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"
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"
346 Use the new version-1 format superblock. This has fewer restrictions.
347 It can easily be moved between hosts with different endian-ness, and a
348 recovery operation can be checkpointed and restarted. The different
349 sub-versions store the superblock at different locations on the
350 device, either at the end (for 1.0), at the start (for 1.1) or 4K from
351 the start (for 1.2). "1" is equivalent to "1.2" (the commonly
352 preferred 1.x format).
353 'if '{DEFAULT_METADATA}'1.2' "default" is equivalent to "1.2".
355 Use the "Industry Standard" DDF (Disk Data Format) format defined by
357 When creating a DDF array a
359 will be created, and normal arrays can be created in that container.
361 Use the Intel(R) Matrix Storage Manager metadata format. This creates a
363 which is managed in a similar manner to DDF, and is supported by an
364 option-rom on some platforms:
366 .B http://www.intel.com/design/chipsets/matrixstorage_sb.htm
372 This will override any
374 setting in the config file and provides the identity of the host which
375 should be considered the home for any arrays.
377 When creating an array, the
379 will be recorded in the metadata. For version-1 superblocks, it will
380 be prefixed to the array name. For version-0.90 superblocks, part of
381 the SHA1 hash of the hostname will be stored in the later half of the
384 When reporting information about an array, any array which is tagged
385 for the given homehost will be reported as such.
387 When using Auto-Assemble, only arrays tagged for the given homehost
388 will be allowed to use 'local' names (i.e. not ending in '_' followed
389 by a digit string). See below under
390 .BR "Auto Assembly" .
396 needs to print the name for a device it normally finds the name in
398 which refers to the device and is shortest. When a path component is
402 will prefer a longer name if it contains that component. For example
403 .B \-\-prefer=by-uuid
404 will prefer a name in a subdirectory of
409 This functionality is currently only provided by
414 .SH For create, build, or grow:
417 .BR \-n ", " \-\-raid\-devices=
418 Specify the number of active devices in the array. This, plus the
419 number of spare devices (see below) must equal the number of
421 (including "\fBmissing\fP" devices)
422 that are listed on the command line for
424 Setting a value of 1 is probably
425 a mistake and so requires that
427 be specified first. A value of 1 will then be allowed for linear,
428 multipath, RAID0 and RAID1. It is never allowed for RAID4, RAID5 or RAID6.
430 This number can only be changed using
432 for RAID1, RAID4, RAID5 and RAID6 arrays, and only on kernels which provide
433 the necessary support.
436 .BR \-x ", " \-\-spare\-devices=
437 Specify the number of spare (eXtra) devices in the initial array.
438 Spares can also be added
439 and removed later. The number of component devices listed
440 on the command line must equal the number of RAID devices plus the
441 number of spare devices.
444 .BR \-z ", " \-\-size=
445 Amount (in Kibibytes) of space to use from each drive in RAID levels 1/4/5/6.
446 This must be a multiple of the chunk size, and must leave about 128Kb
447 of space at the end of the drive for the RAID superblock.
448 If this is not specified
449 (as it normally is not) the smallest drive (or partition) sets the
450 size, though if there is a variance among the drives of greater than 1%, a warning is
453 A suffix of 'M' or 'G' can be given to indicate Megabytes or
454 Gigabytes respectively.
456 Sometimes a replacement drive can be a little smaller than the
457 original drives though this should be minimised by IDEMA standards.
458 Such a replacement drive will be rejected by
460 To guard against this it can be useful to set the initial size
461 slightly smaller than the smaller device with the aim that it will
462 still be larger than any replacement.
464 This value can be set with
466 for RAID level 1/4/5/6 though
468 based arrays such as those with IMSM metadata may not be able to
470 If the array was created with a size smaller than the currently
471 active drives, the extra space can be accessed using
473 The size can be given as
475 which means to choose the largest size that fits on all current drives.
477 Before reducing the size of the array (with
478 .BR "\-\-grow \-\-size=" )
479 you should make sure that space isn't needed. If the device holds a
480 filesystem, you would need to resize the filesystem to use less space.
482 After reducing the array size you should check that the data stored in
483 the device is still available. If the device holds a filesystem, then
484 an 'fsck' of the filesystem is a minimum requirement. If there are
485 problems the array can be made bigger again with no loss with another
486 .B "\-\-grow \-\-size="
489 This value cannot be used when creating a
491 such as with DDF and IMSM metadata, though it perfectly valid when
492 creating an array inside a container.
495 .BR \-Z ", " \-\-array\-size=
496 This is only meaningful with
498 and its effect is not persistent: when the array is stopped and
499 restarted the default array size will be restored.
501 Setting the array-size causes the array to appear smaller to programs
502 that access the data. This is particularly needed before reshaping an
503 array so that it will be smaller. As the reshape is not reversible,
504 but setting the size with
506 is, it is required that the array size is reduced as appropriate
507 before the number of devices in the array is reduced.
509 Before reducing the size of the array you should make sure that space
510 isn't needed. If the device holds a filesystem, you would need to
511 resize the filesystem to use less space.
513 After reducing the array size you should check that the data stored in
514 the device is still available. If the device holds a filesystem, then
515 an 'fsck' of the filesystem is a minimum requirement. If there are
516 problems the array can be made bigger again with no loss with another
517 .B "\-\-grow \-\-array\-size="
520 A suffix of 'M' or 'G' can be given to indicate Megabytes or
521 Gigabytes respectively.
524 restores the apparent size of the array to be whatever the real
525 amount of available space is.
528 .BR \-c ", " \-\-chunk=
529 Specify chunk size of kibibytes. The default when creating an
530 array is 512KB. To ensure compatibility with earlier versions, the
531 default when Building and array with no persistent metadata is 64KB.
532 This is only meaningful for RAID0, RAID4, RAID5, RAID6, and RAID10.
534 RAID4, RAID5, RAID6, and RAID10 require the chunk size to be a power
535 of 2. In any case it must be a multiple of 4KB.
537 A suffix of 'M' or 'G' can be given to indicate Megabytes or
538 Gigabytes respectively.
542 Specify rounding factor for a Linear array. The size of each
543 component will be rounded down to a multiple of this size.
544 This is a synonym for
546 but highlights the different meaning for Linear as compared to other
547 RAID levels. The default is 64K if a kernel earlier than 2.6.16 is in
548 use, and is 0K (i.e. no rounding) in later kernels.
551 .BR \-l ", " \-\-level=
552 Set RAID level. When used with
554 options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
555 raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty, container.
556 Obviously some of these are synonymous.
560 metadata type is requested, only the
562 level is permitted, and it does not need to be explicitly given.
566 only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid.
570 to change the RAID level in some cases. See LEVEL CHANGES below.
573 .BR \-p ", " \-\-layout=
574 This option configures the fine details of data layout for RAID5, RAID6,
575 and RAID10 arrays, and controls the failure modes for
578 The layout of the RAID5 parity block can be one of
579 .BR left\-asymmetric ,
580 .BR left\-symmetric ,
581 .BR right\-asymmetric ,
582 .BR right\-symmetric ,
583 .BR la ", " ra ", " ls ", " rs .
585 .BR left\-symmetric .
587 It is also possible to cause RAID5 to use a RAID4-like layout by
593 Finally for RAID5 there are DDF\-compatible layouts,
594 .BR ddf\-zero\-restart ,
595 .BR ddf\-N\-restart ,
597 .BR ddf\-N\-continue .
599 These same layouts are available for RAID6. There are also 4 layouts
600 that will provide an intermediate stage for converting between RAID5
601 and RAID6. These provide a layout which is identical to the
602 corresponding RAID5 layout on the first N\-1 devices, and has the 'Q'
603 syndrome (the second 'parity' block used by RAID6) on the last device.
605 .BR left\-symmetric\-6 ,
606 .BR right\-symmetric\-6 ,
607 .BR left\-asymmetric\-6 ,
608 .BR right\-asymmetric\-6 ,
610 .BR parity\-first\-6 .
612 When setting the failure mode for level
615 .BR write\-transient ", " wt ,
616 .BR read\-transient ", " rt ,
617 .BR write\-persistent ", " wp ,
618 .BR read\-persistent ", " rp ,
620 .BR read\-fixable ", " rf ,
621 .BR clear ", " flush ", " none .
623 Each failure mode can be followed by a number, which is used as a period
624 between fault generation. Without a number, the fault is generated
625 once on the first relevant request. With a number, the fault will be
626 generated after that many requests, and will continue to be generated
627 every time the period elapses.
629 Multiple failure modes can be current simultaneously by using the
631 option to set subsequent failure modes.
633 "clear" or "none" will remove any pending or periodic failure modes,
634 and "flush" will clear any persistent faults.
636 Finally, the layout options for RAID10 are one of 'n', 'o' or 'f' followed
637 by a small number. The default is 'n2'. The supported options are:
640 signals 'near' copies. Multiple copies of one data block are at
641 similar offsets in different devices.
644 signals 'offset' copies. Rather than the chunks being duplicated
645 within a stripe, whole stripes are duplicated but are rotated by one
646 device so duplicate blocks are on different devices. Thus subsequent
647 copies of a block are in the next drive, and are one chunk further
652 (multiple copies have very different offsets).
653 See md(4) for more detail about 'near', 'offset', and 'far'.
655 The number is the number of copies of each datablock. 2 is normal, 3
656 can be useful. This number can be at most equal to the number of
657 devices in the array. It does not need to divide evenly into that
658 number (e.g. it is perfectly legal to have an 'n2' layout for an array
659 with an odd number of devices).
661 When an array is converted between RAID5 and RAID6 an intermediate
662 RAID6 layout is used in which the second parity block (Q) is always on
663 the last device. To convert a RAID5 to RAID6 and leave it in this new
664 layout (which does not require re-striping) use
665 .BR \-\-layout=preserve .
666 This will try to avoid any restriping.
668 The converse of this is
669 .B \-\-layout=normalise
670 which will change a non-standard RAID6 layout into a more standard
677 (thus explaining the p of
681 .BR \-b ", " \-\-bitmap=
682 Specify a file to store a write-intent bitmap in. The file should not
685 is also given. The same file should be provided
686 when assembling the array. If the word
688 is given, then the bitmap is stored with the metadata on the array,
689 and so is replicated on all devices. If the word
693 mode, then any bitmap that is present is removed.
695 To help catch typing errors, the filename must contain at least one
696 slash ('/') if it is a real file (not 'internal' or 'none').
698 Note: external bitmaps are only known to work on ext2 and ext3.
699 Storing bitmap files on other filesystems may result in serious problems.
702 .BR \-\-bitmap\-chunk=
703 Set the chunksize of the bitmap. Each bit corresponds to that many
704 Kilobytes of storage.
705 When using a file based bitmap, the default is to use the smallest
706 size that is at-least 4 and requires no more than 2^21 chunks.
709 bitmap, the chunksize defaults to 64Meg, or larger if necessary to
710 fit the bitmap into the available space.
712 A suffix of 'M' or 'G' can be given to indicate Megabytes or
713 Gigabytes respectively.
716 .BR \-W ", " \-\-write\-mostly
717 subsequent devices listed in a
722 command will be flagged as 'write-mostly'. This is valid for RAID1
723 only and means that the 'md' driver will avoid reading from these
724 devices if at all possible. This can be useful if mirroring over a
728 .BR \-\-write\-behind=
729 Specify that write-behind mode should be enabled (valid for RAID1
730 only). If an argument is specified, it will set the maximum number
731 of outstanding writes allowed. The default value is 256.
732 A write-intent bitmap is required in order to use write-behind
733 mode, and write-behind is only attempted on drives marked as
737 .BR \-\-assume\-clean
740 that the array pre-existed and is known to be clean. It can be useful
741 when trying to recover from a major failure as you can be sure that no
742 data will be affected unless you actually write to the array. It can
743 also be used when creating a RAID1 or RAID10 if you want to avoid the
744 initial resync, however this practice \(em while normally safe \(em is not
745 recommended. Use this only if you really know what you are doing.
747 When the devices that will be part of a new array were filled
748 with zeros before creation the operator knows the array is
749 actually clean. If that is the case, such as after running
750 badblocks, this argument can be used to tell mdadm the
751 facts the operator knows.
753 When an array is resized to a larger size with
754 .B "\-\-grow \-\-size="
755 the new space is normally resynced in that same way that the whole
756 array is resynced at creation. From Linux version 3.0,
758 can be used with that command to avoid the automatic resync.
761 .BR \-\-backup\-file=
764 is used to increase the number of raid-devices in a RAID5 or RAID6 if
765 there are no spare devices available, or to shrink, change RAID level
766 or layout. See the GROW MODE section below on RAID\-DEVICES CHANGES.
767 The file must be stored on a separate device, not on the RAID array
772 This option is complementary to the
773 .B \-\-freeze-reshape
774 option for assembly. It is needed when
776 operation is interrupted and it is not restarted automatically due to
777 .B \-\-freeze-reshape
778 usage during array assembly. This option is used together with
782 ) command and device for a pending reshape to be continued.
783 All parameters required for reshape continuation will be read from array metadata.
787 .BR \-\-backup\-file=
788 option to be set, continuation option will require to have exactly the same
789 backup file given as well.
791 Any other parameter passed together with
793 option will be ignored.
796 .BR \-N ", " \-\-name=
799 for the array. This is currently only effective when creating an
800 array with a version-1 superblock, or an array in a DDF container.
801 The name is a simple textual string that can be used to identify array
802 components when assembling. If name is needed but not specified, it
803 is taken from the basename of the device that is being created.
815 run the array, even if some of the components
816 appear to be active in another array or filesystem. Normally
818 will ask for confirmation before including such components in an
819 array. This option causes that question to be suppressed.
822 .BR \-f ", " \-\-force
825 accept the geometry and layout specified without question. Normally
827 will not allow creation of an array with only one device, and will try
828 to create a RAID5 array with one missing drive (as this makes the
829 initial resync work faster). With
832 will not try to be so clever.
835 .BR \-a ", " "\-\-auto{=yes,md,mdp,part,p}{NN}"
836 Instruct mdadm how to create the device file if needed, possibly allocating
837 an unused minor number. "md" causes a non-partitionable array
838 to be used (though since Linux 2.6.28, these array devices are in fact
839 partitionable). "mdp", "part" or "p" causes a partitionable array (2.6 and
840 later) to be used. "yes" requires the named md device to have
841 a 'standard' format, and the type and minor number will be determined
842 from this. With mdadm 3.0, device creation is normally left up to
844 so this option is unlikely to be needed.
845 See DEVICE NAMES below.
847 The argument can also come immediately after
852 is not given on the command line or in the config file, then
858 is also given, then any
860 entries in the config file will override the
862 instruction given on the command line.
864 For partitionable arrays,
866 will create the device file for the whole array and for the first 4
867 partitions. A different number of partitions can be specified at the
868 end of this option (e.g.
870 If the device name ends with a digit, the partition names add a 'p',
872 .IR /dev/md/home1p3 .
873 If there is no trailing digit, then the partition names just have a
875 .IR /dev/md/scratch3 .
877 If the md device name is in a 'standard' format as described in DEVICE
878 NAMES, then it will be created, if necessary, with the appropriate
879 device number based on that name. If the device name is not in one of these
880 formats, then a unused device number will be allocated. The device
881 number will be considered unused if there is no active array for that
882 number, and there is no entry in /dev for that number and with a
883 non-standard name. Names that are not in 'standard' format are only
884 allowed in "/dev/md/".
886 This is meaningful with
893 .\".BR \-\-symlink = no
898 .\"to create devices in
900 .\"it will also create symlinks from
902 .\"with names starting with
908 .\"to suppress this, or
909 .\".B \-\-symlink=yes
910 .\"to enforce this even if it is suppressing
916 .BR \-a ", " "\-\-add"
917 This option can be used in Grow mode in two cases.
919 If the target array is a Linear array, then
921 can be used to add one or more devices to the array. They
922 are simply catenated on to the end of the array. Once added, the
923 devices cannot be removed.
927 option is being used to increase the number of devices in an array,
930 can be used to add some extra devices to be included in the array.
931 In most cases this is not needed as the extra devices can be added as
932 spares first, and then the number of raid-disks can be changed.
933 However for RAID0, it is not possible to add spares. So to increase
934 the number of devices in a RAID0, it is necessary to set the new
935 number of devices, and to add the new devices, in the same command.
940 .BR \-u ", " \-\-uuid=
941 uuid of array to assemble. Devices which don't have this uuid are
945 .BR \-m ", " \-\-super\-minor=
946 Minor number of device that array was created for. Devices which
947 don't have this minor number are excluded. If you create an array as
948 /dev/md1, then all superblocks will contain the minor number 1, even if
949 the array is later assembled as /dev/md2.
951 Giving the literal word "dev" for
955 to use the minor number of the md device that is being assembled.
958 .B \-\-super\-minor=dev
959 will look for super blocks with a minor number of 0.
962 is only relevant for v0.90 metadata, and should not normally be used.
968 .BR \-N ", " \-\-name=
969 Specify the name of the array to assemble. This must be the name
970 that was specified when creating the array. It must either match
971 the name stored in the superblock exactly, or it must match
974 prefixed to the start of the given name.
977 .BR \-f ", " \-\-force
978 Assemble the array even if the metadata on some devices appears to be
981 cannot find enough working devices to start the array, but can find
982 some devices that are recorded as having failed, then it will mark
983 those devices as working so that the array can be started.
984 An array which requires
986 to be started may contain data corruption. Use it carefully.
990 Attempt to start the array even if fewer drives were given than were
991 present last time the array was active. Normally if not all the
992 expected drives are found and
994 is not used, then the array will be assembled but not started.
997 an attempt will be made to start it anyway.
1001 This is the reverse of
1003 in that it inhibits the startup of array unless all expected drives
1004 are present. This is only needed with
1006 and can be used if the physical connections to devices are
1007 not as reliable as you would like.
1010 .BR \-a ", " "\-\-auto{=no,yes,md,mdp,part}"
1011 See this option under Create and Build options.
1014 .BR \-b ", " \-\-bitmap=
1015 Specify the bitmap file that was given when the array was created. If
1018 bitmap, there is no need to specify this when assembling the array.
1021 .BR \-\-backup\-file=
1024 was used while reshaping an array (e.g. changing number of devices or
1025 chunk size) and the system crashed during the critical section, then the same
1027 must be presented to
1029 to allow possibly corrupted data to be restored, and the reshape
1033 .BR \-\-invalid\-backup
1034 If the file needed for the above option is not available for any
1035 reason an empty file can be given together with this option to
1036 indicate that the backup file is invalid. In this case the data that
1037 was being rearranged at the time of the crash could be irrecoverably
1038 lost, but the rest of the array may still be recoverable. This option
1039 should only be used as a last resort if there is no way to recover the
1044 .BR \-U ", " \-\-update=
1045 Update the superblock on each device while assembling the array. The
1046 argument given to this flag can be one of
1061 option will adjust the superblock of an array what was created on a Sparc
1062 machine running a patched 2.2 Linux kernel. This kernel got the
1063 alignment of part of the superblock wrong. You can use the
1064 .B "\-\-examine \-\-sparc2.2"
1067 to see what effect this would have.
1071 option will update the
1072 .B "preferred minor"
1073 field on each superblock to match the minor number of the array being
1075 This can be useful if
1077 reports a different "Preferred Minor" to
1079 In some cases this update will be performed automatically
1080 by the kernel driver. In particular the update happens automatically
1081 at the first write to an array with redundancy (RAID level 1 or
1082 greater) on a 2.6 (or later) kernel.
1086 option will change the uuid of the array. If a UUID is given with the
1088 option that UUID will be used as a new UUID and will
1090 be used to help identify the devices in the array.
1093 is given, a random UUID is chosen.
1097 option will change the
1099 of the array as stored in the superblock. This is only supported for
1100 version-1 superblocks.
1104 option will change the
1106 as recorded in the superblock. For version-0 superblocks, this is the
1107 same as updating the UUID.
1108 For version-1 superblocks, this involves updating the name.
1112 option will cause the array to be marked
1114 meaning that any redundancy in the array (e.g. parity for RAID5,
1115 copies for RAID1) may be incorrect. This will cause the RAID system
1116 to perform a "resync" pass to make sure that all redundant information
1121 option allows arrays to be moved between machines with different
1123 When assembling such an array for the first time after a move, giving
1124 .B "\-\-update=byteorder"
1127 to expect superblocks to have their byteorder reversed, and will
1128 correct that order before assembling the array. This is only valid
1129 with original (Version 0.90) superblocks.
1133 option will correct the summaries in the superblock. That is the
1134 counts of total, working, active, failed, and spare devices.
1138 option will rarely be of use. It applies to version 1.1 and 1.2 metadata
1139 only (where the metadata is at the start of the device) and is only
1140 useful when the component device has changed size (typically become
1141 larger). The version 1 metadata records the amount of the device that
1142 can be used to store data, so if a device in a version 1.1 or 1.2
1143 array becomes larger, the metadata will still be visible, but the
1144 extra space will not. In this case it might be useful to assemble the
1146 .BR \-\-update=devicesize .
1149 to determine the maximum usable amount of space on each device and
1150 update the relevant field in the metadata.
1154 option can be used when an array has an internal bitmap which is
1155 corrupt in some way so that assembling the array normally fails. It
1156 will cause any internal bitmap to be ignored.
1159 .BR \-\-freeze\-reshape
1160 Option is intended to be used in start-up scripts during initrd boot phase.
1161 When array under reshape is assembled during initrd phase, this option
1162 stops reshape after reshape critical section is being restored. This happens
1163 before file system pivot operation and avoids loss of file system context.
1164 Losing file system context would cause reshape to be broken.
1166 Reshape can be continued later using the
1168 option for the grow command.
1170 .SH For Manage mode:
1173 .BR \-t ", " \-\-test
1174 Unless a more serious error occurred,
1176 will exit with a status of 2 if no changes were made to the array and
1177 0 if at least one change was made.
1178 This can be useful when an indirect specifier such as
1183 is used in requesting an operation on the array.
1185 will report failure if these specifiers didn't find any match.
1188 .BR \-a ", " \-\-add
1189 hot-add listed devices.
1190 If a device appears to have recently been part of the array
1191 (possibly it failed or was removed) the device is re\-added as described
1193 If that fails or the device was never part of the array, the device is
1194 added as a hot-spare.
1195 If the array is degraded, it will immediately start to rebuild data
1198 Note that this and the following options are only meaningful on array
1199 with redundancy. They don't apply to RAID0 or Linear.
1203 re\-add a device that was previous removed from an array.
1204 If the metadata on the device reports that it is a member of the
1205 array, and the slot that it used is still vacant, then the device will
1206 be added back to the array in the same position. This will normally
1207 cause the data for that device to be recovered. However based on the
1208 event count on the device, the recovery may only require sections that
1209 are flagged a write-intent bitmap to be recovered or may not require
1210 any recovery at all.
1212 When used on an array that has no metadata (i.e. it was built with
1214 it will be assumed that bitmap-based recovery is enough to make the
1215 device fully consistent with the array.
1219 can be accompanied by
1220 .BR \-\-update=devicesize .
1221 See the description of this option when used in Assemble mode for an
1222 explanation of its use.
1224 If the device name given is
1226 then mdadm will try to find any device that looks like it should be
1227 part of the array but isn't and will try to re\-add all such devices.
1230 .BR \-r ", " \-\-remove
1231 remove listed devices. They must not be active. i.e. they should
1232 be failed or spare devices. As well as the name of a device file
1241 The first causes all failed device to be removed. The second causes
1242 any device which is no longer connected to the system (i.e an 'open'
1245 to be removed. This will only succeed for devices that are spares or
1246 have already been marked as failed.
1249 .BR \-f ", " \-\-fail
1250 mark listed devices as faulty.
1251 As well as the name of a device file, the word
1253 can be given. This will cause any device that has been detached from
1254 the system to be marked as failed. It can then be removed.
1262 .BR \-\-write\-mostly
1263 Subsequent devices that are added or re\-added will have the 'write-mostly'
1264 flag set. This is only valid for RAID1 and means that the 'md' driver
1265 will avoid reading from these devices if possible.
1268 Subsequent devices that are added or re\-added will have the 'write-mostly'
1272 Each of these options requires that the first device listed is the array
1273 to be acted upon, and the remainder are component devices to be added,
1274 removed, marked as faulty, etc. Several different operations can be
1275 specified for different devices, e.g.
1277 mdadm /dev/md0 \-\-add /dev/sda1 \-\-fail /dev/sdb1 \-\-remove /dev/sdb1
1279 Each operation applies to all devices listed until the next
1282 If an array is using a write-intent bitmap, then devices which have
1283 been removed can be re\-added in a way that avoids a full
1284 reconstruction but instead just updates the blocks that have changed
1285 since the device was removed. For arrays with persistent metadata
1286 (superblocks) this is done automatically. For arrays created with
1288 mdadm needs to be told that this device we removed recently with
1291 Devices can only be removed from an array if they are not in active
1292 use, i.e. that must be spares or failed devices. To remove an active
1293 device, it must first be marked as
1299 .BR \-Q ", " \-\-query
1300 Examine a device to see
1301 (1) if it is an md device and (2) if it is a component of an md
1303 Information about what is discovered is presented.
1306 .BR \-D ", " \-\-detail
1307 Print details of one or more md devices.
1310 .BR \-\-detail\-platform
1311 Print details of the platform's RAID capabilities (firmware / hardware
1312 topology) for a given metadata format.
1315 .BR \-Y ", " \-\-export
1320 output will be formatted as
1322 pairs for easy import into the environment.
1325 .BR \-E ", " \-\-examine
1326 Print contents of the metadata stored on the named device(s).
1327 Note the contrast between
1332 applies to devices which are components of an array, while
1334 applies to a whole array which is currently active.
1337 If an array was created on a SPARC machine with a 2.2 Linux kernel
1338 patched with RAID support, the superblock will have been created
1339 incorrectly, or at least incompatibly with 2.4 and later kernels.
1344 will fix the superblock before displaying it. If this appears to do
1345 the right thing, then the array can be successfully assembled using
1346 .BR "\-\-assemble \-\-update=sparc2.2" .
1349 .BR \-X ", " \-\-examine\-bitmap
1350 Report information about a bitmap file.
1351 The argument is either an external bitmap file or an array component
1352 in case of an internal bitmap. Note that running this on an array
1355 does not report the bitmap for that array.
1358 .BR \-R ", " \-\-run
1359 start a partially assembled array. If
1361 did not find enough devices to fully start the array, it might leaving
1362 it partially assembled. If you wish, you can then use
1364 to start the array in degraded mode.
1367 .BR \-S ", " \-\-stop
1368 deactivate array, releasing all resources.
1371 .BR \-o ", " \-\-readonly
1372 mark array as readonly.
1375 .BR \-w ", " \-\-readwrite
1376 mark array as readwrite.
1379 .B \-\-zero\-superblock
1380 If the device contains a valid md superblock, the block is
1381 overwritten with zeros. With
1383 the block where the superblock would be is overwritten even if it
1384 doesn't appear to be valid.
1387 .B \-\-kill\-subarray=
1388 If the device is a container and the argument to \-\-kill\-subarray
1389 specifies an inactive subarray in the container, then the subarray is
1390 deleted. Deleting all subarrays will leave an 'empty-container' or
1391 spare superblock on the drives. See \-\-zero\-superblock for completely
1392 removing a superblock. Note that some formats depend on the subarray
1393 index for generating a UUID, this command will fail if it would change
1394 the UUID of an active subarray.
1397 .B \-\-update\-subarray=
1398 If the device is a container and the argument to \-\-update\-subarray
1399 specifies a subarray in the container, then attempt to update the given
1400 superblock field in the subarray. See below in
1405 .BR \-t ", " \-\-test
1410 is set to reflect the status of the device. See below in
1415 .BR \-W ", " \-\-wait
1416 For each md device given, wait for any resync, recovery, or reshape
1417 activity to finish before returning.
1419 will return with success if it actually waited for every device
1420 listed, otherwise it will return failure.
1424 For each md device given, or each device in /proc/mdstat if
1426 is given, arrange for the array to be marked clean as soon as possible.
1428 will return with success if the array uses external metadata and we
1429 successfully waited. For native arrays this returns immediately as the
1430 kernel handles dirty-clean transitions at shutdown. No action is taken
1431 if safe-mode handling is disabled.
1433 .SH For Incremental Assembly mode:
1435 .BR \-\-rebuild\-map ", " \-r
1436 Rebuild the map file
1440 uses to help track which arrays are currently being assembled.
1443 .BR \-\-run ", " \-R
1444 Run any array assembled as soon as a minimal number of devices are
1445 available, rather than waiting until all expected devices are present.
1448 .BR \-\-scan ", " \-s
1449 Only meaningful with
1453 file for arrays that are being incrementally assembled and will try to
1454 start any that are not already started. If any such array is listed
1457 as requiring an external bitmap, that bitmap will be attached first.
1460 .BR \-\-fail ", " \-f
1461 This allows the hot-plug system to remove devices that have fully disappeared
1462 from the kernel. It will first fail and then remove the device from any
1463 array it belongs to.
1464 The device name given should be a kernel device name such as "sda",
1470 Only used with \-\-fail. The 'path' given will be recorded so that if
1471 a new device appears at the same location it can be automatically
1472 added to the same array. This allows the failed device to be
1473 automatically replaced by a new device without metadata if it appears
1474 at specified path. This option is normally only set by a
1478 .SH For Monitor mode:
1480 .BR \-m ", " \-\-mail
1481 Give a mail address to send alerts to.
1484 .BR \-p ", " \-\-program ", " \-\-alert
1485 Give a program to be run whenever an event is detected.
1488 .BR \-y ", " \-\-syslog
1489 Cause all events to be reported through 'syslog'. The messages have
1490 facility of 'daemon' and varying priorities.
1493 .BR \-d ", " \-\-delay
1494 Give a delay in seconds.
1496 polls the md arrays and then waits this many seconds before polling
1497 again. The default is 60 seconds. Since 2.6.16, there is no need to
1498 reduce this as the kernel alerts
1500 immediately when there is any change.
1503 .BR \-r ", " \-\-increment
1504 Give a percentage increment.
1506 will generate RebuildNN events with the given percentage increment.
1509 .BR \-f ", " \-\-daemonise
1512 to run as a background daemon if it decides to monitor anything. This
1513 causes it to fork and run in the child, and to disconnect from the
1514 terminal. The process id of the child is written to stdout.
1517 which will only continue monitoring if a mail address or alert program
1518 is found in the config file.
1521 .BR \-i ", " \-\-pid\-file
1524 is running in daemon mode, write the pid of the daemon process to
1525 the specified file, instead of printing it on standard output.
1528 .BR \-1 ", " \-\-oneshot
1529 Check arrays only once. This will generate
1531 events and more significantly
1537 .B " mdadm \-\-monitor \-\-scan \-1"
1539 from a cron script will ensure regular notification of any degraded arrays.
1542 .BR \-t ", " \-\-test
1545 alert for every array found at startup. This alert gets mailed and
1546 passed to the alert program. This can be used for testing that alert
1547 message do get through successfully.
1551 This inhibits the functionality for moving spares between arrays.
1552 Only one monitoring process started with
1554 but without this flag is allowed, otherwise the two could interfere
1561 .B mdadm \-\-assemble
1562 .I md-device options-and-component-devices...
1565 .B mdadm \-\-assemble \-\-scan
1566 .I md-devices-and-options...
1569 .B mdadm \-\-assemble \-\-scan
1573 This usage assembles one or more RAID arrays from pre-existing components.
1574 For each array, mdadm needs to know the md device, the identity of the
1575 array, and a number of component-devices. These can be found in a number of ways.
1577 In the first usage example (without the
1579 the first device given is the md device.
1580 In the second usage example, all devices listed are treated as md
1581 devices and assembly is attempted.
1582 In the third (where no devices are listed) all md devices that are
1583 listed in the configuration file are assembled. If no arrays are
1584 described by the configuration file, then any arrays that
1585 can be found on unused devices will be assembled.
1587 If precisely one device is listed, but
1593 was given and identity information is extracted from the configuration file.
1595 The identity can be given with the
1601 option, will be taken from the md-device record in the config file, or
1602 will be taken from the super block of the first component-device
1603 listed on the command line.
1605 Devices can be given on the
1607 command line or in the config file. Only devices which have an md
1608 superblock which contains the right identity will be considered for
1611 The config file is only used if explicitly named with
1613 or requested with (a possibly implicit)
1618 .B /etc/mdadm/mdadm.conf
1623 is not given, then the config file will only be used to find the
1624 identity of md arrays.
1626 Normally the array will be started after it is assembled. However if
1628 is not given and not all expected drives were listed, then the array
1629 is not started (to guard against usage errors). To insist that the
1630 array be started in this case (as may work for RAID1, 4, 5, 6, or 10),
1639 does not create any entries in
1643 It does record information in
1647 to choose the correct name.
1651 detects that udev is not configured, it will create the devices in
1655 In Linux kernels prior to version 2.6.28 there were two distinctly
1656 different types of md devices that could be created: one that could be
1657 partitioned using standard partitioning tools and one that could not.
1658 Since 2.6.28 that distinction is no longer relevant as both type of
1659 devices can be partitioned.
1661 will normally create the type that originally could not be partitioned
1662 as it has a well defined major number (9).
1664 Prior to 2.6.28, it is important that mdadm chooses the correct type
1665 of array device to use. This can be controlled with the
1667 option. In particular, a value of "mdp" or "part" or "p" tells mdadm
1668 to use a partitionable device rather than the default.
1670 In the no-udev case, the value given to
1672 can be suffixed by a number. This tells
1674 to create that number of partition devices rather than the default of 4.
1678 can also be given in the configuration file as a word starting
1680 on the ARRAY line for the relevant array.
1687 and no devices are listed,
1689 will first attempt to assemble all the arrays listed in the config
1692 If no arrays are listed in the config (other than those marked
1694 it will look through the available devices for possible arrays and
1695 will try to assemble anything that it finds. Arrays which are tagged
1696 as belonging to the given homehost will be assembled and started
1697 normally. Arrays which do not obviously belong to this host are given
1698 names that are expected not to conflict with anything local, and are
1699 started "read-auto" so that nothing is written to any device until the
1700 array is written to. i.e. automatic resync etc is delayed.
1704 finds a consistent set of devices that look like they should comprise
1705 an array, and if the superblock is tagged as belonging to the given
1706 home host, it will automatically choose a device name and try to
1707 assemble the array. If the array uses version-0.90 metadata, then the
1709 number as recorded in the superblock is used to create a name in
1713 If the array uses version-1 metadata, then the
1715 from the superblock is used to similarly create a name in
1717 (the name will have any 'host' prefix stripped first).
1719 This behaviour can be modified by the
1723 configuration file. This line can indicate that specific metadata
1724 type should, or should not, be automatically assembled. If an array
1725 is found which is not listed in
1727 and has a metadata format that is denied by the
1729 line, then it will not be assembled.
1732 line can also request that all arrays identified as being for this
1733 homehost should be assembled regardless of their metadata type.
1736 for further details.
1738 Note: Auto assembly cannot be used for assembling and activating some
1739 arrays which are undergoing reshape. In particular as the
1741 cannot be given, any reshape which requires a backup-file to continue
1742 cannot be started by auto assembly. An array which is growing to more
1743 devices and has passed the critical section can be assembled using
1754 .BI \-\-raid\-devices= Z
1758 This usage is similar to
1760 The difference is that it creates an array without a superblock. With
1761 these arrays there is no difference between initially creating the array and
1762 subsequently assembling the array, except that hopefully there is useful
1763 data there in the second case.
1765 The level may raid0, linear, raid1, raid10, multipath, or faulty, or
1766 one of their synonyms. All devices must be listed and the array will
1767 be started once complete. It will often be appropriate to use
1768 .B \-\-assume\-clean
1769 with levels raid1 or raid10.
1780 .BI \-\-raid\-devices= Z
1784 This usage will initialise a new md array, associate some devices with
1785 it, and activate the array.
1787 The named device will normally not exist when
1788 .I "mdadm \-\-create"
1789 is run, but will be created by
1791 once the array becomes active.
1793 As devices are added, they are checked to see if they contain RAID
1794 superblocks or filesystems. They are also checked to see if the variance in
1795 device size exceeds 1%.
1797 If any discrepancy is found, the array will not automatically be run, though
1800 can override this caution.
1802 To create a "degraded" array in which some devices are missing, simply
1803 give the word "\fBmissing\fP"
1804 in place of a device name. This will cause
1806 to leave the corresponding slot in the array empty.
1807 For a RAID4 or RAID5 array at most one slot can be
1808 "\fBmissing\fP"; for a RAID6 array at most two slots.
1809 For a RAID1 array, only one real device needs to be given. All of the
1813 When creating a RAID5 array,
1815 will automatically create a degraded array with an extra spare drive.
1816 This is because building the spare into a degraded array is in general
1817 faster than resyncing the parity on a non-degraded, but not clean,
1818 array. This feature can be overridden with the
1822 When creating an array with version-1 metadata a name for the array is
1824 If this is not given with the
1828 will choose a name based on the last component of the name of the
1829 device being created. So if
1831 is being created, then the name
1836 is being created, then the name
1840 When creating a partition based array, using
1842 with version-1.x metadata, the partition type should be set to
1844 (non fs-data). This type selection allows for greater precision since
1845 using any other [RAID auto-detect (0xFD) or a GNU/Linux partition (0x83)],
1846 might create problems in the event of array recovery through a live cdrom.
1848 A new array will normally get a randomly assigned 128bit UUID which is
1849 very likely to be unique. If you have a specific need, you can choose
1850 a UUID for the array by giving the
1852 option. Be warned that creating two arrays with the same UUID is a
1853 recipe for disaster. Also, using
1855 when creating a v0.90 array will silently override any
1860 .\"option is given, it is not necessary to list any component-devices in this command.
1861 .\"They can be added later, before a
1865 .\"is given, the apparent size of the smallest drive given is used.
1867 When creating an array within a
1870 can be given either the list of devices to use, or simply the name of
1871 the container. The former case gives control over which devices in
1872 the container will be used for the array. The latter case allows
1874 to automatically choose which devices to use based on how much spare
1877 The General Management options that are valid with
1882 insist on running the array even if some devices look like they might
1887 start the array readonly \(em not supported yet.
1894 .I options... devices...
1897 This usage will allow individual devices in an array to be failed,
1898 removed or added. It is possible to perform multiple operations with
1899 on command. For example:
1901 .B " mdadm /dev/md0 \-f /dev/hda1 \-r /dev/hda1 \-a /dev/hda1"
1907 and will then remove it from the array and finally add it back
1908 in as a spare. However only one md array can be affected by a single
1911 When a device is added to an active array, mdadm checks to see if it
1912 has metadata on it which suggests that it was recently a member of the
1913 array. If it does, it tries to "re\-add" the device. If there have
1914 been no changes since the device was removed, or if the array has a
1915 write-intent bitmap which has recorded whatever changes there were,
1916 then the device will immediately become a full member of the array and
1917 those differences recorded in the bitmap will be resolved.
1927 MISC mode includes a number of distinct operations that
1928 operate on distinct devices. The operations are:
1931 The device is examined to see if it is
1932 (1) an active md array, or
1933 (2) a component of an md array.
1934 The information discovered is reported.
1938 The device should be an active md device.
1940 will display a detailed description of the array.
1944 will cause the output to be less detailed and the format to be
1945 suitable for inclusion in
1949 will normally be 0 unless
1951 failed to get useful information about the device(s); however, if the
1953 option is given, then the exit status will be:
1957 The array is functioning normally.
1960 The array has at least one failed device.
1963 The array has multiple failed devices such that it is unusable.
1966 There was an error while trying to get information about the device.
1970 .B \-\-detail\-platform
1971 Print detail of the platform's RAID capabilities (firmware / hardware
1972 topology). If the metadata is specified with
1976 then the return status will be:
1980 metadata successfully enumerated its platform components on this system
1983 metadata is platform independent
1986 metadata failed to find its platform components on this system
1990 .B \-\-update\-subarray=
1991 If the device is a container and the argument to \-\-update\-subarray
1992 specifies a subarray in the container, then attempt to update the given
1993 superblock field in the subarray. Similar to updating an array in
1994 "assemble" mode, the field to update is selected by
1998 option. Currently only
2004 option updates the subarray name in the metadata, it may not affect the
2005 device node name or the device node symlink until the subarray is
2006 re\-assembled. If updating
2008 would change the UUID of an active subarray this operation is blocked,
2009 and the command will end in an error.
2013 The device should be a component of an md array.
2015 will read the md superblock of the device and display the contents.
2020 is given, then multiple devices that are components of the one array
2021 are grouped together and reported in a single entry suitable
2027 without listing any devices will cause all devices listed in the
2028 config file to be examined.
2032 The devices should be active md arrays which will be deactivated, as
2033 long as they are not currently in use.
2037 This will fully activate a partially assembled md array.
2041 This will mark an active array as read-only, providing that it is
2042 not currently being used.
2048 array back to being read/write.
2052 For all operations except
2055 will cause the operation to be applied to all arrays listed in
2060 causes all devices listed in the config file to be examined.
2063 .BR \-b ", " \-\-brief
2064 Be less verbose. This is used with
2072 gives an intermediate level of verbosity.
2078 .B mdadm \-\-monitor
2079 .I options... devices...
2084 to periodically poll a number of md arrays and to report on any events
2087 will never exit once it decides that there are arrays to be checked,
2088 so it should normally be run in the background.
2090 As well as reporting events,
2092 may move a spare drive from one array to another if they are in the
2097 and if the destination array has a failed drive but no spares.
2099 If any devices are listed on the command line,
2101 will only monitor those devices. Otherwise all arrays listed in the
2102 configuration file will be monitored. Further, if
2104 is given, then any other md devices that appear in
2106 will also be monitored.
2108 The result of monitoring the arrays is the generation of events.
2109 These events are passed to a separate program (if specified) and may
2110 be mailed to a given E-mail address.
2112 When passing events to a program, the program is run once for each event,
2113 and is given 2 or 3 command-line arguments: the first is the
2114 name of the event (see below), the second is the name of the
2115 md device which is affected, and the third is the name of a related
2116 device if relevant (such as a component device that has failed).
2120 is given, then a program or an E-mail address must be specified on the
2121 command line or in the config file. If neither are available, then
2123 will not monitor anything.
2127 will continue monitoring as long as something was found to monitor. If
2128 no program or email is given, then each event is reported to
2131 The different events are:
2135 .B DeviceDisappeared
2136 An md array which previously was configured appears to no longer be
2137 configured. (syslog priority: Critical)
2141 was told to monitor an array which is RAID0 or Linear, then it will
2143 .B DeviceDisappeared
2144 with the extra information
2146 This is because RAID0 and Linear do not support the device-failed,
2147 hot-spare and resync operations which are monitored.
2151 An md array started reconstruction. (syslog priority: Warning)
2157 is a two-digit number (ie. 05, 48). This indicates that rebuild
2158 has passed that many percent of the total. The events are generated
2159 with fixed increment since 0. Increment size may be specified with
2160 a commandline option (default is 20). (syslog priority: Warning)
2164 An md array that was rebuilding, isn't any more, either because it
2165 finished normally or was aborted. (syslog priority: Warning)
2169 An active component device of an array has been marked as
2170 faulty. (syslog priority: Critical)
2174 A spare component device which was being rebuilt to replace a faulty
2175 device has failed. (syslog priority: Critical)
2179 A spare component device which was being rebuilt to replace a faulty
2180 device has been successfully rebuilt and has been made active.
2181 (syslog priority: Info)
2185 A new md array has been detected in the
2187 file. (syslog priority: Info)
2191 A newly noticed array appears to be degraded. This message is not
2194 notices a drive failure which causes degradation, but only when
2196 notices that an array is degraded when it first sees the array.
2197 (syslog priority: Critical)
2201 A spare drive has been moved from one array in a
2205 to another to allow a failed drive to be replaced.
2206 (syslog priority: Info)
2212 has been told, via the config file, that an array should have a certain
2213 number of spare devices, and
2215 detects that it has fewer than this number when it first sees the
2216 array, it will report a
2219 (syslog priority: Warning)
2223 An array was found at startup, and the
2226 (syslog priority: Info)
2236 cause Email to be sent. All events cause the program to be run.
2237 The program is run with two or three arguments: the event
2238 name, the array device and possibly a second device.
2240 Each event has an associated array device (e.g.
2242 and possibly a second device. For
2247 the second device is the relevant component device.
2250 the second device is the array that the spare was moved from.
2254 to move spares from one array to another, the different arrays need to
2255 be labeled with the same
2257 or the spares must be allowed to migrate through matching POLICY domains
2258 in the configuration file. The
2260 name can be any string; it is only necessary that different spare
2261 groups use different names.
2265 detects that an array in a spare group has fewer active
2266 devices than necessary for the complete array, and has no spare
2267 devices, it will look for another array in the same spare group that
2268 has a full complement of working drive and a spare. It will then
2269 attempt to remove the spare from the second drive and add it to the
2271 If the removal succeeds but the adding fails, then it is added back to
2274 If the spare group for a degraded array is not defined,
2276 will look at the rules of spare migration specified by POLICY lines in
2278 and then follow similar steps as above if a matching spare is found.
2281 The GROW mode is used for changing the size or shape of an active
2283 For this to work, the kernel must support the necessary change.
2284 Various types of growth are being added during 2.6 development.
2286 Currently the supported changes include
2288 change the "size" attribute for RAID1, RAID4, RAID5 and RAID6.
2290 increase or decrease the "raid\-devices" attribute of RAID0, RAID1, RAID4,
2293 change the chunk-size and layout of RAID0, RAID4, RAID5 and RAID6.
2295 convert between RAID1 and RAID5, between RAID5 and RAID6, between
2296 RAID0, RAID4, and RAID5, and between RAID0 and RAID10 (in the near-2 mode).
2298 add a write-intent bitmap to any array which supports these bitmaps, or
2299 remove a write-intent bitmap from such an array.
2302 Using GROW on containers is currently supported only for Intel's IMSM
2303 container format. The number of devices in a container can be
2304 increased - which affects all arrays in the container - or an array
2305 in a container can be converted between levels where those levels are
2306 supported by the container, and the conversion is on of those listed
2307 above. Resizing arrays in an IMSM container with
2309 is not yet supported.
2311 Grow functionality (e.g. expand a number of raid devices) for Intel's
2312 IMSM container format has an experimental status. It is guarded by the
2313 .B MDADM_EXPERIMENTAL
2314 environment variable which must be set to '1' for a GROW command to
2316 This is for the following reasons:
2319 Intel's native IMSM check-pointing is not fully tested yet.
2320 This can causes IMSM incompatibility during the grow process: an array
2321 which is growing cannot roam between Microsoft Windows(R) and Linux
2325 Interrupting a grow operation is not recommended, because it
2326 has not been fully tested for Intel's IMSM container format yet.
2329 Note: Intel's native checkpointing doesn't use
2331 option and it is transparent for assembly feature.
2334 Normally when an array is built the "size" is taken from the smallest
2335 of the drives. If all the small drives in an arrays are, one at a
2336 time, removed and replaced with larger drives, then you could have an
2337 array of large drives with only a small amount used. In this
2338 situation, changing the "size" with "GROW" mode will allow the extra
2339 space to start being used. If the size is increased in this way, a
2340 "resync" process will start to make sure the new parts of the array
2343 Note that when an array changes size, any filesystem that may be
2344 stored in the array will not automatically grow or shrink to use or
2345 vacate the space. The
2346 filesystem will need to be explicitly told to use the extra space
2347 after growing, or to reduce its size
2349 to shrinking the array.
2351 Also the size of an array cannot be changed while it has an active
2352 bitmap. If an array has a bitmap, it must be removed before the size
2353 can be changed. Once the change is complete a new bitmap can be created.
2355 .SS RAID\-DEVICES CHANGES
2357 A RAID1 array can work with any number of devices from 1 upwards
2358 (though 1 is not very useful). There may be times which you want to
2359 increase or decrease the number of active devices. Note that this is
2360 different to hot-add or hot-remove which changes the number of
2363 When reducing the number of devices in a RAID1 array, the slots which
2364 are to be removed from the array must already be vacant. That is, the
2365 devices which were in those slots must be failed and removed.
2367 When the number of devices is increased, any hot spares that are
2368 present will be activated immediately.
2370 Changing the number of active devices in a RAID5 or RAID6 is much more
2371 effort. Every block in the array will need to be read and written
2372 back to a new location. From 2.6.17, the Linux Kernel is able to
2373 increase the number of devices in a RAID5 safely, including restarting
2374 an interrupted "reshape". From 2.6.31, the Linux Kernel is able to
2375 increase or decrease the number of devices in a RAID5 or RAID6.
2377 From 2.6.35, the Linux Kernel is able to convert a RAID0 in to a RAID4
2380 uses this functionality and the ability to add
2381 devices to a RAID4 to allow devices to be added to a RAID0. When
2382 requested to do this,
2384 will convert the RAID0 to a RAID4, add the necessary disks and make
2385 the reshape happen, and then convert the RAID4 back to RAID0.
2387 When decreasing the number of devices, the size of the array will also
2388 decrease. If there was data in the array, it could get destroyed and
2389 this is not reversible, so you should firstly shrink the filesystem on
2390 the array to fit within the new size. To help prevent accidents,
2392 requires that the size of the array be decreased first with
2393 .BR "mdadm --grow --array-size" .
2394 This is a reversible change which simply makes the end of the array
2395 inaccessible. The integrity of any data can then be checked before
2396 the non-reversible reduction in the number of devices is request.
2398 When relocating the first few stripes on a RAID5 or RAID6, it is not
2399 possible to keep the data on disk completely consistent and
2400 crash-proof. To provide the required safety, mdadm disables writes to
2401 the array while this "critical section" is reshaped, and takes a
2402 backup of the data that is in that section. For grows, this backup may be
2403 stored in any spare devices that the array has, however it can also be
2404 stored in a separate file specified with the
2406 option, and is required to be specified for shrinks, RAID level
2407 changes and layout changes. If this option is used, and the system
2408 does crash during the critical period, the same file must be passed to
2410 to restore the backup and reassemble the array. When shrinking rather
2411 than growing the array, the reshape is done from the end towards the
2412 beginning, so the "critical section" is at the end of the reshape.
2416 Changing the RAID level of any array happens instantaneously. However
2417 in the RAID5 to RAID6 case this requires a non-standard layout of the
2418 RAID6 data, and in the RAID6 to RAID5 case that non-standard layout is
2419 required before the change can be accomplished. So while the level
2420 change is instant, the accompanying layout change can take quite a
2423 is required. If the array is not simultaneously being grown or
2424 shrunk, so that the array size will remain the same - for example,
2425 reshaping a 3-drive RAID5 into a 4-drive RAID6 - the backup file will
2426 be used not just for a "cricital section" but throughout the reshape
2427 operation, as described below under LAYOUT CHANGES.
2429 .SS CHUNK-SIZE AND LAYOUT CHANGES
2431 Changing the chunk-size of layout without also changing the number of
2432 devices as the same time will involve re-writing all blocks in-place.
2433 To ensure against data loss in the case of a crash, a
2435 must be provided for these changes. Small sections of the array will
2436 be copied to the backup file while they are being rearranged. This
2437 means that all the data is copied twice, once to the backup and once
2438 to the new layout on the array, so this type of reshape will go very
2441 If the reshape is interrupted for any reason, this backup file must be
2443 .B "mdadm --assemble"
2444 so the array can be reassembled. Consequently the file cannot be
2445 stored on the device being reshaped.
2450 A write-intent bitmap can be added to, or removed from, an active
2451 array. Either internal bitmaps, or bitmaps stored in a separate file,
2452 can be added. Note that if you add a bitmap stored in a file which is
2453 in a filesystem that is on the RAID array being affected, the system
2454 will deadlock. The bitmap must be on a separate filesystem.
2456 .SH INCREMENTAL MODE
2460 .B mdadm \-\-incremental
2466 .B mdadm \-\-incremental \-\-fail
2470 .B mdadm \-\-incremental \-\-rebuild\-map
2473 .B mdadm \-\-incremental \-\-run \-\-scan
2476 This mode is designed to be used in conjunction with a device
2477 discovery system. As devices are found in a system, they can be
2479 .B "mdadm \-\-incremental"
2480 to be conditionally added to an appropriate array.
2482 Conversely, it can also be used with the
2484 flag to do just the opposite and find whatever array a particular device
2485 is part of and remove the device from that array.
2487 If the device passed is a
2489 device created by a previous call to
2491 then rather than trying to add that device to an array, all the arrays
2492 described by the metadata of the container will be started.
2495 performs a number of tests to determine if the device is part of an
2496 array, and which array it should be part of. If an appropriate array
2497 is found, or can be created,
2499 adds the device to the array and conditionally starts the array.
2503 will normally only add devices to an array which were previously working
2504 (active or spare) parts of that array. The support for automatic
2505 inclusion of a new drive as a spare in some array requires
2506 a configuration through POLICY in config file.
2510 makes are as follow:
2512 Is the device permitted by
2514 That is, is it listed in a
2516 line in that file. If
2518 is absent then the default it to allow any device. Similar if
2520 contains the special word
2522 then any device is allowed. Otherwise the device name given to
2524 must match one of the names or patterns in a
2529 Does the device have a valid md superblock? If a specific metadata
2530 version is requested with
2534 then only that style of metadata is accepted, otherwise
2536 finds any known version of metadata. If no
2538 metadata is found, the device may be still added to an array
2539 as a spare if POLICY allows.
2543 Does the metadata match an expected array?
2544 The metadata can match in two ways. Either there is an array listed
2547 which identifies the array (either by UUID, by name, by device list,
2548 or by minor-number), or the array was created with a
2554 or on the command line.
2557 is not able to positively identify the array as belonging to the
2558 current host, the device will be rejected.
2563 keeps a list of arrays that it has partially assembled in
2565 If no array exists which matches
2566 the metadata on the new device,
2568 must choose a device name and unit number. It does this based on any
2571 or any name information stored in the metadata. If this name
2572 suggests a unit number, that number will be used, otherwise a free
2573 unit number will be chosen. Normally
2575 will prefer to create a partitionable array, however if the
2579 suggests that a non-partitionable array is preferred, that will be
2582 If the array is not found in the config file and its metadata does not
2583 identify it as belonging to the "homehost", then
2585 will choose a name for the array which is certain not to conflict with
2586 any array which does belong to this host. It does this be adding an
2587 underscore and a small number to the name preferred by the metadata.
2589 Once an appropriate array is found or created and the device is added,
2591 must decide if the array is ready to be started. It will
2592 normally compare the number of available (non-spare) devices to the
2593 number of devices that the metadata suggests need to be active. If
2594 there are at least that many, the array will be started. This means
2595 that if any devices are missing the array will not be restarted.
2601 in which case the array will be run as soon as there are enough
2602 devices present for the data to be accessible. For a RAID1, that
2603 means one device will start the array. For a clean RAID5, the array
2604 will be started as soon as all but one drive is present.
2606 Note that neither of these approaches is really ideal. If it can
2607 be known that all device discovery has completed, then
2611 can be run which will try to start all arrays that are being
2612 incrementally assembled. They are started in "read-auto" mode in
2613 which they are read-only until the first write request. This means
2614 that no metadata updates are made and no attempt at resync or recovery
2615 happens. Further devices that are found before the first write can
2616 still be added safely.
2619 This section describes environment variables that affect how mdadm
2624 Setting this value to 1 will prevent mdadm from automatically launching
2625 mdmon. This variable is intended primarily for debugging mdadm/mdmon.
2631 does not create any device nodes in /dev, but leaves that task to
2635 appears not to be configured, or if this environment variable is set
2638 will create and devices that are needed.
2642 .B " mdadm \-\-query /dev/name-of-device"
2644 This will find out if a given device is a RAID array, or is part of
2645 one, and will provide brief information about the device.
2647 .B " mdadm \-\-assemble \-\-scan"
2649 This will assemble and start all arrays listed in the standard config
2650 file. This command will typically go in a system startup file.
2652 .B " mdadm \-\-stop \-\-scan"
2654 This will shut down all arrays that can be shut down (i.e. are not
2655 currently in use). This will typically go in a system shutdown script.
2657 .B " mdadm \-\-follow \-\-scan \-\-delay=120"
2659 If (and only if) there is an Email address or program given in the
2660 standard config file, then
2661 monitor the status of all arrays listed in that file by
2662 polling them ever 2 minutes.
2664 .B " mdadm \-\-create /dev/md0 \-\-level=1 \-\-raid\-devices=2 /dev/hd[ac]1"
2666 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
2669 .B " echo 'DEVICE /dev/hd*[0\-9] /dev/sd*[0\-9]' > mdadm.conf"
2671 .B " mdadm \-\-detail \-\-scan >> mdadm.conf"
2673 This will create a prototype config file that describes currently
2674 active arrays that are known to be made from partitions of IDE or SCSI drives.
2675 This file should be reviewed before being used as it may
2676 contain unwanted detail.
2678 .B " echo 'DEVICE /dev/hd[a\-z] /dev/sd*[a\-z]' > mdadm.conf"
2680 .B " mdadm \-\-examine \-\-scan \-\-config=mdadm.conf >> mdadm.conf"
2682 This will find arrays which could be assembled from existing IDE and
2683 SCSI whole drives (not partitions), and store the information in the
2684 format of a config file.
2685 This file is very likely to contain unwanted detail, particularly
2688 entries. It should be reviewed and edited before being used as an
2691 .B " mdadm \-\-examine \-\-brief \-\-scan \-\-config=partitions"
2693 .B " mdadm \-Ebsc partitions"
2695 Create a list of devices by reading
2696 .BR /proc/partitions ,
2697 scan these for RAID superblocks, and printout a brief listing of all
2700 .B " mdadm \-Ac partitions \-m 0 /dev/md0"
2702 Scan all partitions and devices listed in
2703 .BR /proc/partitions
2706 out of all such devices with a RAID superblock with a minor number of 0.
2708 .B " mdadm \-\-monitor \-\-scan \-\-daemonise > /run/mdadm/mon.pid"
2710 If config file contains a mail address or alert program, run mdadm in
2711 the background in monitor mode monitoring all md devices. Also write
2712 pid of mdadm daemon to
2713 .BR /run/mdadm/mon.pid .
2715 .B " mdadm \-Iq /dev/somedevice"
2717 Try to incorporate newly discovered device into some array as
2720 .B " mdadm \-\-incremental \-\-rebuild\-map \-\-run \-\-scan"
2722 Rebuild the array map from any current arrays, and then start any that
2725 .B " mdadm /dev/md4 --fail detached --remove detached"
2727 Any devices which are components of /dev/md4 will be marked as faulty
2728 and then remove from the array.
2730 .B " mdadm --grow /dev/md4 --level=6 --backup-file=/root/backup-md4"
2734 which is currently a RAID5 array will be converted to RAID6. There
2735 should normally already be a spare drive attached to the array as a
2736 RAID6 needs one more drive than a matching RAID5.
2738 .B " mdadm --create /dev/md/ddf --metadata=ddf --raid-disks 6 /dev/sd[a-f]"
2740 Create a DDF array over 6 devices.
2742 .B " mdadm --create /dev/md/home -n3 -l5 -z 30000000 /dev/md/ddf"
2744 Create a RAID5 array over any 3 devices in the given DDF set. Use
2745 only 30 gigabytes of each device.
2747 .B " mdadm -A /dev/md/ddf1 /dev/sd[a-f]"
2749 Assemble a pre-exist ddf array.
2751 .B " mdadm -I /dev/md/ddf1"
2753 Assemble all arrays contained in the ddf array, assigning names as
2756 .B " mdadm \-\-create \-\-help"
2758 Provide help about the Create mode.
2760 .B " mdadm \-\-config \-\-help"
2762 Provide help about the format of the config file.
2764 .B " mdadm \-\-help"
2766 Provide general help.
2776 lists all active md devices with information about them.
2778 uses this to find arrays when
2780 is given in Misc mode, and to monitor array reconstruction
2785 The config file lists which devices may be scanned to see if
2786 they contain MD super block, and gives identifying information
2787 (e.g. UUID) about known MD arrays. See
2794 mode is used, this file gets a list of arrays currently being created.
2799 understand two sorts of names for array devices.
2801 The first is the so-called 'standard' format name, which matches the
2802 names used by the kernel and which appear in
2805 The second sort can be freely chosen, but must reside in
2807 When giving a device name to
2809 to create or assemble an array, either full path name such as
2813 can be given, or just the suffix of the second sort of name, such as
2819 chooses device names during auto-assembly or incremental assembly, it
2820 will sometimes add a small sequence number to the end of the name to
2821 avoid conflicted between multiple arrays that have the same name. If
2823 can reasonably determine that the array really is meant for this host,
2824 either by a hostname in the metadata, or by the presence of the array
2827 then it will leave off the suffix if possible.
2828 Also if the homehost is specified as
2831 will only use a suffix if a different array of the same name already
2832 exists or is listed in the config file.
2834 The standard names for non-partitioned arrays (the only sort of md
2835 array available in 2.4 and earlier) are of the form
2839 where NN is a number.
2840 The standard names for partitionable arrays (as available from 2.6
2841 onwards) are of the form
2845 Partition numbers should be indicated by added "pMM" to these, thus "/dev/md/d1p2".
2847 From kernel version, 2.6.28 the "non-partitioned array" can actually
2848 be partitioned. So the "md_dNN" names are no longer needed, and
2849 partitions such as "/dev/mdNNpXX" are possible.
2853 was previously known as
2857 is completely separate from the
2859 package, and does not use the
2861 configuration file at all.
2864 For further information on mdadm usage, MD and the various levels of
2867 .B http://raid.wiki.kernel.org/
2869 (based upon Jakob \(/Ostergaard's Software\-RAID.HOWTO)
2871 .\"for new releases of the RAID driver check out:
2874 .\".UR ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2875 .\"ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2880 .\".UR http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2881 .\"http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2884 The latest version of
2886 should always be available from
2888 .B http://www.kernel.org/pub/linux/utils/raid/mdadm/