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. It is also possible for there to be confusion
326 about whether the superblock applies to a whole device or just the
327 last partition, if that partition starts on a 64K boundary.
328 .ie '{DEFAULT_METADATA}'0.90'
329 .IP "1, 1.0, 1.1, 1.2"
331 .IP "1, 1.0, 1.1, 1.2 default"
333 Use the new version-1 format superblock. This has fewer restrictions.
334 It can easily be moved between hosts with different endian-ness, and a
335 recovery operation can be checkpointed and restarted. The different
336 sub-versions store the superblock at different locations on the
337 device, either at the end (for 1.0), at the start (for 1.1) or 4K from
338 the start (for 1.2). "1" is equivalent to "1.0".
339 'if '{DEFAULT_METADATA}'1.2' "default" is equivalent to "1.2".
341 Use the "Industry Standard" DDF (Disk Data Format) format defined by
343 When creating a DDF array a
345 will be created, and normal arrays can be created in that container.
347 Use the Intel(R) Matrix Storage Manager metadata format. This creates a
349 which is managed in a similar manner to DDF, and is supported by an
350 option-rom on some platforms:
352 .B http://www.intel.com/design/chipsets/matrixstorage_sb.htm
358 This will override any
360 setting in the config file and provides the identity of the host which
361 should be considered the home for any arrays.
363 When creating an array, the
365 will be recorded in the metadata. For version-1 superblocks, it will
366 be prefixed to the array name. For version-0.90 superblocks, part of
367 the SHA1 hash of the hostname will be stored in the later half of the
370 When reporting information about an array, any array which is tagged
371 for the given homehost will be reported as such.
373 When using Auto-Assemble, only arrays tagged for the given homehost
374 will be allowed to use 'local' names (i.e. not ending in '_' followed
375 by a digit string). See below under
376 .BR "Auto Assembly" .
378 .SH For create, build, or grow:
381 .BR \-n ", " \-\-raid\-devices=
382 Specify the number of active devices in the array. This, plus the
383 number of spare devices (see below) must equal the number of
385 (including "\fBmissing\fP" devices)
386 that are listed on the command line for
388 Setting a value of 1 is probably
389 a mistake and so requires that
391 be specified first. A value of 1 will then be allowed for linear,
392 multipath, RAID0 and RAID1. It is never allowed for RAID4, RAID5 or RAID6.
394 This number can only be changed using
396 for RAID1, RAID4, RAID5 and RAID6 arrays, and only on kernels which provide
397 the necessary support.
400 .BR \-x ", " \-\-spare\-devices=
401 Specify the number of spare (eXtra) devices in the initial array.
402 Spares can also be added
403 and removed later. The number of component devices listed
404 on the command line must equal the number of RAID devices plus the
405 number of spare devices.
408 .BR \-z ", " \-\-size=
409 Amount (in Kibibytes) of space to use from each drive in RAID levels 1/4/5/6.
410 This must be a multiple of the chunk size, and must leave about 128Kb
411 of space at the end of the drive for the RAID superblock.
412 If this is not specified
413 (as it normally is not) the smallest drive (or partition) sets the
414 size, though if there is a variance among the drives of greater than 1%, a warning is
417 A suffix of 'M' or 'G' can be given to indicate Megabytes or
418 Gigabytes respectively.
420 This value can be set with
422 for RAID level 1/4/5/6. If the array was created with a size smaller
423 than the currently active drives, the extra space can be accessed
426 The size can be given as
428 which means to choose the largest size that fits on all current drives.
430 Before reducing the size of the array (with
431 .BR "\-\-grow \-\-size=" )
432 you should make sure that space isn't needed. If the device holds a
433 filesystem, you would need to resize the filesystem to use less space.
435 After reducing the array size you should check that the data stored in
436 the device is still available. If the device holds a filesystem, then
437 an 'fsck' of the filesystem is a minimum requirement. If there are
438 problems the array can be made bigger again with no loss with another
439 .B "\-\-grow \-\-size="
442 This value can not be used with
444 metadata such as DDF and IMSM.
447 .BR \-Z ", " \-\-array\-size=
448 This is only meaningful with
450 and its effect is not persistent: when the array is stopped and
451 restarted the default array size will be restored.
453 Setting the array-size causes the array to appear smaller to programs
454 that access the data. This is particularly needed before reshaping an
455 array so that it will be smaller. As the reshape is not reversible,
456 but setting the size with
458 is, it is required that the array size is reduced as appropriate
459 before the number of devices in the array is reduced.
461 Before reducing the size of the array you should make sure that space
462 isn't needed. If the device holds a filesystem, you would need to
463 resize the filesystem to use less space.
465 After reducing the array size you should check that the data stored in
466 the device is still available. If the device holds a filesystem, then
467 an 'fsck' of the filesystem is a minimum requirement. If there are
468 problems the array can be made bigger again with no loss with another
469 .B "\-\-grow \-\-array\-size="
472 A suffix of 'M' or 'G' can be given to indicate Megabytes or
473 Gigabytes respectively.
476 restores the apparent size of the array to be whatever the real
477 amount of available space is.
480 .BR \-c ", " \-\-chunk=
481 Specify chunk size of kibibytes. The default when creating an
482 array is 512KB. To ensure compatibility with earlier versions, the
483 default when Building and array with no persistent metadata is 64KB.
484 This is only meaningful for RAID0, RAID4, RAID5, RAID6, and RAID10.
486 A suffix of 'M' or 'G' can be given to indicate Megabytes or
487 Gigabytes respectively.
491 Specify rounding factor for a Linear array. The size of each
492 component will be rounded down to a multiple of this size.
493 This is a synonym for
495 but highlights the different meaning for Linear as compared to other
496 RAID levels. The default is 64K if a kernel earlier than 2.6.16 is in
497 use, and is 0K (i.e. no rounding) in later kernels.
500 .BR \-l ", " \-\-level=
501 Set RAID level. When used with
503 options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
504 raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty, container.
505 Obviously some of these are synonymous.
509 metadata type is requested, only the
511 level is permitted, and it does not need to be explicitly given.
515 only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid.
519 to change the RAID level in some cases. See LEVEL CHANGES below.
522 .BR \-p ", " \-\-layout=
523 This option configures the fine details of data layout for RAID5, RAID6,
524 and RAID10 arrays, and controls the failure modes for
527 The layout of the RAID5 parity block can be one of
528 .BR left\-asymmetric ,
529 .BR left\-symmetric ,
530 .BR right\-asymmetric ,
531 .BR right\-symmetric ,
532 .BR la ", " ra ", " ls ", " rs .
534 .BR left\-symmetric .
536 It is also possible to cause RAID5 to use a RAID4-like layout by
542 Finally for RAID5 there are DDF\-compatible layouts,
543 .BR ddf\-zero\-restart ,
544 .BR ddf\-N\-restart ,
546 .BR ddf\-N\-continue .
548 These same layouts are available for RAID6. There are also 4 layouts
549 that will provide an intermediate stage for converting between RAID5
550 and RAID6. These provide a layout which is identical to the
551 corresponding RAID5 layout on the first N\-1 devices, and has the 'Q'
552 syndrome (the second 'parity' block used by RAID6) on the last device.
554 .BR left\-symmetric\-6 ,
555 .BR right\-symmetric\-6 ,
556 .BR left\-asymmetric\-6 ,
557 .BR right\-asymmetric\-6 ,
559 .BR parity\-first\-6 .
561 When setting the failure mode for level
564 .BR write\-transient ", " wt ,
565 .BR read\-transient ", " rt ,
566 .BR write\-persistent ", " wp ,
567 .BR read\-persistent ", " rp ,
569 .BR read\-fixable ", " rf ,
570 .BR clear ", " flush ", " none .
572 Each failure mode can be followed by a number, which is used as a period
573 between fault generation. Without a number, the fault is generated
574 once on the first relevant request. With a number, the fault will be
575 generated after that many requests, and will continue to be generated
576 every time the period elapses.
578 Multiple failure modes can be current simultaneously by using the
580 option to set subsequent failure modes.
582 "clear" or "none" will remove any pending or periodic failure modes,
583 and "flush" will clear any persistent faults.
585 Finally, the layout options for RAID10 are one of 'n', 'o' or 'f' followed
586 by a small number. The default is 'n2'. The supported options are:
589 signals 'near' copies. Multiple copies of one data block are at
590 similar offsets in different devices.
593 signals 'offset' copies. Rather than the chunks being duplicated
594 within a stripe, whole stripes are duplicated but are rotated by one
595 device so duplicate blocks are on different devices. Thus subsequent
596 copies of a block are in the next drive, and are one chunk further
601 (multiple copies have very different offsets).
602 See md(4) for more detail about 'near', 'offset', and 'far'.
604 The number is the number of copies of each datablock. 2 is normal, 3
605 can be useful. This number can be at most equal to the number of
606 devices in the array. It does not need to divide evenly into that
607 number (e.g. it is perfectly legal to have an 'n2' layout for an array
608 with an odd number of devices).
610 When an array is converted between RAID5 and RAID6 an intermediate
611 RAID6 layout is used in which the second parity block (Q) is always on
612 the last device. To convert a RAID5 to RAID6 and leave it in this new
613 layout (which does not require re-striping) use
614 .BR \-\-layout=preserve .
615 This will try to avoid any restriping.
617 The converse of this is
618 .B \-\-layout=normalise
619 which will change a non-standard RAID6 layout into a more standard
626 (thus explaining the p of
630 .BR \-b ", " \-\-bitmap=
631 Specify a file to store a write-intent bitmap in. The file should not
634 is also given. The same file should be provided
635 when assembling the array. If the word
637 is given, then the bitmap is stored with the metadata on the array,
638 and so is replicated on all devices. If the word
642 mode, then any bitmap that is present is removed.
644 To help catch typing errors, the filename must contain at least one
645 slash ('/') if it is a real file (not 'internal' or 'none').
647 Note: external bitmaps are only known to work on ext2 and ext3.
648 Storing bitmap files on other filesystems may result in serious problems.
651 .BR \-\-bitmap\-chunk=
652 Set the chunksize of the bitmap. Each bit corresponds to that many
653 Kilobytes of storage.
654 When using a file based bitmap, the default is to use the smallest
655 size that is at-least 4 and requires no more than 2^21 chunks.
658 bitmap, the chunksize defaults to 64Meg, or larger if necessary to
659 fit the bitmap into the available space.
661 A suffix of 'M' or 'G' can be given to indicate Megabytes or
662 Gigabytes respectively.
665 .BR \-W ", " \-\-write\-mostly
666 subsequent devices listed in a
671 command will be flagged as 'write-mostly'. This is valid for RAID1
672 only and means that the 'md' driver will avoid reading from these
673 devices if at all possible. This can be useful if mirroring over a
677 .BR \-\-write\-behind=
678 Specify that write-behind mode should be enabled (valid for RAID1
679 only). If an argument is specified, it will set the maximum number
680 of outstanding writes allowed. The default value is 256.
681 A write-intent bitmap is required in order to use write-behind
682 mode, and write-behind is only attempted on drives marked as
686 .BR \-\-assume\-clean
689 that the array pre-existed and is known to be clean. It can be useful
690 when trying to recover from a major failure as you can be sure that no
691 data will be affected unless you actually write to the array. It can
692 also be used when creating a RAID1 or RAID10 if you want to avoid the
693 initial resync, however this practice \(em while normally safe \(em is not
694 recommended. Use this only if you really know what you are doing.
696 When the devices that will be part of a new array were filled
697 with zeros before creation the operator knows the array is
698 actually clean. If that is the case, such as after running
699 badblocks, this argument can be used to tell mdadm the
700 facts the operator knows.
703 .BR \-\-backup\-file=
706 is used to increase the number of raid-devices in a RAID5 or RAID6 if
707 there are no spare devices available, or to shrink, change RAID level
708 or layout. See the GROW MODE section below on RAID\-DEVICES CHANGES.
709 The file must be stored on a separate device, not on the RAID array
713 .BR \-N ", " \-\-name=
716 for the array. This is currently only effective when creating an
717 array with a version-1 superblock, or an array in a DDF container.
718 The name is a simple textual string that can be used to identify array
719 components when assembling. If name is needed but not specified, it
720 is taken from the basename of the device that is being created.
732 run the array, even if some of the components
733 appear to be active in another array or filesystem. Normally
735 will ask for confirmation before including such components in an
736 array. This option causes that question to be suppressed.
739 .BR \-f ", " \-\-force
742 accept the geometry and layout specified without question. Normally
744 will not allow creation of an array with only one device, and will try
745 to create a RAID5 array with one missing drive (as this makes the
746 initial resync work faster). With
749 will not try to be so clever.
752 .BR \-a ", " "\-\-auto{=yes,md,mdp,part,p}{NN}"
753 Instruct mdadm how to create the device file if needed, possibly allocating
754 an unused minor number. "md" causes a non-partitionable array
755 to be used (though since Linux 2.6.28, these array devices are in fact
756 partitionable). "mdp", "part" or "p" causes a partitionable array (2.6 and
757 later) to be used. "yes" requires the named md device to have
758 a 'standard' format, and the type and minor number will be determined
759 from this. With mdadm 3.0, device creation is normally left up to
761 so this option is unlikely to be needed.
762 See DEVICE NAMES below.
764 The argument can also come immediately after
769 is not given on the command line or in the config file, then
775 is also given, then any
777 entries in the config file will override the
779 instruction given on the command line.
781 For partitionable arrays,
783 will create the device file for the whole array and for the first 4
784 partitions. A different number of partitions can be specified at the
785 end of this option (e.g.
787 If the device name ends with a digit, the partition names add a 'p',
789 .IR /dev/md/home1p3 .
790 If there is no trailing digit, then the partition names just have a
792 .IR /dev/md/scratch3 .
794 If the md device name is in a 'standard' format as described in DEVICE
795 NAMES, then it will be created, if necessary, with the appropriate
796 device number based on that name. If the device name is not in one of these
797 formats, then a unused device number will be allocated. The device
798 number will be considered unused if there is no active array for that
799 number, and there is no entry in /dev for that number and with a
800 non-standard name. Names that are not in 'standard' format are only
801 allowed in "/dev/md/".
805 .\".BR \-\-symlink = no
810 .\"to create devices in
812 .\"it will also create symlinks from
814 .\"with names starting with
820 .\"to suppress this, or
821 .\".B \-\-symlink=yes
822 .\"to enforce this even if it is suppressing
830 .BR \-u ", " \-\-uuid=
831 uuid of array to assemble. Devices which don't have this uuid are
835 .BR \-m ", " \-\-super\-minor=
836 Minor number of device that array was created for. Devices which
837 don't have this minor number are excluded. If you create an array as
838 /dev/md1, then all superblocks will contain the minor number 1, even if
839 the array is later assembled as /dev/md2.
841 Giving the literal word "dev" for
845 to use the minor number of the md device that is being assembled.
848 .B \-\-super\-minor=dev
849 will look for super blocks with a minor number of 0.
852 is only relevant for v0.90 metadata, and should not normally be used.
858 .BR \-N ", " \-\-name=
859 Specify the name of the array to assemble. This must be the name
860 that was specified when creating the array. It must either match
861 the name stored in the superblock exactly, or it must match
864 prefixed to the start of the given name.
867 .BR \-f ", " \-\-force
868 Assemble the array even if the metadata on some devices appears to be
871 cannot find enough working devices to start the array, but can find
872 some devices that are recorded as having failed, then it will mark
873 those devices as working so that the array can be started.
874 An array which requires
876 to be started may contain data corruption. Use it carefully.
880 Attempt to start the array even if fewer drives were given than were
881 present last time the array was active. Normally if not all the
882 expected drives are found and
884 is not used, then the array will be assembled but not started.
887 an attempt will be made to start it anyway.
891 This is the reverse of
893 in that it inhibits the startup of array unless all expected drives
894 are present. This is only needed with
896 and can be used if the physical connections to devices are
897 not as reliable as you would like.
900 .BR \-a ", " "\-\-auto{=no,yes,md,mdp,part}"
901 See this option under Create and Build options.
904 .BR \-b ", " \-\-bitmap=
905 Specify the bitmap file that was given when the array was created. If
908 bitmap, there is no need to specify this when assembling the array.
911 .BR \-\-backup\-file=
914 was used while reshaping an array (e.g. changing number of devices or
915 chunk size) and the system crashed during the critical section, then the same
919 to allow possibly corrupted data to be restored, and the reshape
923 .BR \-\-invalid\-backup
924 If the file needed for the above option is not available for any
925 reason an empty file can be given together with this option to
926 indicate that the backup file is invalid. In this case the data that
927 was being rearranged at the time of the crash could be irrecoverably
928 lost, but the rest of the array may still be recoverable. This option
929 should only be used as a last resort if there is no way to recover the
934 .BR \-U ", " \-\-update=
935 Update the superblock on each device while assembling the array. The
936 argument given to this flag can be one of
951 option will adjust the superblock of an array what was created on a Sparc
952 machine running a patched 2.2 Linux kernel. This kernel got the
953 alignment of part of the superblock wrong. You can use the
954 .B "\-\-examine \-\-sparc2.2"
957 to see what effect this would have.
961 option will update the
963 field on each superblock to match the minor number of the array being
965 This can be useful if
967 reports a different "Preferred Minor" to
969 In some cases this update will be performed automatically
970 by the kernel driver. In particular the update happens automatically
971 at the first write to an array with redundancy (RAID level 1 or
972 greater) on a 2.6 (or later) kernel.
976 option will change the uuid of the array. If a UUID is given with the
978 option that UUID will be used as a new UUID and will
980 be used to help identify the devices in the array.
983 is given, a random UUID is chosen.
987 option will change the
989 of the array as stored in the superblock. This is only supported for
990 version-1 superblocks.
994 option will change the
996 as recorded in the superblock. For version-0 superblocks, this is the
997 same as updating the UUID.
998 For version-1 superblocks, this involves updating the name.
1002 option will cause the array to be marked
1004 meaning that any redundancy in the array (e.g. parity for RAID5,
1005 copies for RAID1) may be incorrect. This will cause the RAID system
1006 to perform a "resync" pass to make sure that all redundant information
1011 option allows arrays to be moved between machines with different
1013 When assembling such an array for the first time after a move, giving
1014 .B "\-\-update=byteorder"
1017 to expect superblocks to have their byteorder reversed, and will
1018 correct that order before assembling the array. This is only valid
1019 with original (Version 0.90) superblocks.
1023 option will correct the summaries in the superblock. That is the
1024 counts of total, working, active, failed, and spare devices.
1028 option will rarely be of use. It applies to version 1.1 and 1.2 metadata
1029 only (where the metadata is at the start of the device) and is only
1030 useful when the component device has changed size (typically become
1031 larger). The version 1 metadata records the amount of the device that
1032 can be used to store data, so if a device in a version 1.1 or 1.2
1033 array becomes larger, the metadata will still be visible, but the
1034 extra space will not. In this case it might be useful to assemble the
1036 .BR \-\-update=devicesize .
1039 to determine the maximum usable amount of space on each device and
1040 update the relevant field in the metadata.
1044 option can be used when an array has an internal bitmap which is
1045 corrupt in some way so that assembling the array normally fails. It
1046 will cause any internal bitmap to be ignored.
1050 .B \-\-auto\-update\-homehost
1051 This flag is only meaningful with auto-assembly (see discussion below).
1052 In that situation, if no suitable arrays are found for this homehost,
1054 will rescan for any arrays at all and will assemble them and update the
1055 homehost to match the current host.
1058 .SH For Manage mode:
1061 .BR \-t ", " \-\-test
1062 Unless a more serious error occurred,
1064 will exit with a status of 2 if no changes were made to the array and
1065 0 if at least one change was made.
1066 This can be useful when an indirect specifier such as
1071 is used in requesting an operation on the array.
1073 will report failure if these specifiers didn't find any match.
1076 .BR \-a ", " \-\-add
1077 hot-add listed devices.
1078 If a device appears to have recently been part of the array
1079 (possibly it failed or was removed) the device is re\-added as describe
1081 If that fails or the device was never part of the array, the device is
1082 added as a hot-spare.
1083 If the array is degraded, it will immediately start to rebuild data
1086 Note that this and the following options are only meaningful on array
1087 with redundancy. They don't apply to RAID0 or Linear.
1091 re\-add a device that was previous removed from an array.
1092 If the metadata on the device reports that it is a member of the
1093 array, and the slot that it used is still vacant, then the device will
1094 be added back to the array in the same position. This will normally
1095 cause the data for that device to be recovered. However based on the
1096 event count on the device, the recovery may only require sections that
1097 are flagged a write-intent bitmap to be recovered or may not require
1098 any recovery at all.
1100 When used on an array that has no metadata (i.e. it was built with
1102 it will be assumed that bitmap-based recovery is enough to make the
1103 device fully consistent with the array.
1107 can be accompanied by
1108 .BR \-\-update=devicesize .
1109 See the description of this option when used in Assemble mode for an
1110 explanation of its use.
1112 If the device name given is
1114 then mdadm will try to find any device that looks like it should be
1115 part of the array but isn't and will try to re\-add all such devices.
1118 .BR \-r ", " \-\-remove
1119 remove listed devices. They must not be active. i.e. they should
1120 be failed or spare devices. As well as the name of a device file
1129 The first causes all failed device to be removed. The second causes
1130 any device which is no longer connected to the system (i.e an 'open'
1133 to be removed. This will only succeed for devices that are spares or
1134 have already been marked as failed.
1137 .BR \-f ", " \-\-fail
1138 mark listed devices as faulty.
1139 As well as the name of a device file, the word
1141 can be given. This will cause any device that has been detached from
1142 the system to be marked as failed. It can then be removed.
1150 .BR \-\-write\-mostly
1151 Subsequent devices that are added or re\-added will have the 'write-mostly'
1152 flag set. This is only valid for RAID1 and means that the 'md' driver
1153 will avoid reading from these devices if possible.
1156 Subsequent devices that are added or re\-added will have the 'write-mostly'
1160 Each of these options requires that the first device listed is the array
1161 to be acted upon, and the remainder are component devices to be added,
1162 removed, marked as faulty, etc. Several different operations can be
1163 specified for different devices, e.g.
1165 mdadm /dev/md0 \-\-add /dev/sda1 \-\-fail /dev/sdb1 \-\-remove /dev/sdb1
1167 Each operation applies to all devices listed until the next
1170 If an array is using a write-intent bitmap, then devices which have
1171 been removed can be re\-added in a way that avoids a full
1172 reconstruction but instead just updates the blocks that have changed
1173 since the device was removed. For arrays with persistent metadata
1174 (superblocks) this is done automatically. For arrays created with
1176 mdadm needs to be told that this device we removed recently with
1179 Devices can only be removed from an array if they are not in active
1180 use, i.e. that must be spares or failed devices. To remove an active
1181 device, it must first be marked as
1187 .BR \-Q ", " \-\-query
1188 Examine a device to see
1189 (1) if it is an md device and (2) if it is a component of an md
1191 Information about what is discovered is presented.
1194 .BR \-D ", " \-\-detail
1195 Print details of one or more md devices.
1198 .BR \-\-detail\-platform
1199 Print details of the platform's RAID capabilities (firmware / hardware
1200 topology) for a given metadata format.
1203 .BR \-Y ", " \-\-export
1208 output will be formatted as
1210 pairs for easy import into the environment.
1213 .BR \-E ", " \-\-examine
1214 Print contents of the metadata stored on the named device(s).
1215 Note the contrast between
1220 applies to devices which are components of an array, while
1222 applies to a whole array which is currently active.
1225 If an array was created on a SPARC machine with a 2.2 Linux kernel
1226 patched with RAID support, the superblock will have been created
1227 incorrectly, or at least incompatibly with 2.4 and later kernels.
1232 will fix the superblock before displaying it. If this appears to do
1233 the right thing, then the array can be successfully assembled using
1234 .BR "\-\-assemble \-\-update=sparc2.2" .
1237 .BR \-X ", " \-\-examine\-bitmap
1238 Report information about a bitmap file.
1239 The argument is either an external bitmap file or an array component
1240 in case of an internal bitmap. Note that running this on an array
1243 does not report the bitmap for that array.
1246 .BR \-R ", " \-\-run
1247 start a partially assembled array. If
1249 did not find enough devices to fully start the array, it might leaving
1250 it partially assembled. If you wish, you can then use
1252 to start the array in degraded mode.
1255 .BR \-S ", " \-\-stop
1256 deactivate array, releasing all resources.
1259 .BR \-o ", " \-\-readonly
1260 mark array as readonly.
1263 .BR \-w ", " \-\-readwrite
1264 mark array as readwrite.
1267 .B \-\-zero\-superblock
1268 If the device contains a valid md superblock, the block is
1269 overwritten with zeros. With
1271 the block where the superblock would be is overwritten even if it
1272 doesn't appear to be valid.
1275 .B \-\-kill\-subarray=
1276 If the device is a container and the argument to \-\-kill\-subarray
1277 specifies an inactive subarray in the container, then the subarray is
1278 deleted. Deleting all subarrays will leave an 'empty-container' or
1279 spare superblock on the drives. See \-\-zero\-superblock for completely
1280 removing a superblock. Note that some formats depend on the subarray
1281 index for generating a UUID, this command will fail if it would change
1282 the UUID of an active subarray.
1285 .B \-\-update\-subarray=
1286 If the device is a container and the argument to \-\-update\-subarray
1287 specifies a subarray in the container, then attempt to update the given
1288 superblock field in the subarray. See below in
1293 .BR \-t ", " \-\-test
1298 is set to reflect the status of the device. See below in
1303 .BR \-W ", " \-\-wait
1304 For each md device given, wait for any resync, recovery, or reshape
1305 activity to finish before returning.
1307 will return with success if it actually waited for every device
1308 listed, otherwise it will return failure.
1312 For each md device given, or each device in /proc/mdstat if
1314 is given, arrange for the array to be marked clean as soon as possible.
1316 will return with success if the array uses external metadata and we
1317 successfully waited. For native arrays this returns immediately as the
1318 kernel handles dirty-clean transitions at shutdown. No action is taken
1319 if safe-mode handling is disabled.
1321 .SH For Incremental Assembly mode:
1323 .BR \-\-rebuild\-map ", " \-r
1324 Rebuild the map file
1325 .RB ( /var/run/mdadm/map )
1328 uses to help track which arrays are currently being assembled.
1331 .BR \-\-run ", " \-R
1332 Run any array assembled as soon as a minimal number of devices are
1333 available, rather than waiting until all expected devices are present.
1336 .BR \-\-scan ", " \-s
1337 Only meaningful with
1341 file for arrays that are being incrementally assembled and will try to
1342 start any that are not already started. If any such array is listed
1345 as requiring an external bitmap, that bitmap will be attached first.
1348 .BR \-\-fail ", " \-f
1349 This allows the hot-plug system to remove devices that have fully disappeared
1350 from the kernel. It will first fail and then remove the device from any
1351 array it belongs to.
1352 The device name given should be a kernel device name such as "sda",
1358 Only used with \-\-fail. The 'path' given will be recorded so that if
1359 a new device appears at the same location it can be automatically
1360 added to the same array. This allows the failed device to be
1361 automatically replaced by a new device without metadata if it appears
1362 at specified path. This option is normally only set by a
1366 .SH For Monitor mode:
1368 .BR \-m ", " \-\-mail
1369 Give a mail address to send alerts to.
1372 .BR \-p ", " \-\-program ", " \-\-alert
1373 Give a program to be run whenever an event is detected.
1376 .BR \-y ", " \-\-syslog
1377 Cause all events to be reported through 'syslog'. The messages have
1378 facility of 'daemon' and varying priorities.
1381 .BR \-d ", " \-\-delay
1382 Give a delay in seconds.
1384 polls the md arrays and then waits this many seconds before polling
1385 again. The default is 60 seconds. Since 2.6.16, there is no need to
1386 reduce this as the kernel alerts
1388 immediately when there is any change.
1391 .BR \-r ", " \-\-increment
1392 Give a percentage increment.
1394 will generate RebuildNN events with the given percentage increment.
1397 .BR \-f ", " \-\-daemonise
1400 to run as a background daemon if it decides to monitor anything. This
1401 causes it to fork and run in the child, and to disconnect from the
1402 terminal. The process id of the child is written to stdout.
1405 which will only continue monitoring if a mail address or alert program
1406 is found in the config file.
1409 .BR \-i ", " \-\-pid\-file
1412 is running in daemon mode, write the pid of the daemon process to
1413 the specified file, instead of printing it on standard output.
1416 .BR \-1 ", " \-\-oneshot
1417 Check arrays only once. This will generate
1419 events and more significantly
1425 .B " mdadm \-\-monitor \-\-scan \-1"
1427 from a cron script will ensure regular notification of any degraded arrays.
1430 .BR \-t ", " \-\-test
1433 alert for every array found at startup. This alert gets mailed and
1434 passed to the alert program. This can be used for testing that alert
1435 message do get through successfully.
1439 This inhibits the functionality for moving spares between arrays.
1440 Only one monitoring process started with
1442 but without this flag is allowed, otherwise the two could interfere
1449 .B mdadm \-\-assemble
1450 .I md-device options-and-component-devices...
1453 .B mdadm \-\-assemble \-\-scan
1454 .I md-devices-and-options...
1457 .B mdadm \-\-assemble \-\-scan
1461 This usage assembles one or more RAID arrays from pre-existing components.
1462 For each array, mdadm needs to know the md device, the identity of the
1463 array, and a number of component-devices. These can be found in a number of ways.
1465 In the first usage example (without the
1467 the first device given is the md device.
1468 In the second usage example, all devices listed are treated as md
1469 devices and assembly is attempted.
1470 In the third (where no devices are listed) all md devices that are
1471 listed in the configuration file are assembled. If not arrays are
1472 described by the configuration file, then any arrays that
1473 can be found on unused devices will be assembled.
1475 If precisely one device is listed, but
1481 was given and identity information is extracted from the configuration file.
1483 The identity can be given with the
1489 option, will be taken from the md-device record in the config file, or
1490 will be taken from the super block of the first component-device
1491 listed on the command line.
1493 Devices can be given on the
1495 command line or in the config file. Only devices which have an md
1496 superblock which contains the right identity will be considered for
1499 The config file is only used if explicitly named with
1501 or requested with (a possibly implicit)
1506 .B /etc/mdadm/mdadm.conf
1511 is not given, then the config file will only be used to find the
1512 identity of md arrays.
1514 Normally the array will be started after it is assembled. However if
1516 is not given and not all expected drives were listed, then the array
1517 is not started (to guard against usage errors). To insist that the
1518 array be started in this case (as may work for RAID1, 4, 5, 6, or 10),
1527 does not create any entries in
1531 It does record information in
1532 .B /var/run/mdadm/map
1535 to choose the correct name.
1539 detects that udev is not configured, it will create the devices in
1543 In Linux kernels prior to version 2.6.28 there were two distinctly
1544 different types of md devices that could be created: one that could be
1545 partitioned using standard partitioning tools and one that could not.
1546 Since 2.6.28 that distinction is no longer relevant as both type of
1547 devices can be partitioned.
1549 will normally create the type that originally could not be partitioned
1550 as it has a well defined major number (9).
1552 Prior to 2.6.28, it is important that mdadm chooses the correct type
1553 of array device to use. This can be controlled with the
1555 option. In particular, a value of "mdp" or "part" or "p" tells mdadm
1556 to use a partitionable device rather than the default.
1558 In the no-udev case, the value given to
1560 can be suffixed by a number. This tells
1562 to create that number of partition devices rather than the default of 4.
1566 can also be given in the configuration file as a word starting
1568 on the ARRAY line for the relevant array.
1575 and no devices are listed,
1577 will first attempt to assemble all the arrays listed in the config
1580 In no array at listed in the config (other than those marked
1582 it will look through the available devices for possible arrays and
1583 will try to assemble anything that it finds. Arrays which are tagged
1584 as belonging to the given homehost will be assembled and started
1585 normally. Arrays which do not obviously belong to this host are given
1586 names that are expected not to conflict with anything local, and are
1587 started "read-auto" so that nothing is written to any device until the
1588 array is written to. i.e. automatic resync etc is delayed.
1592 finds a consistent set of devices that look like they should comprise
1593 an array, and if the superblock is tagged as belonging to the given
1594 home host, it will automatically choose a device name and try to
1595 assemble the array. If the array uses version-0.90 metadata, then the
1597 number as recorded in the superblock is used to create a name in
1601 If the array uses version-1 metadata, then the
1603 from the superblock is used to similarly create a name in
1605 (the name will have any 'host' prefix stripped first).
1607 This behaviour can be modified by the
1611 configuration file. This line can indicate that specific metadata
1612 type should, or should not, be automatically assembled. If an array
1613 is found which is not listed in
1615 and has a metadata format that is denied by the
1617 line, then it will not be assembled.
1620 line can also request that all arrays identified as being for this
1621 homehost should be assembled regardless of their metadata type.
1624 for further details.
1629 cannot find any array for the given host at all, and if
1630 .B \-\-auto\-update\-homehost
1633 will search again for any array (not just an array created for this
1634 host) and will assemble each assuming
1635 .BR \-\-update=homehost .
1636 This will change the host tag in the superblock so that on the next run,
1637 these arrays will be found without the second pass. The intention of
1638 this feature is to support transitioning a set of md arrays to using
1641 The reason for requiring arrays to be tagged with the homehost for
1642 auto assembly is to guard against problems that can arise when moving
1643 devices from one host to another.
1654 .BI \-\-raid\-devices= Z
1658 This usage is similar to
1660 The difference is that it creates an array without a superblock. With
1661 these arrays there is no difference between initially creating the array and
1662 subsequently assembling the array, except that hopefully there is useful
1663 data there in the second case.
1665 The level may raid0, linear, raid1, raid10, multipath, or faulty, or
1666 one of their synonyms. All devices must be listed and the array will
1667 be started once complete. It will often be appropriate to use
1668 .B \-\-assume\-clean
1669 with levels raid1 or raid10.
1680 .BI \-\-raid\-devices= Z
1684 This usage will initialise a new md array, associate some devices with
1685 it, and activate the array.
1687 The named device will normally not exist when
1688 .I "mdadm \-\-create"
1689 is run, but will be created by
1691 once the array becomes active.
1693 As devices are added, they are checked to see if they contain RAID
1694 superblocks or filesystems. They are also checked to see if the variance in
1695 device size exceeds 1%.
1697 If any discrepancy is found, the array will not automatically be run, though
1700 can override this caution.
1702 To create a "degraded" array in which some devices are missing, simply
1703 give the word "\fBmissing\fP"
1704 in place of a device name. This will cause
1706 to leave the corresponding slot in the array empty.
1707 For a RAID4 or RAID5 array at most one slot can be
1708 "\fBmissing\fP"; for a RAID6 array at most two slots.
1709 For a RAID1 array, only one real device needs to be given. All of the
1713 When creating a RAID5 array,
1715 will automatically create a degraded array with an extra spare drive.
1716 This is because building the spare into a degraded array is in general
1717 faster than resyncing the parity on a non-degraded, but not clean,
1718 array. This feature can be overridden with the
1722 When creating an array with version-1 metadata a name for the array is
1724 If this is not given with the
1728 will choose a name based on the last component of the name of the
1729 device being created. So if
1731 is being created, then the name
1736 is being created, then the name
1740 When creating a partition based array, using
1742 with version-1.x metadata, the partition type should be set to
1744 (non fs-data). This type selection allows for greater precision since
1745 using any other [RAID auto-detect (0xFD) or a GNU/Linux partition (0x83)],
1746 might create problems in the event of array recovery through a live cdrom.
1748 A new array will normally get a randomly assigned 128bit UUID which is
1749 very likely to be unique. If you have a specific need, you can choose
1750 a UUID for the array by giving the
1752 option. Be warned that creating two arrays with the same UUID is a
1753 recipe for disaster. Also, using
1755 when creating a v0.90 array will silently override any
1760 .\"option is given, it is not necessary to list any component-devices in this command.
1761 .\"They can be added later, before a
1765 .\"is given, the apparent size of the smallest drive given is used.
1767 When creating an array within a
1770 can be given either the list of devices to use, or simply the name of
1771 the container. The former case gives control over which devices in
1772 the container will be used for the array. The latter case allows
1774 to automatically choose which devices to use based on how much spare
1777 The General Management options that are valid with
1782 insist on running the array even if some devices look like they might
1787 start the array readonly \(em not supported yet.
1794 .I options... devices...
1797 This usage will allow individual devices in an array to be failed,
1798 removed or added. It is possible to perform multiple operations with
1799 on command. For example:
1801 .B " mdadm /dev/md0 \-f /dev/hda1 \-r /dev/hda1 \-a /dev/hda1"
1807 and will then remove it from the array and finally add it back
1808 in as a spare. However only one md array can be affected by a single
1811 When a device is added to an active array, mdadm checks to see if it
1812 has metadata on it which suggests that it was recently a member of the
1813 array. If it does, it tries to "re\-add" the device. If there have
1814 been no changes since the device was removed, or if the array has a
1815 write-intent bitmap which has recorded whatever changes there were,
1816 then the device will immediately become a full member of the array and
1817 those differences recorded in the bitmap will be resolved.
1827 MISC mode includes a number of distinct operations that
1828 operate on distinct devices. The operations are:
1831 The device is examined to see if it is
1832 (1) an active md array, or
1833 (2) a component of an md array.
1834 The information discovered is reported.
1838 The device should be an active md device.
1840 will display a detailed description of the array.
1844 will cause the output to be less detailed and the format to be
1845 suitable for inclusion in
1849 will normally be 0 unless
1851 failed to get useful information about the device(s); however, if the
1853 option is given, then the exit status will be:
1857 The array is functioning normally.
1860 The array has at least one failed device.
1863 The array has multiple failed devices such that it is unusable.
1866 There was an error while trying to get information about the device.
1870 .B \-\-detail\-platform
1871 Print detail of the platform's RAID capabilities (firmware / hardware
1872 topology). If the metadata is specified with
1876 then the return status will be:
1880 metadata successfully enumerated its platform components on this system
1883 metadata is platform independent
1886 metadata failed to find its platform components on this system
1890 .B \-\-update\-subarray=
1891 If the device is a container and the argument to \-\-update\-subarray
1892 specifies a subarray in the container, then attempt to update the given
1893 superblock field in the subarray. Similar to updating an array in
1894 "assemble" mode, the field to update is selected by
1898 option. Currently only
1904 option updates the subarray name in the metadata, it may not affect the
1905 device node name or the device node symlink until the subarray is
1906 re\-assembled. If updating
1908 would change the UUID of an active subarray this operation is blocked,
1909 and the command will end in an error.
1913 The device should be a component of an md array.
1915 will read the md superblock of the device and display the contents.
1920 is given, then multiple devices that are components of the one array
1921 are grouped together and reported in a single entry suitable
1927 without listing any devices will cause all devices listed in the
1928 config file to be examined.
1932 The devices should be active md arrays which will be deactivated, as
1933 long as they are not currently in use.
1937 This will fully activate a partially assembled md array.
1941 This will mark an active array as read-only, providing that it is
1942 not currently being used.
1948 array back to being read/write.
1952 For all operations except
1955 will cause the operation to be applied to all arrays listed in
1960 causes all devices listed in the config file to be examined.
1963 .BR \-b ", " \-\-brief
1964 Be less verbose. This is used with
1972 gives an intermediate level of verbosity.
1978 .B mdadm \-\-monitor
1979 .I options... devices...
1984 to periodically poll a number of md arrays and to report on any events
1987 will never exit once it decides that there are arrays to be checked,
1988 so it should normally be run in the background.
1990 As well as reporting events,
1992 may move a spare drive from one array to another if they are in the
1997 and if the destination array has a failed drive but no spares.
1999 If any devices are listed on the command line,
2001 will only monitor those devices. Otherwise all arrays listed in the
2002 configuration file will be monitored. Further, if
2004 is given, then any other md devices that appear in
2006 will also be monitored.
2008 The result of monitoring the arrays is the generation of events.
2009 These events are passed to a separate program (if specified) and may
2010 be mailed to a given E-mail address.
2012 When passing events to a program, the program is run once for each event,
2013 and is given 2 or 3 command-line arguments: the first is the
2014 name of the event (see below), the second is the name of the
2015 md device which is affected, and the third is the name of a related
2016 device if relevant (such as a component device that has failed).
2020 is given, then a program or an E-mail address must be specified on the
2021 command line or in the config file. If neither are available, then
2023 will not monitor anything.
2027 will continue monitoring as long as something was found to monitor. If
2028 no program or email is given, then each event is reported to
2031 The different events are:
2035 .B DeviceDisappeared
2036 An md array which previously was configured appears to no longer be
2037 configured. (syslog priority: Critical)
2041 was told to monitor an array which is RAID0 or Linear, then it will
2043 .B DeviceDisappeared
2044 with the extra information
2046 This is because RAID0 and Linear do not support the device-failed,
2047 hot-spare and resync operations which are monitored.
2051 An md array started reconstruction. (syslog priority: Warning)
2057 is a two-digit number (ie. 05, 48). This indicates that rebuild
2058 has passed that many percent of the total. The events are generated
2059 with fixed increment since 0. Increment size may be specified with
2060 a commandline option (default is 20). (syslog priority: Warning)
2064 An md array that was rebuilding, isn't any more, either because it
2065 finished normally or was aborted. (syslog priority: Warning)
2069 An active component device of an array has been marked as
2070 faulty. (syslog priority: Critical)
2074 A spare component device which was being rebuilt to replace a faulty
2075 device has failed. (syslog priority: Critical)
2079 A spare component device which was being rebuilt to replace a faulty
2080 device has been successfully rebuilt and has been made active.
2081 (syslog priority: Info)
2085 A new md array has been detected in the
2087 file. (syslog priority: Info)
2091 A newly noticed array appears to be degraded. This message is not
2094 notices a drive failure which causes degradation, but only when
2096 notices that an array is degraded when it first sees the array.
2097 (syslog priority: Critical)
2101 A spare drive has been moved from one array in a
2105 to another to allow a failed drive to be replaced.
2106 (syslog priority: Info)
2112 has been told, via the config file, that an array should have a certain
2113 number of spare devices, and
2115 detects that it has fewer than this number when it first sees the
2116 array, it will report a
2119 (syslog priority: Warning)
2123 An array was found at startup, and the
2126 (syslog priority: Info)
2136 cause Email to be sent. All events cause the program to be run.
2137 The program is run with two or three arguments: the event
2138 name, the array device and possibly a second device.
2140 Each event has an associated array device (e.g.
2142 and possibly a second device. For
2147 the second device is the relevant component device.
2150 the second device is the array that the spare was moved from.
2154 to move spares from one array to another, the different arrays need to
2155 be labeled with the same
2157 or the spares must be allowed to migrate through matching POLICY domains
2158 in the configuration file. The
2160 name can be any string; it is only necessary that different spare
2161 groups use different names.
2165 detects that an array in a spare group has fewer active
2166 devices than necessary for the complete array, and has no spare
2167 devices, it will look for another array in the same spare group that
2168 has a full complement of working drive and a spare. It will then
2169 attempt to remove the spare from the second drive and add it to the
2171 If the removal succeeds but the adding fails, then it is added back to
2174 If the spare group for a degraded array is not defined,
2176 will look at the rules of spare migration specified by POLICY lines in
2178 and then follow similar steps as above if a matching spare is found.
2181 The GROW mode is used for changing the size or shape of an active
2183 For this to work, the kernel must support the necessary change.
2184 Various types of growth are being added during 2.6 development,
2185 including restructuring a RAID5 array to have more active devices.
2187 Currently the only support available is to
2189 change the "size" attribute
2190 for RAID1, RAID5 and RAID6.
2192 increase or decrease the "raid\-devices" attribute of RAID1, RAID5,
2195 change the chunk-size and layout of RAID5 and RAID6.
2197 convert between RAID1 and RAID5, and between RAID5 and RAID6.
2199 add a write-intent bitmap to any array which supports these bitmaps, or
2200 remove a write-intent bitmap from such an array.
2203 GROW mode is not currently supported for
2205 or arrays inside containers.
2208 Normally when an array is built the "size" it taken from the smallest
2209 of the drives. If all the small drives in an arrays are, one at a
2210 time, removed and replaced with larger drives, then you could have an
2211 array of large drives with only a small amount used. In this
2212 situation, changing the "size" with "GROW" mode will allow the extra
2213 space to start being used. If the size is increased in this way, a
2214 "resync" process will start to make sure the new parts of the array
2217 Note that when an array changes size, any filesystem that may be
2218 stored in the array will not automatically grow to use the space. The
2219 filesystem will need to be explicitly told to use the extra space.
2221 Also the size of an array cannot be changed while it has an active
2222 bitmap. If an array has a bitmap, it must be removed before the size
2223 can be changed. Once the change it complete a new bitmap can be created.
2225 .SS RAID\-DEVICES CHANGES
2227 A RAID1 array can work with any number of devices from 1 upwards
2228 (though 1 is not very useful). There may be times which you want to
2229 increase or decrease the number of active devices. Note that this is
2230 different to hot-add or hot-remove which changes the number of
2233 When reducing the number of devices in a RAID1 array, the slots which
2234 are to be removed from the array must already be vacant. That is, the
2235 devices which were in those slots must be failed and removed.
2237 When the number of devices is increased, any hot spares that are
2238 present will be activated immediately.
2240 Changing the number of active devices in a RAID5 or RAID6 is much more
2241 effort. Every block in the array will need to be read and written
2242 back to a new location. From 2.6.17, the Linux Kernel is able to
2243 increase the number of devices in a RAID5 safely, including restarting
2244 an interrupted "reshape". From 2.6.31, the Linux Kernel is able to
2245 increase or decrease the number of devices in a RAID5 or RAID6.
2247 When decreasing the number of devices, the size of the array will also
2248 decrease. If there was data in the array, it could get destroyed and
2249 this is not reversible. To help prevent accidents,
2251 requires that the size of the array be decreased first with
2252 .BR "mdadm --grow --array-size" .
2253 This is a reversible change which simply makes the end of the array
2254 inaccessible. The integrity of any data can then be checked before
2255 the non-reversible reduction in the number of devices is request.
2257 When relocating the first few stripes on a RAID5 or RAID6, it is not
2258 possible to keep the data on disk completely consistent and
2259 crash-proof. To provide the required safety, mdadm disables writes to
2260 the array while this "critical section" is reshaped, and takes a
2261 backup of the data that is in that section. For grows, this backup may be
2262 stored in any spare devices that the array has, however it can also be
2263 stored in a separate file specified with the
2265 option, and is required to be specified for shrinks, RAID level
2266 changes and layout changes. If this option is used, and the system
2267 does crash during the critical period, the same file must be passed to
2269 to restore the backup and reassemble the array. When shrinking rather
2270 than growing the array, the reshape is done from the end towards the
2271 beginning, so the "critical section" is at the end of the reshape.
2275 Changing the RAID level of any array happens instantaneously. However
2276 in the RAID5 to RAID6 case this requires a non-standard layout of the
2277 RAID6 data, and in the RAID6 to RAID5 case that non-standard layout is
2278 required before the change can be accomplished. So while the level
2279 change is instant, the accompanying layout change can take quite a
2282 is required. If the array is not simultaneously being grown or
2283 shrunk, so that the array size will remain the same - for example,
2284 reshaping a 3-drive RAID5 into a 4-drive RAID6 - the backup file will
2285 be used not just for a "cricital section" but throughout the reshape
2286 operation, as described below under LAYOUT CHANGES.
2288 .SS CHUNK-SIZE AND LAYOUT CHANGES
2290 Changing the chunk-size of layout without also changing the number of
2291 devices as the same time will involve re-writing all blocks in-place.
2292 To ensure against data loss in the case of a crash, a
2294 must be provided for these changes. Small sections of the array will
2295 be copied to the backup file while they are being rearranged. This
2296 means that all the data is copied twice, once to the backup and once
2297 to the new layout on the array, so this type of reshape will go very
2300 If the reshape is interrupted for any reason, this backup file must be
2302 .B "mdadm --assemble"
2303 so the array can be reassembled. Consequently the file cannot be
2304 stored on the device being reshaped.
2309 A write-intent bitmap can be added to, or removed from, an active
2310 array. Either internal bitmaps, or bitmaps stored in a separate file,
2311 can be added. Note that if you add a bitmap stored in a file which is
2312 in a filesystem that is on the RAID array being affected, the system
2313 will deadlock. The bitmap must be on a separate filesystem.
2315 .SH INCREMENTAL MODE
2319 .B mdadm \-\-incremental
2325 .B mdadm \-\-incremental \-\-fail
2329 .B mdadm \-\-incremental \-\-rebuild\-map
2332 .B mdadm \-\-incremental \-\-run \-\-scan
2335 This mode is designed to be used in conjunction with a device
2336 discovery system. As devices are found in a system, they can be
2338 .B "mdadm \-\-incremental"
2339 to be conditionally added to an appropriate array.
2341 Conversely, it can also be used with the
2343 flag to do just the opposite and find whatever array a particular device
2344 is part of and remove the device from that array.
2346 If the device passed is a
2348 device created by a previous call to
2350 then rather than trying to add that device to an array, all the arrays
2351 described by the metadata of the container will be started.
2354 performs a number of tests to determine if the device is part of an
2355 array, and which array it should be part of. If an appropriate array
2356 is found, or can be created,
2358 adds the device to the array and conditionally starts the array.
2362 will normally only add devices to an array which were previously working
2363 (active or spare) parts of that array. The support for automatic
2364 inclusion of a new drive as a spare in some array requires
2365 a configuration through POLICY in config file.
2369 makes are as follow:
2371 Is the device permitted by
2373 That is, is it listed in a
2375 line in that file. If
2377 is absent then the default it to allow any device. Similar if
2379 contains the special word
2381 then any device is allowed. Otherwise the device name given to
2383 must match one of the names or patterns in a
2388 Does the device have a valid md superblock. If a specific metadata
2389 version is request with
2393 then only that style of metadata is accepted, otherwise
2395 finds any known version of metadata. If no
2397 metadata is found, the device may be still added to an array
2398 as a spare if POLICY allows.
2402 Does the metadata match an expected array?
2403 The metadata can match in two ways. Either there is an array listed
2406 which identifies the array (either by UUID, by name, by device list,
2407 or by minor-number), or the array was created with a
2413 or on the command line.
2416 is not able to positively identify the array as belonging to the
2417 current host, the device will be rejected.
2421 keeps a list of arrays that it has partially assembled in
2422 .B /var/run/mdadm/map
2424 .B /var/run/mdadm.map
2425 if the directory doesn't exist. Or maybe even
2426 .BR /dev/.mdadm.map ).
2427 If no array exists which matches
2428 the metadata on the new device,
2430 must choose a device name and unit number. It does this based on any
2433 or any name information stored in the metadata. If this name
2434 suggests a unit number, that number will be used, otherwise a free
2435 unit number will be chosen. Normally
2437 will prefer to create a partitionable array, however if the
2441 suggests that a non-partitionable array is preferred, that will be
2444 If the array is not found in the config file and its metadata does not
2445 identify it as belonging to the "homehost", then
2447 will choose a name for the array which is certain not to conflict with
2448 any array which does belong to this host. It does this be adding an
2449 underscore and a small number to the name preferred by the metadata.
2451 Once an appropriate array is found or created and the device is added,
2453 must decide if the array is ready to be started. It will
2454 normally compare the number of available (non-spare) devices to the
2455 number of devices that the metadata suggests need to be active. If
2456 there are at least that many, the array will be started. This means
2457 that if any devices are missing the array will not be restarted.
2463 in which case the array will be run as soon as there are enough
2464 devices present for the data to be accessible. For a RAID1, that
2465 means one device will start the array. For a clean RAID5, the array
2466 will be started as soon as all but one drive is present.
2468 Note that neither of these approaches is really ideal. If it can
2469 be known that all device discovery has completed, then
2473 can be run which will try to start all arrays that are being
2474 incrementally assembled. They are started in "read-auto" mode in
2475 which they are read-only until the first write request. This means
2476 that no metadata updates are made and no attempt at resync or recovery
2477 happens. Further devices that are found before the first write can
2478 still be added safely.
2481 This section describes environment variables that affect how mdadm
2486 Setting this value to 1 will prevent mdadm from automatically launching
2487 mdmon. This variable is intended primarily for debugging mdadm/mdmon.
2493 does not create any device nodes in /dev, but leaves that task to
2497 appears not to be configured, or if this environment variable is set
2500 will create and devices that are needed.
2504 .B " mdadm \-\-query /dev/name-of-device"
2506 This will find out if a given device is a RAID array, or is part of
2507 one, and will provide brief information about the device.
2509 .B " mdadm \-\-assemble \-\-scan"
2511 This will assemble and start all arrays listed in the standard config
2512 file. This command will typically go in a system startup file.
2514 .B " mdadm \-\-stop \-\-scan"
2516 This will shut down all arrays that can be shut down (i.e. are not
2517 currently in use). This will typically go in a system shutdown script.
2519 .B " mdadm \-\-follow \-\-scan \-\-delay=120"
2521 If (and only if) there is an Email address or program given in the
2522 standard config file, then
2523 monitor the status of all arrays listed in that file by
2524 polling them ever 2 minutes.
2526 .B " mdadm \-\-create /dev/md0 \-\-level=1 \-\-raid\-devices=2 /dev/hd[ac]1"
2528 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
2531 .B " echo 'DEVICE /dev/hd*[0\-9] /dev/sd*[0\-9]' > mdadm.conf"
2533 .B " mdadm \-\-detail \-\-scan >> mdadm.conf"
2535 This will create a prototype config file that describes currently
2536 active arrays that are known to be made from partitions of IDE or SCSI drives.
2537 This file should be reviewed before being used as it may
2538 contain unwanted detail.
2540 .B " echo 'DEVICE /dev/hd[a\-z] /dev/sd*[a\-z]' > mdadm.conf"
2542 .B " mdadm \-\-examine \-\-scan \-\-config=mdadm.conf >> mdadm.conf"
2544 This will find arrays which could be assembled from existing IDE and
2545 SCSI whole drives (not partitions), and store the information in the
2546 format of a config file.
2547 This file is very likely to contain unwanted detail, particularly
2550 entries. It should be reviewed and edited before being used as an
2553 .B " mdadm \-\-examine \-\-brief \-\-scan \-\-config=partitions"
2555 .B " mdadm \-Ebsc partitions"
2557 Create a list of devices by reading
2558 .BR /proc/partitions ,
2559 scan these for RAID superblocks, and printout a brief listing of all
2562 .B " mdadm \-Ac partitions \-m 0 /dev/md0"
2564 Scan all partitions and devices listed in
2565 .BR /proc/partitions
2568 out of all such devices with a RAID superblock with a minor number of 0.
2570 .B " mdadm \-\-monitor \-\-scan \-\-daemonise > /var/run/mdadm"
2572 If config file contains a mail address or alert program, run mdadm in
2573 the background in monitor mode monitoring all md devices. Also write
2574 pid of mdadm daemon to
2575 .BR /var/run/mdadm .
2577 .B " mdadm \-Iq /dev/somedevice"
2579 Try to incorporate newly discovered device into some array as
2582 .B " mdadm \-\-incremental \-\-rebuild\-map \-\-run \-\-scan"
2584 Rebuild the array map from any current arrays, and then start any that
2587 .B " mdadm /dev/md4 --fail detached --remove detached"
2589 Any devices which are components of /dev/md4 will be marked as faulty
2590 and then remove from the array.
2592 .B " mdadm --grow /dev/md4 --level=6 --backup-file=/root/backup-md4
2596 which is currently a RAID5 array will be converted to RAID6. There
2597 should normally already be a spare drive attached to the array as a
2598 RAID6 needs one more drive than a matching RAID5.
2600 .B " mdadm --create /dev/md/ddf --metadata=ddf --raid-disks 6 /dev/sd[a-f]"
2602 Create a DDF array over 6 devices.
2604 .B " mdadm --create /dev/md/home -n3 -l5 -z 30000000 /dev/md/ddf"
2606 Create a RAID5 array over any 3 devices in the given DDF set. Use
2607 only 30 gigabytes of each device.
2609 .B " mdadm -A /dev/md/ddf1 /dev/sd[a-f]"
2611 Assemble a pre-exist ddf array.
2613 .B " mdadm -I /dev/md/ddf1"
2615 Assemble all arrays contained in the ddf array, assigning names as
2618 .B " mdadm \-\-create \-\-help"
2620 Provide help about the Create mode.
2622 .B " mdadm \-\-config \-\-help"
2624 Provide help about the format of the config file.
2626 .B " mdadm \-\-help"
2628 Provide general help.
2638 lists all active md devices with information about them.
2640 uses this to find arrays when
2642 is given in Misc mode, and to monitor array reconstruction
2647 The config file lists which devices may be scanned to see if
2648 they contain MD super block, and gives identifying information
2649 (e.g. UUID) about known MD arrays. See
2653 .SS /var/run/mdadm/map
2656 mode is used, this file gets a list of arrays currently being created.
2659 does not exist as a directory, then
2660 .B /var/run/mdadm.map
2663 is not available (as may be the case during early boot),
2665 is used on the basis that
2667 is usually available very early in boot.
2672 understand two sorts of names for array devices.
2674 The first is the so-called 'standard' format name, which matches the
2675 names used by the kernel and which appear in
2678 The second sort can be freely chosen, but must reside in
2680 When giving a device name to
2682 to create or assemble an array, either full path name such as
2686 can be given, or just the suffix of the second sort of name, such as
2692 chooses device names during auto-assembly or incremental assembly, it
2693 will sometimes add a small sequence number to the end of the name to
2694 avoid conflicted between multiple arrays that have the same name. If
2696 can reasonably determine that the array really is meant for this host,
2697 either by a hostname in the metadata, or by the presence of the array
2700 then it will leave off the suffix if possible.
2701 Also if the homehost is specified as
2704 will only use a suffix if a different array of the same name already
2705 exists or is listed in the config file.
2707 The standard names for non-partitioned arrays (the only sort of md
2708 array available in 2.4 and earlier) are of the form
2712 where NN is a number.
2713 The standard names for partitionable arrays (as available from 2.6
2714 onwards) are of the form
2718 Partition numbers should be indicated by added "pMM" to these, thus "/dev/md/d1p2".
2720 From kernel version, 2.6.28 the "non-partitioned array" can actually
2721 be partitioned. So the "md_dNN" names are no longer needed, and
2722 partitions such as "/dev/mdNNpXX" are possible.
2726 was previously known as
2730 is completely separate from the
2732 package, and does not use the
2734 configuration file at all.
2737 For further information on mdadm usage, MD and the various levels of
2740 .B http://linux\-raid.osdl.org/
2742 (based upon Jakob \(/Ostergaard's Software\-RAID.HOWTO)
2744 .\"for new releases of the RAID driver check out:
2747 .\".UR ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2748 .\"ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2753 .\".UR http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2754 .\"http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2757 The latest version of
2759 should always be available from
2761 .B http://www.kernel.org/pub/linux/utils/raid/mdadm/