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 RAID
126 levels 1/4/5/6, changing the RAID level between 1, 5, and 6, changing
127 the chunk size and layout for RAID5 and RAID5, as well as adding or
128 removing a write-intent bitmap.
131 .B "Incremental Assembly"
132 Add a single device to an appropriate array. If the addition of the
133 device makes the array runnable, the array will be started.
134 This provides a convenient interface to a
136 system. As each device is detected,
138 has a chance to include it in some array as appropriate.
141 flag is passed in we will remove the device from any active array
142 instead of adding it.
148 in this mode, then any arrays within that container will be assembled
153 This is for doing things to specific components of an array such as
154 adding new spares and removing faulty devices.
158 This is an 'everything else' mode that supports operations on active
159 arrays, operations on component devices such as erasing old superblocks, and
160 information gathering operations.
161 .\"This mode allows operations on independent devices such as examine MD
162 .\"superblocks, erasing old superblocks and stopping active arrays.
166 This mode does not act on a specific device or array, but rather it
167 requests the Linux Kernel to activate any auto-detected arrays.
170 .SH Options for selecting a mode are:
173 .BR \-A ", " \-\-assemble
174 Assemble a pre-existing array.
177 .BR \-B ", " \-\-build
178 Build a legacy array without superblocks.
181 .BR \-C ", " \-\-create
185 .BR \-F ", " \-\-follow ", " \-\-monitor
191 .BR \-G ", " \-\-grow
192 Change the size or shape of an active array.
195 .BR \-I ", " \-\-incremental
196 Add/remove a single device to/from an appropriate array, and possibly start the array.
200 Request that the kernel starts any auto-detected arrays. This can only
203 is compiled into the kernel \(em not if it is a module.
204 Arrays can be auto-detected by the kernel if all the components are in
205 primary MS-DOS partitions with partition type
207 and all use v0.90 metadata.
208 In-kernel autodetect is not recommended for new installations. Using
210 to detect and assemble arrays \(em possibly in an
212 \(em is substantially more flexible and should be preferred.
215 If a device is given before any options, or if the first option is
220 then the MANAGE mode is assumed.
221 Anything other than these will cause the
225 .SH Options that are not mode-specific are:
228 .BR \-h ", " \-\-help
229 Display general help message or, after one of the above options, a
230 mode-specific help message.
234 Display more detailed help about command line parsing and some commonly
238 .BR \-V ", " \-\-version
239 Print version information for mdadm.
242 .BR \-v ", " \-\-verbose
243 Be more verbose about what is happening. This can be used twice to be
245 The extra verbosity currently only affects
246 .B \-\-detail \-\-scan
248 .BR "\-\-examine \-\-scan" .
251 .BR \-q ", " \-\-quiet
252 Avoid printing purely informative messages. With this,
254 will be silent unless there is something really important to report.
257 .BR \-f ", " \-\-force
258 Be more forceful about certain operations. See the various modes for
259 the exact meaning of this option in different contexts.
262 .BR \-c ", " \-\-config=
263 Specify the config file. Default is to use
264 .BR /etc/mdadm.conf ,
265 or if that is missing then
266 .BR /etc/mdadm/mdadm.conf .
267 If the config file given is
269 then nothing will be read, but
271 will act as though the config file contained exactly
272 .B "DEVICE partitions containers"
275 to find a list of devices to scan, and
277 to find a list of containers to examine.
280 is given for the config file, then
282 will act as though the config file were empty.
285 .BR \-s ", " \-\-scan
288 for missing information.
289 In general, this option gives
291 permission to get any missing information (like component devices,
292 array devices, array identities, and alert destination) from the
293 configuration file (see previous option);
294 one exception is MISC mode when using
300 says to get a list of array devices from
304 .BR \-e ", " \-\-metadata=
305 Declare the style of RAID metadata (superblock) to be used. The
306 default is {DEFAULT_METADATA} for
308 and to guess for other operations.
309 The default can be overridden by setting the
318 .ie '{DEFAULT_METADATA}'0.90'
319 .IP "0, 0.90, default"
323 Use the original 0.90 format superblock. This format limits arrays to
324 28 component devices and limits component devices of levels 1 and
325 greater to 2 terabytes.
326 .ie '{DEFAULT_METADATA}'0.90'
327 .IP "1, 1.0, 1.1, 1.2"
329 .IP "1, 1.0, 1.1, 1.2 default"
331 Use the new version-1 format superblock. This has few restrictions.
332 The different sub-versions store the superblock at different locations
333 on the device, either at the end (for 1.0), at the start (for 1.1) or
334 4K from the start (for 1.2). "1" is equivalent to "1.0".
335 'if '{DEFAULT_METADATA}'1.2' "default" is equivalent to "1.2".
337 Use the "Industry Standard" DDF (Disk Data Format) format defined by
339 When creating a DDF array a
341 will be created, and normal arrays can be created in that container.
343 Use the Intel(R) Matrix Storage Manager metadata format. This creates a
345 which is managed in a similar manner to DDF, and is supported by an
346 option-rom on some platforms:
348 .B http://www.intel.com/design/chipsets/matrixstorage_sb.htm
354 This will override any
356 setting in the config file and provides the identity of the host which
357 should be considered the home for any arrays.
359 When creating an array, the
361 will be recorded in the metadata. For version-1 superblocks, it will
362 be prefixed to the array name. For version-0.90 superblocks, part of
363 the SHA1 hash of the hostname will be stored in the later half of the
366 When reporting information about an array, any array which is tagged
367 for the given homehost will be reported as such.
369 When using Auto-Assemble, only arrays tagged for the given homehost
370 will be allowed to use 'local' names (i.e. not ending in '_' followed
371 by a digit string). See below under
372 .BR "Auto Assembly" .
374 .SH For create, build, or grow:
377 .BR \-n ", " \-\-raid\-devices=
378 Specify the number of active devices in the array. This, plus the
379 number of spare devices (see below) must equal the number of
381 (including "\fBmissing\fP" devices)
382 that are listed on the command line for
384 Setting a value of 1 is probably
385 a mistake and so requires that
387 be specified first. A value of 1 will then be allowed for linear,
388 multipath, RAID0 and RAID1. It is never allowed for RAID4, RAID5 or RAID6.
390 This number can only be changed using
392 for RAID1, RAID4, RAID5 and RAID6 arrays, and only on kernels which provide
393 the necessary support.
396 .BR \-x ", " \-\-spare\-devices=
397 Specify the number of spare (eXtra) devices in the initial array.
398 Spares can also be added
399 and removed later. The number of component devices listed
400 on the command line must equal the number of RAID devices plus the
401 number of spare devices.
404 .BR \-z ", " \-\-size=
405 Amount (in Kibibytes) of space to use from each drive in RAID levels 1/4/5/6.
406 This must be a multiple of the chunk size, and must leave about 128Kb
407 of space at the end of the drive for the RAID superblock.
408 If this is not specified
409 (as it normally is not) the smallest drive (or partition) sets the
410 size, though if there is a variance among the drives of greater than 1%, a warning is
413 This value can be set with
415 for RAID level 1/4/5/6. If the array was created with a size smaller
416 than the currently active drives, the extra space can be accessed
419 The size can be given as
421 which means to choose the largest size that fits on all current drives.
423 This value can not be used with
425 metadata such as DDF and IMSM.
428 .BR \-Z ", " \-\-array-size=
429 This is only meaningful with
431 and its effect is not persistent: when the array is stopped an
432 restarted the default array size will be restored.
434 Setting the array-size causes the array to appear smaller to programs
435 that access the data. This is particularly needed before reshaping an
436 array so that it will be smaller. As the reshape is not reversible,
437 but setting the size with
439 is, it is required that the array size is reduced as appropriate
440 before the number of devices in the array is reduced.
443 .BR \-c ", " \-\-chunk=
444 Specify chunk size of kibibytes. The default when creating an
445 array is 512KB. To ensure compatibility with earlier versions, the
446 default when Building and array with no persistent metadata is 64KB.
447 This is only meaningful for RAID0, RAID4, RAID5, RAID6, and RAID10.
451 Specify rounding factor for a Linear array. The size of each
452 component will be rounded down to a multiple of this size.
453 This is a synonym for
455 but highlights the different meaning for Linear as compared to other
456 RAID levels. The default is 64K if a kernel earlier than 2.6.16 is in
457 use, and is 0K (i.e. no rounding) in later kernels.
460 .BR \-l ", " \-\-level=
461 Set RAID level. When used with
463 options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
464 raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty, container.
465 Obviously some of these are synonymous.
469 metadata type is requested, only the
471 level is permitted, and it does not need to be explicitly given.
475 only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid.
479 to change the RAID level in some cases. See LEVEL CHANGES below.
482 .BR \-p ", " \-\-layout=
483 This option configures the fine details of data layout for RAID5, RAID6,
484 and RAID10 arrays, and controls the failure modes for
487 The layout of the RAID5 parity block can be one of
488 .BR left\-asymmetric ,
489 .BR left\-symmetric ,
490 .BR right\-asymmetric ,
491 .BR right\-symmetric ,
492 .BR la ", " ra ", " ls ", " rs .
494 .BR left\-symmetric .
496 It is also possibly to cause RAID5 to use a RAID4-like layout by
502 Finally for RAID5 there are DDF\-compatible layouts,
503 .BR ddf\-zero\-restart ,
504 .BR ddf\-N\-restart ,
506 .BR ddf\-N\-continue .
508 These same layouts are available for RAID6. There are also 4 layouts
509 that will provide an intermediate stage for converting between RAID5
510 and RAID6. These provide a layout which is identical to the
511 corresponding RAID5 layout on the first N\-1 devices, and has the 'Q'
512 syndrome (the second 'parity' block used by RAID6) on the last device.
514 .BR left\-symmetric\-6 ,
515 .BR right\-symmetric\-6 ,
516 .BR left\-asymmetric\-6 ,
517 .BR right\-asymmetric\-6 ,
519 .BR parity\-first\-6 .
521 When setting the failure mode for level
524 .BR write\-transient ", " wt ,
525 .BR read\-transient ", " rt ,
526 .BR write\-persistent ", " wp ,
527 .BR read\-persistent ", " rp ,
529 .BR read\-fixable ", " rf ,
530 .BR clear ", " flush ", " none .
532 Each failure mode can be followed by a number, which is used as a period
533 between fault generation. Without a number, the fault is generated
534 once on the first relevant request. With a number, the fault will be
535 generated after that many requests, and will continue to be generated
536 every time the period elapses.
538 Multiple failure modes can be current simultaneously by using the
540 option to set subsequent failure modes.
542 "clear" or "none" will remove any pending or periodic failure modes,
543 and "flush" will clear any persistent faults.
545 Finally, the layout options for RAID10 are one of 'n', 'o' or 'f' followed
546 by a small number. The default is 'n2'. The supported options are:
549 signals 'near' copies. Multiple copies of one data block are at
550 similar offsets in different devices.
553 signals 'offset' copies. Rather than the chunks being duplicated
554 within a stripe, whole stripes are duplicated but are rotated by one
555 device so duplicate blocks are on different devices. Thus subsequent
556 copies of a block are in the next drive, and are one chunk further
561 (multiple copies have very different offsets).
562 See md(4) for more detail about 'near', 'offset', and 'far'.
564 The number is the number of copies of each datablock. 2 is normal, 3
565 can be useful. This number can be at most equal to the number of
566 devices in the array. It does not need to divide evenly into that
567 number (e.g. it is perfectly legal to have an 'n2' layout for an array
568 with an odd number of devices).
570 When an array is converted between RAID5 and RAID6 an intermediate
571 RAID6 layout is used in which the second parity block (Q) is always on
572 the last device. To convert a RAID5 to RAID6 and leave it in this new
573 layout (which does not require re-striping) use
574 .BR \-\-layout=preserve .
575 This will try to avoid any restriping.
577 The converse of this is
578 .B \-\-layout=normalise
579 which will change a non-standard RAID6 layout into a more standard
586 (thus explaining the p of
590 .BR \-b ", " \-\-bitmap=
591 Specify a file to store a write-intent bitmap in. The file should not
594 is also given. The same file should be provided
595 when assembling the array. If the word
597 is given, then the bitmap is stored with the metadata on the array,
598 and so is replicated on all devices. If the word
602 mode, then any bitmap that is present is removed.
604 To help catch typing errors, the filename must contain at least one
605 slash ('/') if it is a real file (not 'internal' or 'none').
607 Note: external bitmaps are only known to work on ext2 and ext3.
608 Storing bitmap files on other filesystems may result in serious problems.
611 .BR \-\-bitmap\-chunk=
612 Set the chunksize of the bitmap. Each bit corresponds to that many
613 Kilobytes of storage.
614 When using a file based bitmap, the default is to use the smallest
615 size that is at-least 4 and requires no more than 2^21 chunks.
618 bitmap, the chunksize defaults to 64Meg, or larger if necessary to
619 fit the bitmap into the available space.
622 .BR \-W ", " \-\-write\-mostly
623 subsequent devices listed in a
628 command will be flagged as 'write-mostly'. This is valid for RAID1
629 only and means that the 'md' driver will avoid reading from these
630 devices if at all possible. This can be useful if mirroring over a
634 .BR \-\-write\-behind=
635 Specify that write-behind mode should be enabled (valid for RAID1
636 only). If an argument is specified, it will set the maximum number
637 of outstanding writes allowed. The default value is 256.
638 A write-intent bitmap is required in order to use write-behind
639 mode, and write-behind is only attempted on drives marked as
643 .BR \-\-assume\-clean
646 that the array pre-existed and is known to be clean. It can be useful
647 when trying to recover from a major failure as you can be sure that no
648 data will be affected unless you actually write to the array. It can
649 also be used when creating a RAID1 or RAID10 if you want to avoid the
650 initial resync, however this practice \(em while normally safe \(em is not
651 recommended. Use this only if you really know what you are doing.
653 When the devices that will be part of a new array were filled
654 with zeros before creation the operator knows the array is
655 actually clean. If that is the case, such as after running
656 badblocks, this argument can be used to tell mdadm the
657 facts the operator knows.
660 .BR \-\-backup\-file=
663 is used to increase the number of
664 raid-devices in a RAID5 if there are no spare devices available.
665 See the GROW MODE section below on RAID\-DEVICES CHANGES. The file
666 should be stored on a separate device, not on the RAID array being
670 .BR \-\-array-size= ", " \-Z
671 Set the size of the array which is seen by users of the device such as
672 filesystems. This can be less that the real size, but never greater.
673 The size set this way does not persist across restarts of the array.
675 This is most useful when reducing the number of devices in a RAID5 or
676 RAID6. Such arrays require the array-size to be reduced before a
677 reshape can be performed that reduces the real size.
681 restores the apparent size of the array to be whatever the real
682 amount of available space is.
685 .BR \-N ", " \-\-name=
688 for the array. This is currently only effective when creating an
689 array with a version-1 superblock, or an array in a DDF container.
690 The name is a simple textual string that can be used to identify array
691 components when assembling. If name is needed but not specified, it
692 is taken from the basename of the device that is being created.
704 run the array, even if some of the components
705 appear to be active in another array or filesystem. Normally
707 will ask for confirmation before including such components in an
708 array. This option causes that question to be suppressed.
711 .BR \-f ", " \-\-force
714 accept the geometry and layout specified without question. Normally
716 will not allow creation of an array with only one device, and will try
717 to create a RAID5 array with one missing drive (as this makes the
718 initial resync work faster). With
721 will not try to be so clever.
724 .BR \-a ", " "\-\-auto{=yes,md,mdp,part,p}{NN}"
725 Instruct mdadm how to create the device file if needed, possibly allocating
726 an unused minor number. "md" causes a non-partitionable array
727 to be used (though since Linux 2.6.28, these array devices are in fact
728 partitionable). "mdp", "part" or "p" causes a partitionable array (2.6 and
729 later) to be used. "yes" requires the named md device to have
730 a 'standard' format, and the type and minor number will be determined
731 from this. With mdadm 3.0, device creation is normally left up to
733 so this option is unlikely to be needed.
734 See DEVICE NAMES below.
736 The argument can also come immediately after
741 is not given on the command line or in the config file, then
747 is also given, then any
749 entries in the config file will override the
751 instruction given on the command line.
753 For partitionable arrays,
755 will create the device file for the whole array and for the first 4
756 partitions. A different number of partitions can be specified at the
757 end of this option (e.g.
759 If the device name ends with a digit, the partition names add a 'p',
761 .IR /dev/md/home1p3 .
762 If there is no trailing digit, then the partition names just have a
764 .IR /dev/md/scratch3 .
766 If the md device name is in a 'standard' format as described in DEVICE
767 NAMES, then it will be created, if necessary, with the appropriate
768 device number based on that name. If the device name is not in one of these
769 formats, then a unused device number will be allocated. The device
770 number will be considered unused if there is no active array for that
771 number, and there is no entry in /dev for that number and with a
772 non-standard name. Names that are not in 'standard' format are only
773 allowed in "/dev/md/".
777 .\".BR \-\-symlink = no
782 .\"to create devices in
784 .\"it will also create symlinks from
786 .\"with names starting with
792 .\"to suppress this, or
793 .\".B \-\-symlink=yes
794 .\"to enforce this even if it is suppressing
802 .BR \-u ", " \-\-uuid=
803 uuid of array to assemble. Devices which don't have this uuid are
807 .BR \-m ", " \-\-super\-minor=
808 Minor number of device that array was created for. Devices which
809 don't have this minor number are excluded. If you create an array as
810 /dev/md1, then all superblocks will contain the minor number 1, even if
811 the array is later assembled as /dev/md2.
813 Giving the literal word "dev" for
817 to use the minor number of the md device that is being assembled.
820 .B \-\-super\-minor=dev
821 will look for super blocks with a minor number of 0.
824 is only relevant for v0.90 metadata, and should not normally be used.
830 .BR \-N ", " \-\-name=
831 Specify the name of the array to assemble. This must be the name
832 that was specified when creating the array. It must either match
833 the name stored in the superblock exactly, or it must match
836 prefixed to the start of the given name.
839 .BR \-f ", " \-\-force
840 Assemble the array even if the metadata on some devices appears to be
843 cannot find enough working devices to start the array, but can find
844 some devices that are recorded as having failed, then it will mark
845 those devices as working so that the array can be started.
846 An array which requires
848 to be started may contain data corruption. Use it carefully.
852 Attempt to start the array even if fewer drives were given than were
853 present last time the array was active. Normally if not all the
854 expected drives are found and
856 is not used, then the array will be assembled but not started.
859 an attempt will be made to start it anyway.
863 This is the reverse of
865 in that it inhibits the startup of array unless all expected drives
866 are present. This is only needed with
868 and can be used if the physical connections to devices are
869 not as reliable as you would like.
872 .BR \-a ", " "\-\-auto{=no,yes,md,mdp,part}"
873 See this option under Create and Build options.
876 .BR \-b ", " \-\-bitmap=
877 Specify the bitmap file that was given when the array was created. If
880 bitmap, there is no need to specify this when assembling the array.
883 .BR \-\-backup\-file=
886 was used to grow the number of raid-devices in a RAID5, and the system
887 crashed during the critical section, then the same
891 to allow possibly corrupted data to be restored.
894 .BR \-U ", " \-\-update=
895 Update the superblock on each device while assembling the array. The
896 argument given to this flag can be one of
910 option will adjust the superblock of an array what was created on a Sparc
911 machine running a patched 2.2 Linux kernel. This kernel got the
912 alignment of part of the superblock wrong. You can use the
913 .B "\-\-examine \-\-sparc2.2"
916 to see what effect this would have.
920 option will update the
922 field on each superblock to match the minor number of the array being
924 This can be useful if
926 reports a different "Preferred Minor" to
928 In some cases this update will be performed automatically
929 by the kernel driver. In particular the update happens automatically
930 at the first write to an array with redundancy (RAID level 1 or
931 greater) on a 2.6 (or later) kernel.
935 option will change the uuid of the array. If a UUID is given with the
937 option that UUID will be used as a new UUID and will
939 be used to help identify the devices in the array.
942 is given, a random UUID is chosen.
946 option will change the
948 of the array as stored in the superblock. This is only supported for
949 version-1 superblocks.
953 option will change the
955 as recorded in the superblock. For version-0 superblocks, this is the
956 same as updating the UUID.
957 For version-1 superblocks, this involves updating the name.
961 option will cause the array to be marked
963 meaning that any redundancy in the array (e.g. parity for RAID5,
964 copies for RAID1) may be incorrect. This will cause the RAID system
965 to perform a "resync" pass to make sure that all redundant information
970 option allows arrays to be moved between machines with different
972 When assembling such an array for the first time after a move, giving
973 .B "\-\-update=byteorder"
976 to expect superblocks to have their byteorder reversed, and will
977 correct that order before assembling the array. This is only valid
978 with original (Version 0.90) superblocks.
982 option will correct the summaries in the superblock. That is the
983 counts of total, working, active, failed, and spare devices.
987 will rarely be of use. It applies to version 1.1 and 1.2 metadata
988 only (where the metadata is at the start of the device) and is only
989 useful when the component device has changed size (typically become
990 larger). The version 1 metadata records the amount of the device that
991 can be used to store data, so if a device in a version 1.1 or 1.2
992 array becomes larger, the metadata will still be visible, but the
993 extra space will not. In this case it might be useful to assemble the
995 .BR \-\-update=devicesize .
998 to determine the maximum usable amount of space on each device and
999 update the relevant field in the metadata.
1003 .B \-\-auto\-update\-homehost
1004 This flag is only meaningful with auto-assembly (see discussion below).
1005 In that situation, if no suitable arrays are found for this homehost,
1007 will rescan for any arrays at all and will assemble them and update the
1008 homehost to match the current host.
1011 .SH For Manage mode:
1014 .BR \-a ", " \-\-add
1015 hot-add listed devices. For arrays with redundancy, the listed
1016 devices become available as spares. If the array is degraded, it will
1017 immediately start recovering data on to one of these spares.
1021 re\-add a device that was recently removed from an array. This is
1022 normally only needed for arrays that have be built (i.e. with
1024 For created arrays, devices are always re\-added if that is possible,
1025 however using \-\-re\-add will ensure the device isn't made into a
1026 spare if the \-\-re\-add failed.
1028 When re\-adding a device, if nothing has changed on the array since the
1029 device was removed, no recovery is performed. Also, if the array has
1030 a write-intent bitmap, then the recovery performed after a re\-add will
1031 be limited to those blocks which, according to the bitmap, might have
1032 changed since the device was removed.
1034 If the device name given is
1036 then mdadm will try to find any device that looks like it should be
1037 part of the array but isn't and will try to re\-add all such devices.
1040 .BR \-r ", " \-\-remove
1041 remove listed devices. They must not be active. i.e. they should
1042 be failed or spare devices. As well as the name of a device file
1051 The first causes all failed device to be removed. The second causes
1052 any device which is no longer connected to the system (i.e an 'open'
1055 to be removed. This will only succeed for devices that are spares or
1056 have already been marked as failed.
1059 .BR \-f ", " \-\-fail
1060 mark listed devices as faulty.
1061 As well as the name of a device file, the word
1063 can be given. This will cause any device that has been detached from
1064 the system to be marked as failed. It can then be removed.
1072 .BR \-\-write\-mostly
1073 Subsequent devices that are added or re\-added will have the 'write-mostly'
1074 flag set. This is only valid for RAID1 and means that the 'md' driver
1075 will avoid reading from these devices if possible.
1078 Subsequent devices that are added or re\-added will have the 'write-mostly'
1082 Each of these options requires that the first device listed is the array
1083 to be acted upon, and the remainder are component devices to be added,
1084 removed, marked as faulty, etc. Several different operations can be
1085 specified for different devices, e.g.
1087 mdadm /dev/md0 \-\-add /dev/sda1 \-\-fail /dev/sdb1 \-\-remove /dev/sdb1
1089 Each operation applies to all devices listed until the next
1092 If an array is using a write-intent bitmap, then devices which have
1093 been removed can be re\-added in a way that avoids a full
1094 reconstruction but instead just updates the blocks that have changed
1095 since the device was removed. For arrays with persistent metadata
1096 (superblocks) this is done automatically. For arrays created with
1098 mdadm needs to be told that this device we removed recently with
1101 Devices can only be removed from an array if they are not in active
1102 use, i.e. that must be spares or failed devices. To remove an active
1103 device, it must first be marked as
1109 .BR \-Q ", " \-\-query
1110 Examine a device to see
1111 (1) if it is an md device and (2) if it is a component of an md
1113 Information about what is discovered is presented.
1116 .BR \-D ", " \-\-detail
1117 Print details of one or more md devices.
1120 .BR \-\-detail\-platform
1121 Print details of the platform's RAID capabilities (firmware / hardware
1122 topology) for a given metadata format.
1125 .BR \-Y ", " \-\-export
1130 output will be formatted as
1132 pairs for easy import into the environment.
1135 .BR \-E ", " \-\-examine
1136 Print contents of the metadata stored on the named device(s).
1137 Note the contrast between
1142 applies to devices which are components of an array, while
1144 applies to a whole array which is currently active.
1147 If an array was created on a SPARC machine with a 2.2 Linux kernel
1148 patched with RAID support, the superblock will have been created
1149 incorrectly, or at least incompatibly with 2.4 and later kernels.
1154 will fix the superblock before displaying it. If this appears to do
1155 the right thing, then the array can be successfully assembled using
1156 .BR "\-\-assemble \-\-update=sparc2.2" .
1159 .BR \-X ", " \-\-examine\-bitmap
1160 Report information about a bitmap file.
1161 The argument is either an external bitmap file or an array component
1162 in case of an internal bitmap. Note that running this on an array
1165 does not report the bitmap for that array.
1168 .BR \-R ", " \-\-run
1169 start a partially assembled array. If
1171 did not find enough devices to fully start the array, it might leaving
1172 it partially assembled. If you wish, you can then use
1174 to start the array in degraded mode.
1177 .BR \-S ", " \-\-stop
1178 deactivate array, releasing all resources.
1181 .BR \-o ", " \-\-readonly
1182 mark array as readonly.
1185 .BR \-w ", " \-\-readwrite
1186 mark array as readwrite.
1189 .B \-\-zero\-superblock
1190 If the device contains a valid md superblock, the block is
1191 overwritten with zeros. With
1193 the block where the superblock would be is overwritten even if it
1194 doesn't appear to be valid.
1197 .B \-\-kill\-subarray=
1198 If the device is a container and the argument to \-\-kill\-subarray
1199 specifies an inactive subarray in the container, then the subarray is
1200 deleted. Deleting all subarrays will leave an 'empty-container' or
1201 spare superblock on the drives. See \-\-zero\-superblock for completely
1202 removing a superblock. Note that some formats depend on the subarray
1203 index for generating a UUID, this command will fail if it would change
1204 the UUID of an active subarray.
1207 .B \-\-update\-subarray=
1208 If the device is a container and the argument to \-\-update\-subarray
1209 specifies a subarray in the container, then attempt to update the given
1210 superblock field in the subarray. See below in
1215 .BR \-t ", " \-\-test
1220 is set to reflect the status of the device. See below in
1225 .BR \-W ", " \-\-wait
1226 For each md device given, wait for any resync, recovery, or reshape
1227 activity to finish before returning.
1229 will return with success if it actually waited for every device
1230 listed, otherwise it will return failure.
1234 For each md device given, or each device in /proc/mdstat if
1236 is given, arrange for the array to be marked clean as soon as possible.
1238 will return with success if the array uses external metadata and we
1239 successfully waited. For native arrays this returns immediately as the
1240 kernel handles dirty-clean transitions at shutdown. No action is taken
1241 if safe-mode handling is disabled.
1243 .SH For Incremental Assembly mode:
1245 .BR \-\-rebuild\-map ", " \-r
1246 Rebuild the map file
1247 .RB ( /var/run/mdadm/map )
1250 uses to help track which arrays are currently being assembled.
1253 .BR \-\-run ", " \-R
1254 Run any array assembled as soon as a minimal number of devices are
1255 available, rather than waiting until all expected devices are present.
1258 .BR \-\-scan ", " \-s
1259 Only meaningful with
1263 file for arrays that are being incrementally assembled and will try to
1264 start any that are not already started. If any such array is listed
1267 as requiring an external bitmap, that bitmap will be attached first.
1270 .BR \-\-fail ", " \-f
1271 This allows the hot-plug system to remove devices that have fully disappeared
1272 from the kernel. It will first fail and then remove the device from any
1273 array it belongs to.
1274 The device name given should be a kernel device name such as "sda",
1278 .SH For Monitor mode:
1280 .BR \-m ", " \-\-mail
1281 Give a mail address to send alerts to.
1284 .BR \-p ", " \-\-program ", " \-\-alert
1285 Give a program to be run whenever an event is detected.
1288 .BR \-y ", " \-\-syslog
1289 Cause all events to be reported through 'syslog'. The messages have
1290 facility of 'daemon' and varying priorities.
1293 .BR \-d ", " \-\-delay
1294 Give a delay in seconds.
1296 polls the md arrays and then waits this many seconds before polling
1297 again. The default is 60 seconds. Since 2.6.16, there is no need to
1298 reduce this as the kernel alerts
1300 immediately when there is any change.
1303 .BR \-r ", " \-\-increment
1304 Give a percentage increment.
1306 will generate RebuildNN events with the given percentage increment.
1309 .BR \-f ", " \-\-daemonise
1312 to run as a background daemon if it decides to monitor anything. This
1313 causes it to fork and run in the child, and to disconnect from the
1314 terminal. The process id of the child is written to stdout.
1317 which will only continue monitoring if a mail address or alert program
1318 is found in the config file.
1321 .BR \-i ", " \-\-pid\-file
1324 is running in daemon mode, write the pid of the daemon process to
1325 the specified file, instead of printing it on standard output.
1328 .BR \-1 ", " \-\-oneshot
1329 Check arrays only once. This will generate
1331 events and more significantly
1337 .B " mdadm \-\-monitor \-\-scan \-1"
1339 from a cron script will ensure regular notification of any degraded arrays.
1342 .BR \-t ", " \-\-test
1345 alert for every array found at startup. This alert gets mailed and
1346 passed to the alert program. This can be used for testing that alert
1347 message do get through successfully.
1353 .B mdadm \-\-assemble
1354 .I md-device options-and-component-devices...
1357 .B mdadm \-\-assemble \-\-scan
1358 .I md-devices-and-options...
1361 .B mdadm \-\-assemble \-\-scan
1365 This usage assembles one or more RAID arrays from pre-existing components.
1366 For each array, mdadm needs to know the md device, the identity of the
1367 array, and a number of component-devices. These can be found in a number of ways.
1369 In the first usage example (without the
1371 the first device given is the md device.
1372 In the second usage example, all devices listed are treated as md
1373 devices and assembly is attempted.
1374 In the third (where no devices are listed) all md devices that are
1375 listed in the configuration file are assembled. If not arrays are
1376 described by the configuration file, then any arrays that
1377 can be found on unused devices will be assembled.
1379 If precisely one device is listed, but
1385 was given and identity information is extracted from the configuration file.
1387 The identity can be given with the
1393 option, will be taken from the md-device record in the config file, or
1394 will be taken from the super block of the first component-device
1395 listed on the command line.
1397 Devices can be given on the
1399 command line or in the config file. Only devices which have an md
1400 superblock which contains the right identity will be considered for
1403 The config file is only used if explicitly named with
1405 or requested with (a possibly implicit)
1410 .B /etc/mdadm/mdadm.conf
1415 is not given, then the config file will only be used to find the
1416 identity of md arrays.
1418 Normally the array will be started after it is assembled. However if
1420 is not given and not all expected drives were listed, then the array
1421 is not started (to guard against usage errors). To insist that the
1422 array be started in this case (as may work for RAID1, 4, 5, 6, or 10),
1431 does not create any entries in
1435 It does record information in
1436 .B /var/run/mdadm/map
1439 to choose the correct name.
1443 detects that udev is not configured, it will create the devices in
1447 In Linux kernels prior to version 2.6.28 there were two distinctly
1448 different types of md devices that could be created: one that could be
1449 partitioned using standard partitioning tools and one that could not.
1450 Since 2.6.28 that distinction is no longer relevant as both type of
1451 devices can be partitioned.
1453 will normally create the type that originally could not be partitioned
1454 as it has a well defined major number (9).
1456 Prior to 2.6.28, it is important that mdadm chooses the correct type
1457 of array device to use. This can be controlled with the
1459 option. In particular, a value of "mdp" or "part" or "p" tells mdadm
1460 to use a partitionable device rather than the default.
1462 In the no-udev case, the value given to
1464 can be suffixed by a number. This tells
1466 to create that number of partition devices rather than the default of 4.
1470 can also be given in the configuration file as a word starting
1472 on the ARRAY line for the relevant array.
1479 and no devices are listed,
1481 will first attempt to assemble all the arrays listed in the config
1484 In no array at listed in the config (other than those marked
1486 it will look through the available devices for possible arrays and
1487 will try to assemble anything that it finds. Arrays which are tagged
1488 as belonging to the given homehost will be assembled and started
1489 normally. Arrays which do not obviously belong to this host are given
1490 names that are expected not to conflict with anything local, and are
1491 started "read-auto" so that nothing is written to any device until the
1492 array is written to. i.e. automatic resync etc is delayed.
1496 finds a consistent set of devices that look like they should comprise
1497 an array, and if the superblock is tagged as belonging to the given
1498 home host, it will automatically choose a device name and try to
1499 assemble the array. If the array uses version-0.90 metadata, then the
1501 number as recorded in the superblock is used to create a name in
1505 If the array uses version-1 metadata, then the
1507 from the superblock is used to similarly create a name in
1509 (the name will have any 'host' prefix stripped first).
1511 This behaviour can be modified by the
1515 configuration file. This line can indicate that specific metadata
1516 type should, or should not, be automatically assembled. If an array
1517 is found which is not listed in
1519 and has a metadata format that is denied by the
1521 line, then it will not be assembled.
1524 line can also request that all arrays identified as being for this
1525 homehost should be assembled regardless of their metadata type.
1528 for further details.
1533 cannot find any array for the given host at all, and if
1534 .B \-\-auto\-update\-homehost
1537 will search again for any array (not just an array created for this
1538 host) and will assemble each assuming
1539 .BR \-\-update=homehost .
1540 This will change the host tag in the superblock so that on the next run,
1541 these arrays will be found without the second pass. The intention of
1542 this feature is to support transitioning a set of md arrays to using
1545 The reason for requiring arrays to be tagged with the homehost for
1546 auto assembly is to guard against problems that can arise when moving
1547 devices from one host to another.
1558 .BI \-\-raid\-devices= Z
1562 This usage is similar to
1564 The difference is that it creates an array without a superblock. With
1565 these arrays there is no difference between initially creating the array and
1566 subsequently assembling the array, except that hopefully there is useful
1567 data there in the second case.
1569 The level may raid0, linear, raid1, raid10, multipath, or faulty, or
1570 one of their synonyms. All devices must be listed and the array will
1571 be started once complete. It will often be appropriate to use
1572 .B \-\-assume\-clean
1573 with levels raid1 or raid10.
1584 .BI \-\-raid\-devices= Z
1588 This usage will initialise a new md array, associate some devices with
1589 it, and activate the array.
1591 The named device will normally not exist when
1592 .I "mdadm \-\-create"
1593 is run, but will be created by
1595 once the array becomes active.
1597 As devices are added, they are checked to see if they contain RAID
1598 superblocks or filesystems. They are also checked to see if the variance in
1599 device size exceeds 1%.
1601 If any discrepancy is found, the array will not automatically be run, though
1604 can override this caution.
1606 To create a "degraded" array in which some devices are missing, simply
1607 give the word "\fBmissing\fP"
1608 in place of a device name. This will cause
1610 to leave the corresponding slot in the array empty.
1611 For a RAID4 or RAID5 array at most one slot can be
1612 "\fBmissing\fP"; for a RAID6 array at most two slots.
1613 For a RAID1 array, only one real device needs to be given. All of the
1617 When creating a RAID5 array,
1619 will automatically create a degraded array with an extra spare drive.
1620 This is because building the spare into a degraded array is in general
1621 faster than resyncing the parity on a non-degraded, but not clean,
1622 array. This feature can be overridden with the
1626 When creating an array with version-1 metadata a name for the array is
1628 If this is not given with the
1632 will choose a name based on the last component of the name of the
1633 device being created. So if
1635 is being created, then the name
1640 is being created, then the name
1644 When creating a partition based array, using
1646 with version-1.x metadata, the partition type should be set to
1648 (non fs-data). This type selection allows for greater precision since
1649 using any other [RAID auto-detect (0xFD) or a GNU/Linux partition (0x83)],
1650 might create problems in the event of array recovery through a live cdrom.
1652 A new array will normally get a randomly assigned 128bit UUID which is
1653 very likely to be unique. If you have a specific need, you can choose
1654 a UUID for the array by giving the
1656 option. Be warned that creating two arrays with the same UUID is a
1657 recipe for disaster. Also, using
1659 when creating a v0.90 array will silently override any
1664 .\"option is given, it is not necessary to list any component-devices in this command.
1665 .\"They can be added later, before a
1669 .\"is given, the apparent size of the smallest drive given is used.
1671 When creating an array within a
1674 can be given either the list of devices to use, or simply the name of
1675 the container. The former case gives control over which devices in
1676 the container will be used for the array. The latter case allows
1678 to automatically choose which devices to use based on how much spare
1681 The General Management options that are valid with
1686 insist on running the array even if some devices look like they might
1691 start the array readonly \(em not supported yet.
1698 .I options... devices...
1701 This usage will allow individual devices in an array to be failed,
1702 removed or added. It is possible to perform multiple operations with
1703 on command. For example:
1705 .B " mdadm /dev/md0 \-f /dev/hda1 \-r /dev/hda1 \-a /dev/hda1"
1711 and will then remove it from the array and finally add it back
1712 in as a spare. However only one md array can be affected by a single
1715 When a device is added to an active array, mdadm checks to see if it
1716 has metadata on it which suggests that it was recently a member of the
1717 array. If it does, it tries to "re\-add" the device. If there have
1718 been no changes since the device was removed, or if the array has a
1719 write-intent bitmap which has recorded whatever changes there were,
1720 then the device will immediately become a full member of the array and
1721 those differences recorded in the bitmap will be resolved.
1731 MISC mode includes a number of distinct operations that
1732 operate on distinct devices. The operations are:
1735 The device is examined to see if it is
1736 (1) an active md array, or
1737 (2) a component of an md array.
1738 The information discovered is reported.
1742 The device should be an active md device.
1744 will display a detailed description of the array.
1748 will cause the output to be less detailed and the format to be
1749 suitable for inclusion in
1750 .BR /etc/mdadm.conf .
1753 will normally be 0 unless
1755 failed to get useful information about the device(s); however, if the
1757 option is given, then the exit status will be:
1761 The array is functioning normally.
1764 The array has at least one failed device.
1767 The array has multiple failed devices such that it is unusable.
1770 There was an error while trying to get information about the device.
1774 .B \-\-detail\-platform
1775 Print detail of the platform's RAID capabilities (firmware / hardware
1776 topology). If the metadata is specified with
1780 then the return status will be:
1784 metadata successfully enumerated its platform components on this system
1787 metadata is platform independent
1790 metadata failed to find its platform components on this system
1794 .B \-\-update\-subarray=
1795 If the device is a container and the argument to \-\-update\-subarray
1796 specifies a subarray in the container, then attempt to update the given
1797 superblock field in the subarray. Similar to updating an array in
1798 "assemble" mode, the field to update is selected by
1802 option. Currently only
1808 option updates the subarray name in the metadata, it may not affect the
1809 device node name or the device node symlink until the subarray is
1810 re\-assembled. If updating
1812 would change the UUID of an active subarray this operation is blocked,
1813 and the command will end in an error.
1817 The device should be a component of an md array.
1819 will read the md superblock of the device and display the contents.
1824 is given, then multiple devices that are components of the one array
1825 are grouped together and reported in a single entry suitable
1827 .BR /etc/mdadm.conf .
1831 without listing any devices will cause all devices listed in the
1832 config file to be examined.
1836 The devices should be active md arrays which will be deactivated, as
1837 long as they are not currently in use.
1841 This will fully activate a partially assembled md array.
1845 This will mark an active array as read-only, providing that it is
1846 not currently being used.
1852 array back to being read/write.
1856 For all operations except
1859 will cause the operation to be applied to all arrays listed in
1864 causes all devices listed in the config file to be examined.
1867 .BR \-b ", " \-\-brief
1868 Be less verbose. This is used with
1876 gives an intermediate level of verbosity.
1882 .B mdadm \-\-monitor
1883 .I options... devices...
1888 to periodically poll a number of md arrays and to report on any events
1891 will never exit once it decides that there are arrays to be checked,
1892 so it should normally be run in the background.
1894 As well as reporting events,
1896 may move a spare drive from one array to another if they are in the
1899 and if the destination array has a failed drive but no spares.
1901 If any devices are listed on the command line,
1903 will only monitor those devices. Otherwise all arrays listed in the
1904 configuration file will be monitored. Further, if
1906 is given, then any other md devices that appear in
1908 will also be monitored.
1910 The result of monitoring the arrays is the generation of events.
1911 These events are passed to a separate program (if specified) and may
1912 be mailed to a given E-mail address.
1914 When passing events to a program, the program is run once for each event,
1915 and is given 2 or 3 command-line arguments: the first is the
1916 name of the event (see below), the second is the name of the
1917 md device which is affected, and the third is the name of a related
1918 device if relevant (such as a component device that has failed).
1922 is given, then a program or an E-mail address must be specified on the
1923 command line or in the config file. If neither are available, then
1925 will not monitor anything.
1929 will continue monitoring as long as something was found to monitor. If
1930 no program or email is given, then each event is reported to
1933 The different events are:
1937 .B DeviceDisappeared
1938 An md array which previously was configured appears to no longer be
1939 configured. (syslog priority: Critical)
1943 was told to monitor an array which is RAID0 or Linear, then it will
1945 .B DeviceDisappeared
1946 with the extra information
1948 This is because RAID0 and Linear do not support the device-failed,
1949 hot-spare and resync operations which are monitored.
1953 An md array started reconstruction. (syslog priority: Warning)
1959 is a two-digit number (ie. 05, 48). This indicates that rebuild
1960 has passed that many percent of the total. The events are generated
1961 with fixed increment since 0. Increment size may be specified with
1962 a commandline option (default is 20). (syslog priority: Warning)
1966 An md array that was rebuilding, isn't any more, either because it
1967 finished normally or was aborted. (syslog priority: Warning)
1971 An active component device of an array has been marked as
1972 faulty. (syslog priority: Critical)
1976 A spare component device which was being rebuilt to replace a faulty
1977 device has failed. (syslog priority: Critical)
1981 A spare component device which was being rebuilt to replace a faulty
1982 device has been successfully rebuilt and has been made active.
1983 (syslog priority: Info)
1987 A new md array has been detected in the
1989 file. (syslog priority: Info)
1993 A newly noticed array appears to be degraded. This message is not
1996 notices a drive failure which causes degradation, but only when
1998 notices that an array is degraded when it first sees the array.
1999 (syslog priority: Critical)
2003 A spare drive has been moved from one array in a
2005 to another to allow a failed drive to be replaced.
2006 (syslog priority: Info)
2012 has been told, via the config file, that an array should have a certain
2013 number of spare devices, and
2015 detects that it has fewer than this number when it first sees the
2016 array, it will report a
2019 (syslog priority: Warning)
2023 An array was found at startup, and the
2026 (syslog priority: Info)
2036 cause Email to be sent. All events cause the program to be run.
2037 The program is run with two or three arguments: the event
2038 name, the array device and possibly a second device.
2040 Each event has an associated array device (e.g.
2042 and possibly a second device. For
2047 the second device is the relevant component device.
2050 the second device is the array that the spare was moved from.
2054 to move spares from one array to another, the different arrays need to
2055 be labeled with the same
2057 in the configuration file. The
2059 name can be any string; it is only necessary that different spare
2060 groups use different names.
2064 detects that an array in a spare group has fewer active
2065 devices than necessary for the complete array, and has no spare
2066 devices, it will look for another array in the same spare group that
2067 has a full complement of working drive and a spare. It will then
2068 attempt to remove the spare from the second drive and add it to the
2070 If the removal succeeds but the adding fails, then it is added back to
2074 The GROW mode is used for changing the size or shape of an active
2076 For this to work, the kernel must support the necessary change.
2077 Various types of growth are being added during 2.6 development,
2078 including restructuring a RAID5 array to have more active devices.
2080 Currently the only support available is to
2082 change the "size" attribute
2083 for RAID1, RAID5 and RAID6.
2085 increase or decrease the "raid\-devices" attribute of RAID1, RAID5,
2088 change the chunk-size and layout of RAID5 and RAID6.
2090 convert between RAID1 and RAID5, and between RAID5 and RAID6.
2092 add a write-intent bitmap to any array which supports these bitmaps, or
2093 remove a write-intent bitmap from such an array.
2096 GROW mode is not currently supported for
2098 or arrays inside containers.
2101 Normally when an array is built the "size" it taken from the smallest
2102 of the drives. If all the small drives in an arrays are, one at a
2103 time, removed and replaced with larger drives, then you could have an
2104 array of large drives with only a small amount used. In this
2105 situation, changing the "size" with "GROW" mode will allow the extra
2106 space to start being used. If the size is increased in this way, a
2107 "resync" process will start to make sure the new parts of the array
2110 Note that when an array changes size, any filesystem that may be
2111 stored in the array will not automatically grow to use the space. The
2112 filesystem will need to be explicitly told to use the extra space.
2114 Also the size of an array cannot be changed while it has an active
2115 bitmap. If an array has a bitmap, it must be removed before the size
2116 can be changed. Once the change it complete a new bitmap can be created.
2118 .SS RAID\-DEVICES CHANGES
2120 A RAID1 array can work with any number of devices from 1 upwards
2121 (though 1 is not very useful). There may be times which you want to
2122 increase or decrease the number of active devices. Note that this is
2123 different to hot-add or hot-remove which changes the number of
2126 When reducing the number of devices in a RAID1 array, the slots which
2127 are to be removed from the array must already be vacant. That is, the
2128 devices which were in those slots must be failed and removed.
2130 When the number of devices is increased, any hot spares that are
2131 present will be activated immediately.
2133 Changing the number of active devices in a RAID5 or RAID6 is much more
2134 effort. Every block in the array will need to be read and written
2135 back to a new location. From 2.6.17, the Linux Kernel is able to
2136 increase the number of devices in a RAID5 safely, including restarting
2137 an interrupted "reshape". From 2.6.31, the Linux Kernel is able to
2138 increase or decrease the number of devices in a RAID5 or RAID6.
2140 When decreasing the number of devices, the size of the array will also
2141 decrease. If there was data in the array, it could get destroyed and
2142 this is not reversible. To help prevent accidents,
2144 requires that the size of the array be decreased first with
2145 .BR "mdadm --grow --array-size" .
2146 This is a reversible change which simply makes the end of the array
2147 inaccessible. The integrity of any data can then be checked before
2148 the non-reversible reduction in the number of devices is request.
2150 When relocating the first few stripes on a RAID5, it is not possible
2151 to keep the data on disk completely consistent and crash-proof. To
2152 provide the required safety, mdadm disables writes to the array while
2153 this "critical section" is reshaped, and takes a backup of the data
2154 that is in that section. This backup is normally stored in any spare
2155 devices that the array has, however it can also be stored in a
2156 separate file specified with the
2158 option. If this option is used, and the system does crash during the
2159 critical period, the same file must be passed to
2161 to restore the backup and reassemble the array.
2165 Changing the RAID level of any array happens instantaneously. However
2166 in the RAID to RAID6 case this requires a non-standard layout of the
2167 RAID6 data, and in the RAID6 to RAID5 case that non-standard layout is
2168 required before the change can be accomplish. So while the level
2169 change is instant, the accompanying layout change can take quite a
2172 .SS CHUNK-SIZE AND LAYOUT CHANGES
2174 Changing the chunk-size of layout without also changing the number of
2175 devices as the same time will involve re-writing all blocks in-place.
2176 To ensure against data loss in the case of a crash, a
2178 must be provided for these changes. Small sections of the array will
2179 be copied to the backup file while they are being rearranged.
2181 If the reshape is interrupted for any reason, this backup file must be
2183 .B "mdadm --assemble"
2184 so the array can be reassembled. Consequently the file cannot be
2185 stored on the device being reshaped.
2190 A write-intent bitmap can be added to, or removed from, an active
2191 array. Either internal bitmaps, or bitmaps stored in a separate file,
2192 can be added. Note that if you add a bitmap stored in a file which is
2193 in a filesystem that is on the RAID array being affected, the system
2194 will deadlock. The bitmap must be on a separate filesystem.
2196 .SH INCREMENTAL MODE
2200 .B mdadm \-\-incremental
2206 .B mdadm \-\-incremental \-\-fail
2210 .B mdadm \-\-incremental \-\-rebuild\-map
2213 .B mdadm \-\-incremental \-\-run \-\-scan
2216 This mode is designed to be used in conjunction with a device
2217 discovery system. As devices are found in a system, they can be
2219 .B "mdadm \-\-incremental"
2220 to be conditionally added to an appropriate array.
2222 Conversely, it can also be used with the
2224 flag to do just the opposite and find whatever array a particular device
2225 is part of and remove the device from that array.
2227 If the device passed is a
2229 device created by a previous call to
2231 then rather than trying to add that device to an array, all the arrays
2232 described by the metadata of the container will be started.
2235 performs a number of tests to determine if the device is part of an
2236 array, and which array it should be part of. If an appropriate array
2237 is found, or can be created,
2239 adds the device to the array and conditionally starts the array.
2243 will only add devices to an array which were previously working
2244 (active or spare) parts of that array. It does not currently support
2245 automatic inclusion of a new drive as a spare in some array.
2249 makes are as follow:
2251 Is the device permitted by
2253 That is, is it listed in a
2255 line in that file. If
2257 is absent then the default it to allow any device. Similar if
2259 contains the special word
2261 then any device is allowed. Otherwise the device name given to
2263 must match one of the names or patterns in a
2268 Does the device have a valid md superblock. If a specific metadata
2269 version is request with
2273 then only that style of metadata is accepted, otherwise
2275 finds any known version of metadata. If no
2277 metadata is found, the device is rejected.
2281 Does the metadata match an expected array?
2282 The metadata can match in two ways. Either there is an array listed
2285 which identifies the array (either by UUID, by name, by device list,
2286 or by minor-number), or the array was created with a
2292 or on the command line.
2295 is not able to positively identify the array as belonging to the
2296 current host, the device will be rejected.
2300 keeps a list of arrays that it has partially assembled in
2301 .B /var/run/mdadm/map
2303 .B /var/run/mdadm.map
2304 if the directory doesn't exist. Or maybe even
2305 .BR /dev/.mdadm.map ).
2306 If no array exists which matches
2307 the metadata on the new device,
2309 must choose a device name and unit number. It does this based on any
2312 or any name information stored in the metadata. If this name
2313 suggests a unit number, that number will be used, otherwise a free
2314 unit number will be chosen. Normally
2316 will prefer to create a partitionable array, however if the
2320 suggests that a non-partitionable array is preferred, that will be
2323 If the array is not found in the config file and its metadata does not
2324 identify it as belonging to the "homehost", then
2326 will choose a name for the array which is certain not to conflict with
2327 any array which does belong to this host. It does this be adding an
2328 underscore and a small number to the name preferred by the metadata.
2330 Once an appropriate array is found or created and the device is added,
2332 must decide if the array is ready to be started. It will
2333 normally compare the number of available (non-spare) devices to the
2334 number of devices that the metadata suggests need to be active. If
2335 there are at least that many, the array will be started. This means
2336 that if any devices are missing the array will not be restarted.
2342 in which case the array will be run as soon as there are enough
2343 devices present for the data to be accessible. For a RAID1, that
2344 means one device will start the array. For a clean RAID5, the array
2345 will be started as soon as all but one drive is present.
2347 Note that neither of these approaches is really ideal. If it can
2348 be known that all device discovery has completed, then
2352 can be run which will try to start all arrays that are being
2353 incrementally assembled. They are started in "read-auto" mode in
2354 which they are read-only until the first write request. This means
2355 that no metadata updates are made and no attempt at resync or recovery
2356 happens. Further devices that are found before the first write can
2357 still be added safely.
2360 This section describes environment variables that affect how mdadm
2365 Setting this value to 1 will prevent mdadm from automatically launching
2366 mdmon. This variable is intended primarily for debugging mdadm/mdmon.
2372 does not create any device nodes in /dev, but leaves that task to
2376 appears not to be configured, or if this environment variable is set
2379 will create and devices that are needed.
2383 .B " mdadm \-\-query /dev/name-of-device"
2385 This will find out if a given device is a RAID array, or is part of
2386 one, and will provide brief information about the device.
2388 .B " mdadm \-\-assemble \-\-scan"
2390 This will assemble and start all arrays listed in the standard config
2391 file. This command will typically go in a system startup file.
2393 .B " mdadm \-\-stop \-\-scan"
2395 This will shut down all arrays that can be shut down (i.e. are not
2396 currently in use). This will typically go in a system shutdown script.
2398 .B " mdadm \-\-follow \-\-scan \-\-delay=120"
2400 If (and only if) there is an Email address or program given in the
2401 standard config file, then
2402 monitor the status of all arrays listed in that file by
2403 polling them ever 2 minutes.
2405 .B " mdadm \-\-create /dev/md0 \-\-level=1 \-\-raid\-devices=2 /dev/hd[ac]1"
2407 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
2410 .B " echo 'DEVICE /dev/hd*[0\-9] /dev/sd*[0\-9]' > mdadm.conf"
2412 .B " mdadm \-\-detail \-\-scan >> mdadm.conf"
2414 This will create a prototype config file that describes currently
2415 active arrays that are known to be made from partitions of IDE or SCSI drives.
2416 This file should be reviewed before being used as it may
2417 contain unwanted detail.
2419 .B " echo 'DEVICE /dev/hd[a\-z] /dev/sd*[a\-z]' > mdadm.conf"
2421 .B " mdadm \-\-examine \-\-scan \-\-config=mdadm.conf >> mdadm.conf"
2423 This will find arrays which could be assembled from existing IDE and
2424 SCSI whole drives (not partitions), and store the information in the
2425 format of a config file.
2426 This file is very likely to contain unwanted detail, particularly
2429 entries. It should be reviewed and edited before being used as an
2432 .B " mdadm \-\-examine \-\-brief \-\-scan \-\-config=partitions"
2434 .B " mdadm \-Ebsc partitions"
2436 Create a list of devices by reading
2437 .BR /proc/partitions ,
2438 scan these for RAID superblocks, and printout a brief listing of all
2441 .B " mdadm \-Ac partitions \-m 0 /dev/md0"
2443 Scan all partitions and devices listed in
2444 .BR /proc/partitions
2447 out of all such devices with a RAID superblock with a minor number of 0.
2449 .B " mdadm \-\-monitor \-\-scan \-\-daemonise > /var/run/mdadm"
2451 If config file contains a mail address or alert program, run mdadm in
2452 the background in monitor mode monitoring all md devices. Also write
2453 pid of mdadm daemon to
2454 .BR /var/run/mdadm .
2456 .B " mdadm \-Iq /dev/somedevice"
2458 Try to incorporate newly discovered device into some array as
2461 .B " mdadm \-\-incremental \-\-rebuild\-map \-\-run \-\-scan"
2463 Rebuild the array map from any current arrays, and then start any that
2466 .B " mdadm /dev/md4 --fail detached --remove detached"
2468 Any devices which are components of /dev/md4 will be marked as faulty
2469 and then remove from the array.
2471 .B " mdadm --grow /dev/md4 --level=6 --backup-file=/root/backup-md4
2475 which is currently a RAID5 array will be converted to RAID6. There
2476 should normally already be a spare drive attached to the array as a
2477 RAID6 needs one more drive than a matching RAID5.
2479 .B " mdadm --create /dev/md/ddf --metadata=ddf --raid-disks 6 /dev/sd[a-f]"
2481 Create a DDF array over 6 devices.
2483 .B " mdadm --create /dev/md/home -n3 -l5 -z 30000000 /dev/md/ddf"
2485 Create a RAID5 array over any 3 devices in the given DDF set. Use
2486 only 30 gigabytes of each device.
2488 .B " mdadm -A /dev/md/ddf1 /dev/sd[a-f]"
2490 Assemble a pre-exist ddf array.
2492 .B " mdadm -I /dev/md/ddf1"
2494 Assemble all arrays contained in the ddf array, assigning names as
2497 .B " mdadm \-\-create \-\-help"
2499 Provide help about the Create mode.
2501 .B " mdadm \-\-config \-\-help"
2503 Provide help about the format of the config file.
2505 .B " mdadm \-\-help"
2507 Provide general help.
2517 lists all active md devices with information about them.
2519 uses this to find arrays when
2521 is given in Misc mode, and to monitor array reconstruction
2526 The config file lists which devices may be scanned to see if
2527 they contain MD super block, and gives identifying information
2528 (e.g. UUID) about known MD arrays. See
2532 .SS /var/run/mdadm/map
2535 mode is used, this file gets a list of arrays currently being created.
2538 does not exist as a directory, then
2539 .B /var/run/mdadm.map
2542 is not available (as may be the case during early boot),
2544 is used on the basis that
2546 is usually available very early in boot.
2551 understand two sorts of names for array devices.
2553 The first is the so-called 'standard' format name, which matches the
2554 names used by the kernel and which appear in
2557 The second sort can be freely chosen, but must reside in
2559 When giving a device name to
2561 to create or assemble an array, either full path name such as
2565 can be given, or just the suffix of the second sort of name, such as
2571 chooses device names during auto-assembly or incremental assembly, it
2572 will sometimes add a small sequence number to the end of the name to
2573 avoid conflicted between multiple arrays that have the same name. If
2575 can reasonably determine that the array really is meant for this host,
2576 either by a hostname in the metadata, or by the presence of the array
2577 in /etc/mdadm.conf, then it will leave off the suffix if possible.
2578 Also if the homehost is specified as
2581 will only use a suffix if a different array of the same name already
2582 exists or is listed in the config file.
2584 The standard names for non-partitioned arrays (the only sort of md
2585 array available in 2.4 and earlier) are of the form
2589 where NN is a number.
2590 The standard names for partitionable arrays (as available from 2.6
2591 onwards) are of the form
2595 Partition numbers should be indicated by added "pMM" to these, thus "/dev/md/d1p2".
2597 From kernel version, 2.6.28 the "non-partitioned array" can actually
2598 be partitioned. So the "md_dNN" names are no longer needed, and
2599 partitions such as "/dev/mdNNpXX" are possible.
2603 was previously known as
2607 is completely separate from the
2609 package, and does not use the
2611 configuration file at all.
2614 For further information on mdadm usage, MD and the various levels of
2617 .B http://linux\-raid.osdl.org/
2619 (based upon Jakob \(/Ostergaard's Software\-RAID.HOWTO)
2621 .\"for new releases of the RAID driver check out:
2624 .\".UR ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2625 .\"ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2630 .\".UR http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2631 .\"http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2634 The latest version of
2636 should always be available from
2638 .B http://www.kernel.org/pub/linux/utils/raid/mdadm/