 .B " mdadm \-\-query /dev/name-of-device"
2618 This will find out if a given device is a RAID array, or is part of
2619 one, and will provide brief information about the device.
2621 .B " mdadm \-\-assemble \-\-scan"
2623 This will assemble and start all arrays listed in the standard config
2624 file. This command will typically go in a system startup file.
2626 .B " mdadm \-\-stop \-\-scan"
2628 This will shut down all arrays that can be shut down (i.e. are not
2629 currently in use). This will typically go in a system shutdown script.
2631 .B " mdadm \-\-follow \-\-scan \-\-delay=120"
2633 If (and only if) there is an Email address or program given in the
2634 standard config file, then
2635 monitor the status of all arrays listed in that file by
2636 polling them ever 2 minutes.
2638 .B " mdadm \-\-create /dev/md0 \-\-level=1 \-\-raid\-devices=2 /dev/hd[ac]1"
2640 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
2643 .B " echo 'DEVICE /dev/hd*[0\-9] /dev/sd*[0\-9]' > mdadm.conf"
2645 .B " mdadm \-\-detail \-\-scan >> mdadm.conf"
2647 This will create a prototype config file that describes currently
2648 active arrays that are known to be made from partitions of IDE or SCSI drives.
2649 This file should be reviewed before being used as it may
2650 contain unwanted detail.
2652 .B " echo 'DEVICE /dev/hd[a\-z] /dev/sd*[a\-z]' > mdadm.conf"
2654 .B " mdadm \-\-examine \-\-scan \-\-config=mdadm.conf >> mdadm.conf"
2656 This will find arrays which could be assembled from existing IDE and
2657 SCSI whole drives (not partitions), and store the information in the
2658 format of a config file.
2659 This file is very likely to contain unwanted detail, particularly
2662 entries. It should be reviewed and edited before being used as an
2665 .B " mdadm \-\-examine \-\-brief \-\-scan \-\-config=partitions"
2667 .B " mdadm \-Ebsc partitions"
2669 Create a list of devices by reading
2670 .BR /proc/partitions ,
2671 scan these for RAID superblocks, and printout a brief listing of all
2674 .B " mdadm \-Ac partitions \-m 0 /dev/md0"
2676 Scan all partitions and devices listed in
2677 .BR /proc/partitions
2680 out of all such devices with a RAID superblock with a minor number of 0.
2682 .B " mdadm \-\-monitor \-\-scan \-\-daemonise > /run/mdadm/mon.pid"
2684 If config file contains a mail address or alert program, run mdadm in
2685 the background in monitor mode monitoring all md devices. Also write
2686 pid of mdadm daemon to
2687 .BR /run/mdadm/mon.pid .
2689 .B " mdadm \-Iq /dev/somedevice"
2691 Try to incorporate newly discovered device into some array as
2694 .B " mdadm \-\-incremental \-\-rebuild\-map \-\-run \-\-scan"
2696 Rebuild the array map from any current arrays, and then start any that
2699 .B " mdadm /dev/md4 --fail detached --remove detached"
2701 Any devices which are components of /dev/md4 will be marked as faulty
2702 and then remove from the array.
2704 .B " mdadm --grow /dev/md4 --level=6 --backup-file=/root/backup-md4"
2708 which is currently a RAID5 array will be converted to RAID6. There
2709 should normally already be a spare drive attached to the array as a
2710 RAID6 needs one more drive than a matching RAID5.
2712 .B " mdadm --create /dev/md/ddf --metadata=ddf --raid-disks 6 /dev/sd[a-f]"
2714 Create a DDF array over 6 devices.
2716 .B " mdadm --create /dev/md/home -n3 -l5 -z 30000000 /dev/md/ddf"
2718 Create a RAID5 array over any 3 devices in the given DDF set. Use
2719 only 30 gigabytes of each device.
2721 .B " mdadm -A /dev/md/ddf1 /dev/sd[a-f]"
2723 Assemble a pre-exist ddf array.
2725 .B " mdadm -I /dev/md/ddf1"
2727 Assemble all arrays contained in the ddf array, assigning names as
2730 .B " mdadm \-\-create \-\-help"
2732 Provide help about the Create mode.
2734 .B " mdadm \-\-config \-\-help"
2736 Provide help about the format of the config file.
2738 .B " mdadm \-\-help"
2740 Provide general help.
2750 lists all active md devices with information about them.
2752 uses this to find arrays when
2754 is given in Misc mode, and to monitor array reconstruction
2759 The config file lists which devices may be scanned to see if
2760 they contain MD super block, and gives identifying information
2761 (e.g. UUID) about known MD arrays. See
2768 mode is used, this file gets a list of arrays currently being created.
2773 understand two sorts of names for array devices.
2775 The first is the so-called 'standard' format name, which matches the
2776 names used by the kernel and which appear in
2779 The second sort can be freely chosen, but must reside in
2781 When giving a device name to
2783 to create or assemble an array, either full path name such as
2787 can be given, or just the suffix of the second sort of name, such as
2793 chooses device names during auto-assembly or incremental assembly, it
2794 will sometimes add a small sequence number to the end of the name to
2795 avoid conflicted between multiple arrays that have the same name. If
2797 can reasonably determine that the array really is meant for this host,
2798 either by a hostname in the metadata, or by the presence of the array
2801 then it will leave off the suffix if possible.
2802 Also if the homehost is specified as
2805 will only use a suffix if a different array of the same name already
2806 exists or is listed in the config file.
2808 The standard names for non-partitioned arrays (the only sort of md
2809 array available in 2.4 and earlier) are of the form
2813 where NN is a number.
2814 The standard names for partitionable arrays (as available from 2.6
2815 onwards) are of the form
2819 Partition numbers should be indicated by added "pMM" to these, thus "/dev/md/d1p2".
2821 From kernel version, 2.6.28 the "non-partitioned array" can actually
2822 be partitioned. So the "md_dNN" names are no longer needed, and
2823 partitions such as "/dev/mdNNpXX" are possible.
2827 was previously known as
2831 is completely separate from the
2833 package, and does not use the
2835 configuration file at all.
2838 For further information on mdadm usage, MD and the various levels of
2841 .B http://raid.wiki.kernel.org/
2843 (based upon Jakob \(/Ostergaard's Software\-RAID.HOWTO)
2845 .\"for new releases of the RAID driver check out:
2848 .\".UR ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2849 .\"ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2854 .\".UR http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2855 .\"http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2858 The latest version of
2860 should always be available from
2862 .B http://www.kernel.org/pub/linux/utils/raid/mdadm/