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.
8 .TH MDADM 8 "" v3.0-devel3
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.
51 is also not true RAID, and it only involves one device. It
52 provides a layer over a true device that can be used to inject faults.
57 is a collection of devices that are
58 managed as a set. This is similar to the set of devices connected to
59 a hardware RAID controller. The set of devices may contain a number
60 of different RAID arrays each utilising some (or all) blocks from a
61 number of the devices in the set. For example, two devices in a 5-device set
62 might form a RAID1 used the whole devices. The remaining three might
63 have a RAID5 over the first half of each device, and a RAID0 over the
68 there is one set of metadata that describes all of
69 the arrays in the container. So when
73 device, it just represents the metadata. Other normal arrays (RAID1
74 etc) can be created inside that container.
78 .\"is a program that can be used to create, manage, and monitor
80 .\"such it provides a similar set of functionality to the
83 .\"The key differences between
90 .\"is a single program and not a collection of programs.
93 .\"can perform (almost) all of its functions without having a
94 .\"configuration file and does not use one by default. Also
96 .\"helps with management of the configuration
100 .\"can provide information about your arrays (through Query, Detail, and Examine)
107 .\".IR /etc/raidtab ,
110 .\"configuration file, at all. It has a different configuration file
111 .\"with a different format and a different purpose.
114 mdadm has several major modes of operation:
117 Assemble the components of a previously created
118 array into an active array. Components can be explicitly given
119 or can be searched for.
121 checks that the components
122 do form a bona fide array, and can, on request, fiddle superblock
123 information so as to assemble a faulty array.
127 Build an array that doesn't have per-device superblocks. For these
130 cannot differentiate between initial creation and subsequent assembly
131 of an array. It also cannot perform any checks that appropriate
132 components have been requested. Because of this, the
134 mode should only be used together with a complete understanding of
139 Create a new array with per-device superblocks.
141 .\"in several step create-add-add-run or it can all happen with one command.
144 .B "Follow or Monitor"
145 Monitor one or more md devices and act on any state changes. This is
146 only meaningful for raid1, 4, 5, 6, 10 or multipath arrays, as
147 only these have interesting state. raid0 or linear never have
148 missing, spare, or failed drives, so there is nothing to monitor.
152 Grow (or shrink) an array, or otherwise reshape it in some way.
153 Currently supported growth options including changing the active size
154 of component devices and changing the number of active devices in RAID
155 levels 1/4/5/6, as well as adding or removing a write-intent bitmap.
158 .B "Incremental Assembly"
159 Add a single device to an appropriate array. If the addition of the
160 device makes the array runnable, the array will be started.
161 This provides a convenient interface to a
163 system. As each device is detected,
165 has a chance to include it in some array as appropriate.
171 in this mode, then any arrays within that container will be assembled
176 This is for doing things to specific components of an array such as
177 adding new spares and removing faulty devices.
181 This is an 'everything else' mode that supports operations on active
182 arrays, operations on component devices such as erasing old superblocks, and
183 information gathering operations.
184 .\"This mode allows operations on independent devices such as examine MD
185 .\"superblocks, erasing old superblocks and stopping active arrays.
189 This mode does not act on a specific device or array, but rather it
190 requests the Linux Kernel to activate any auto-detected arrays.
193 .SH Options for selecting a mode are:
196 .BR \-A ", " \-\-assemble
197 Assemble a pre-existing array.
200 .BR \-B ", " \-\-build
201 Build a legacy array without superblocks.
204 .BR \-C ", " \-\-create
208 .BR \-F ", " \-\-follow ", " \-\-monitor
214 .BR \-G ", " \-\-grow
215 Change the size or shape of an active array.
218 .BR \-I ", " \-\-incremental
219 Add a single device into an appropriate array, and possibly start the array.
223 Request that the kernel starts any auto-detected arrays. This can only
226 is compiled into the kernel \(em not if it is a module.
227 Arrays can be auto-detected by the kernel if all the components are in
228 primary MS-DOS partitions with partition type
230 In-kernel autodetect is not recommended for new installations. Using
232 to detect and assemble arrays \(em possibly in an
234 \(em is substantially more flexible and should be preferred.
237 If a device is given before any options, or if the first option is
242 then the MANAGE mode is assume.
243 Anything other than these will cause the
247 .SH Options that are not mode-specific are:
250 .BR \-h ", " \-\-help
251 Display general help message or, after one of the above options, a
252 mode-specific help message.
256 Display more detailed help about command line parsing and some commonly
260 .BR \-V ", " \-\-version
261 Print version information for mdadm.
264 .BR \-v ", " \-\-verbose
265 Be more verbose about what is happening. This can be used twice to be
267 The extra verbosity currently only affects
268 .B \-\-detail \-\-scan
270 .BR "\-\-examine \-\-scan" .
273 .BR \-q ", " \-\-quiet
274 Avoid printing purely informative messages. With this,
276 will be silent unless there is something really important to report.
279 .BR \-b ", " \-\-brief
280 Be less verbose. This is used with
288 gives an intermediate level of verbosity.
291 .BR \-f ", " \-\-force
292 Be more forceful about certain operations. See the various modes for
293 the exact meaning of this option in different contexts.
296 .BR \-c ", " \-\-config=
297 Specify the config file. Default is to use
298 .BR /etc/mdadm.conf ,
299 or if that is missing then
300 .BR /etc/mdadm/mdadm.conf .
301 If the config file given is
303 then nothing will be read, but
305 will act as though the config file contained exactly
306 .B "DEVICE partitions containers"
309 to find a list of devices to scan, and
311 to find a list of containers to examine.
314 is given for the config file, then
316 will act as though the config file were empty.
319 .BR \-s ", " \-\-scan
322 for missing information.
323 In general, this option gives
325 permission to get any missing information (like component devices,
326 array devices, array identities, and alert destination) from the
327 configuration file (see previous option);
328 one exception is MISC mode when using
334 says to get a list of array devices from
338 .B \-e ", " \-\-metadata=
339 Declare the style of superblock (raid metadata) to be used. The
342 and to guess for other operations.
343 The default can be overridden by setting the
352 .IP "0, 0.90, default"
353 Use the original 0.90 format superblock. This format limits arrays to
354 28 component devices and limits component devices of levels 1 and
355 greater to 2 terabytes.
356 .IP "1, 1.0, 1.1, 1.2"
357 Use the new version-1 format superblock. This has few restrictions.
358 The different sub-versions store the superblock at different locations
359 on the device, either at the end (for 1.0), at the start (for 1.1) or
360 4K from the start (for 1.2).
362 Use the "Industry Standard" DDF (Disk Data Format) format. When
363 creating a DDF array a
365 will be created, and normal arrays can be created in that container.
367 Use the Intel(R) Matrix Storage Manager metadata format. This creates a
369 which is managed in a similar manner to DDF, and is supported by an
370 option-rom on some platforms:
372 .B http://www.intel.com/design/chipsets/matrixstorage_sb.htm
378 This will override any
380 setting in the config file and provides the identity of the host which
381 should be considered the home for any arrays.
383 When creating an array, the
385 will be recorded in the superblock. For version-1 superblocks, it will
386 be prefixed to the array name. For version-0.90 superblocks, part of
387 the SHA1 hash of the hostname will be stored in the later half of the
390 When reporting information about an array, any array which is tagged
391 for the given homehost will be reported as such.
393 When using Auto-Assemble, only arrays tagged for the given homehost
396 .SH For create, build, or grow:
399 .BR \-n ", " \-\-raid\-devices=
400 Specify the number of active devices in the array. This, plus the
401 number of spare devices (see below) must equal the number of
403 (including "\fBmissing\fP" devices)
404 that are listed on the command line for
406 Setting a value of 1 is probably
407 a mistake and so requires that
409 be specified first. A value of 1 will then be allowed for linear,
410 multipath, raid0 and raid1. It is never allowed for raid4 or raid5.
412 This number can only be changed using
414 for RAID1, RAID5 and RAID6 arrays, and only on kernels which provide
418 .BR \-x ", " \-\-spare\-devices=
419 Specify the number of spare (eXtra) devices in the initial array.
420 Spares can also be added
421 and removed later. The number of component devices listed
422 on the command line must equal the number of raid devices plus the
423 number of spare devices.
427 .BR \-z ", " \-\-size=
428 Amount (in Kibibytes) of space to use from each drive in RAID level 1/4/5/6.
429 This must be a multiple of the chunk size, and must leave about 128Kb
430 of space at the end of the drive for the RAID superblock.
431 If this is not specified
432 (as it normally is not) the smallest drive (or partition) sets the
433 size, though if there is a variance among the drives of greater than 1%, a warning is
436 This value can be set with
438 for RAID level 1/4/5/6. If the array was created with a size smaller
439 than the currently active drives, the extra space can be accessed
442 The size can be given as
444 which means to choose the largest size that fits on all current drives.
446 This value can not be used with
448 metadata such as DDF and IMSM.
451 .BR \-c ", " \-\-chunk=
452 Specify chunk size of kibibytes. The default is 64.
456 Specify rounding factor for linear array (==chunk size)
459 .BR \-l ", " \-\-level=
460 Set raid level. When used with
462 options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
463 raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty, container.
464 Obviously some of these are synonymous.
468 metadata type is requested, only the
470 level is permitted, and it does not need to be explicitly given.
474 only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid.
476 Not yet supported with
480 .BR \-p ", " \-\-layout=
481 This option configures the fine details of data layout for raid5,
482 and raid10 arrays, and controls the failure modes for
485 The layout of the raid5 parity block can be one of
486 .BR left\-asymmetric ,
487 .BR left\-symmetric ,
488 .BR right\-asymmetric ,
489 .BR right\-symmetric ,
490 .BR la ", " ra ", " ls ", " rs .
492 .BR left\-symmetric .
494 When setting the failure mode for level
497 .BR write\-transient ", " wt ,
498 .BR read\-transient ", " rt ,
499 .BR write\-persistent ", " wp ,
500 .BR read\-persistent ", " rp ,
502 .BR read\-fixable ", " rf ,
503 .BR clear ", " flush ", " none .
505 Each failure mode can be followed by a number, which is used as a period
506 between fault generation. Without a number, the fault is generated
507 once on the first relevant request. With a number, the fault will be
508 generated after that many requests, and will continue to be generated
509 every time the period elapses.
511 Multiple failure modes can be current simultaneously by using the
513 option to set subsequent failure modes.
515 "clear" or "none" will remove any pending or periodic failure modes,
516 and "flush" will clear any persistent faults.
518 To set the parity with
520 the level of the array ("faulty")
521 must be specified before the fault mode is specified.
523 Finally, the layout options for RAID10 are one of 'n', 'o' or 'f' followed
524 by a small number. The default is 'n2'. The supported options are:
527 signals 'near' copies. Multiple copies of one data block are at
528 similar offsets in different devices.
531 signals 'offset' copies. Rather than the chunks being duplicated
532 within a stripe, whole stripes are duplicated but are rotated by one
533 device so duplicate blocks are on different devices. Thus subsequent
534 copies of a block are in the next drive, and are one chunk further
539 (multiple copies have very different offsets).
540 See md(4) for more detail about 'near' and 'far'.
542 The number is the number of copies of each datablock. 2 is normal, 3
543 can be useful. This number can be at most equal to the number of
544 devices in the array. It does not need to divide evenly into that
545 number (e.g. it is perfectly legal to have an 'n2' layout for an array
546 with an odd number of devices).
552 (thus explaining the p of
556 .BR \-b ", " \-\-bitmap=
557 Specify a file to store a write-intent bitmap in. The file should not
560 is also given. The same file should be provided
561 when assembling the array. If the word
563 is given, then the bitmap is stored with the metadata on the array,
564 and so is replicated on all devices. If the word
568 mode, then any bitmap that is present is removed.
570 To help catch typing errors, the filename must contain at least one
571 slash ('/') if it is a real file (not 'internal' or 'none').
573 Note: external bitmaps are only known to work on ext2 and ext3.
574 Storing bitmap files on other filesystems may result in serious problems.
577 .BR \-\-bitmap\-chunk=
578 Set the chunksize of the bitmap. Each bit corresponds to that many
579 Kilobytes of storage.
580 When using a file based bitmap, the default is to use the smallest
581 size that is at-least 4 and requires no more than 2^21 chunks.
584 bitmap, the chunksize is automatically determined to make best use of
589 .BR \-W ", " \-\-write\-mostly
590 subsequent devices lists in a
595 command will be flagged as 'write-mostly'. This is valid for RAID1
596 only and means that the 'md' driver will avoid reading from these
597 devices if at all possible. This can be useful if mirroring over a
601 .BR \-\-write\-behind=
602 Specify that write-behind mode should be enabled (valid for RAID1
603 only). If an argument is specified, it will set the maximum number
604 of outstanding writes allowed. The default value is 256.
605 A write-intent bitmap is required in order to use write-behind
606 mode, and write-behind is only attempted on drives marked as
610 .BR \-\-assume\-clean
613 that the array pre-existed and is known to be clean. It can be useful
614 when trying to recover from a major failure as you can be sure that no
615 data will be affected unless you actually write to the array. It can
616 also be used when creating a RAID1 or RAID10 if you want to avoid the
617 initial resync, however this practice \(em while normally safe \(em is not
618 recommended. Use this only if you really know what you are doing.
621 .BR \-\-backup\-file=
624 is used to increase the number of
625 raid-devices in a RAID5 if there are no spare devices available.
626 See the section below on RAID_DEVICE CHANGES. The file should be
627 stored on a separate device, not on the raid array being reshaped.
630 .BR \-N ", " \-\-name=
633 for the array. This is currently only effective when creating an
634 array with a version-1 superblock. The name is a simple textual
635 string that can be used to identify array components when assembling.
641 run the array, even if some of the components
642 appear to be active in another array or filesystem. Normally
644 will ask for confirmation before including such components in an
645 array. This option causes that question to be suppressed.
648 .BR \-f ", " \-\-force
651 accept the geometry and layout specified without question. Normally
653 will not allow creation of an array with only one device, and will try
654 to create a raid5 array with one missing drive (as this makes the
655 initial resync work faster). With
658 will not try to be so clever.
661 .BR \-a ", " "\-\-auto{=no,yes,md,mdp,part,p}{NN}"
662 Instruct mdadm to create the device file if needed, possibly allocating
663 an unused minor number. "md" causes a non-partitionable array
664 to be used. "mdp", "part" or "p" causes a partitionable array (2.6 and
665 later) to be used. "yes" requires the named md device to have
666 a 'standard' format, and the type and minor number will be determined
667 from this. See DEVICE NAMES below.
669 The argument can also come immediately after
674 is not given on the command line or in the config file, then
680 is also given, then any
682 entries in the config file will override the
684 instruction given on the command line.
686 For partitionable arrays,
688 will create the device file for the whole array and for the first 4
689 partitions. A different number of partitions can be specified at the
690 end of this option (e.g.
692 If the device name ends with a digit, the partition names add a 'p',
693 and a number, e.g. "/dev/md/home1p3". If there is no
694 trailing digit, then the partition names just have a number added,
695 e.g. "/dev/md/scratch3".
697 If the md device name is in a 'standard' format as described in DEVICE
698 NAMES, then it will be created, if necessary, with the appropriate
699 number based on that name. If the device name is not in one of these
700 formats, then a unused minor number will be allocated. The minor
701 number will be considered unused if there is no active array for that
702 number, and there is no entry in /dev for that number and with a
703 non-standard name. Name that are not in 'standard' format are only
704 allowed in "/dev/md/".
707 \".BR \-\-symlink = no
712 \"to create devices in
714 \"it will also create symlinks from
716 \"with names starting with
722 \"to suppress this, or
724 \"to enforce this even if it is suppressing
731 .BR \-u ", " \-\-uuid=
732 uuid of array to assemble. Devices which don't have this uuid are
736 .BR \-m ", " \-\-super\-minor=
737 Minor number of device that array was created for. Devices which
738 don't have this minor number are excluded. If you create an array as
739 /dev/md1, then all superblocks will contain the minor number 1, even if
740 the array is later assembled as /dev/md2.
742 Giving the literal word "dev" for
746 to use the minor number of the md device that is being assembled.
749 .B \-\-super\-minor=dev
750 will look for super blocks with a minor number of 0.
753 .BR \-N ", " \-\-name=
754 Specify the name of the array to assemble. This must be the name
755 that was specified when creating the array. It must either match
756 the name stored in the superblock exactly, or it must match
759 prefixed to the start of the given name.
762 .BR \-f ", " \-\-force
763 Assemble the array even if some superblocks appear out-of-date
767 Attempt to start the array even if fewer drives were given than were
768 present last time the array was active. Normally if not all the
769 expected drives are found and
771 is not used, then the array will be assembled but not started.
774 an attempt will be made to start it anyway.
778 This is the reverse of
780 in that it inhibits the startup of array unless all expected drives
781 are present. This is only needed with
783 and can be used if the physical connections to devices are
784 not as reliable as you would like.
787 .BR \-a ", " "\-\-auto{=no,yes,md,mdp,part}"
788 See this option under Create and Build options.
791 .BR \-b ", " \-\-bitmap=
792 Specify the bitmap file that was given when the array was created. If
795 bitmap, there is no need to specify this when assembling the array.
798 .BR \-\-backup\-file=
801 was used to grow the number of raid-devices in a RAID5, and the system
802 crashed during the critical section, then the same
806 to allow possibly corrupted data to be restored.
809 .BR \-U ", " \-\-update=
810 Update the superblock on each device while assembling the array. The
811 argument given to this flag can be one of
825 option will adjust the superblock of an array what was created on a Sparc
826 machine running a patched 2.2 Linux kernel. This kernel got the
827 alignment of part of the superblock wrong. You can use the
828 .B "\-\-examine \-\-sparc2.2"
831 to see what effect this would have.
835 option will update the
837 field on each superblock to match the minor number of the array being
839 This can be useful if
841 reports a different "Preferred Minor" to
843 In some cases this update will be performed automatically
844 by the kernel driver. In particular the update happens automatically
845 at the first write to an array with redundancy (RAID level 1 or
846 greater) on a 2.6 (or later) kernel.
850 option will change the uuid of the array. If a UUID is given with the
852 option that UUID will be used as a new UUID and will
854 be used to help identify the devices in the array.
857 is given, a random UUID is chosen.
861 option will change the
863 of the array as stored in the superblock. This is only supported for
864 version-1 superblocks.
868 option will change the
870 as recorded in the superblock. For version-0 superblocks, this is the
871 same as updating the UUID.
872 For version-1 superblocks, this involves updating the name.
876 option will cause the array to be marked
878 meaning that any redundancy in the array (e.g. parity for raid5,
879 copies for raid1) may be incorrect. This will cause the raid system
880 to perform a "resync" pass to make sure that all redundant information
885 option allows arrays to be moved between machines with different
887 When assembling such an array for the first time after a move, giving
888 .B "\-\-update=byteorder"
891 to expect superblocks to have their byteorder reversed, and will
892 correct that order before assembling the array. This is only valid
893 with original (Version 0.90) superblocks.
897 option will correct the summaries in the superblock. That is the
898 counts of total, working, active, failed, and spare devices.
902 will rarely be of use. It applies to version 1.1 and 1.2 metadata
903 only (where the metadata is at the start of the device) and is only
904 useful when the component device has changed size (typically become
905 larger). The version 1 metadata records the amount of the device that
906 can be used to store data, so if a device in a version 1.1 or 1.2
907 array becomes larger, the metadata will still be visible, but the
908 extra space will not. In this case it might be useful to assemble the
910 .BR \-\-update=devicesize .
913 to determine the maximum usable amount of space on each device and
914 update the relevant field in the metadata.
917 .B \-\-auto\-update\-homehost
918 This flag is only meaningful with auto-assembly (see discussion below).
919 In that situation, if no suitable arrays are found for this homehost,
921 will rescan for any arrays at all and will assemble them and update the
922 homehost to match the current host.
928 hot-add listed devices.
932 re-add a device that was recently removed from an array.
935 .BR \-r ", " \-\-remove
936 remove listed devices. They must not be active. i.e. they should
937 be failed or spare devices. As well as the name of a device file
946 The first causes all failed device to be removed. The second causes
947 any device which is no longer connected to the system (i.e an 'open'
950 to be removed. This will only succeed for devices that are spares or
951 have already been marked as failed.
954 .BR \-f ", " \-\-fail
955 mark listed devices as faulty.
956 As well as the name of a device file, the word
958 can be given. This will cause any device that has been detached from
959 the system to be marked as failed. It can then be removed.
967 .BR \-\-write\-mostly
968 Subsequent devices that are added or re-added will have the 'write-mostly'
969 flag set. This is only valid for RAID! and means that the 'md' driver
970 will avoid reading from these devices if possible.
973 Subsequent devices that are added or re-added will have the 'write-mostly'
978 Each of these options require that the first device listed is the array
979 to be acted upon, and the remainder are component devices to be added,
980 removed, or marked as faulty. Several different operations can be
981 specified for different devices, e.g.
983 mdadm /dev/md0 \-\-add /dev/sda1 \-\-fail /dev/sdb1 \-\-remove /dev/sdb1
985 Each operation applies to all devices listed until the next
988 If an array is using a write-intent bitmap, then devices which have
989 been removed can be re-added in a way that avoids a full
990 reconstruction but instead just updates the blocks that have changed
991 since the device was removed. For arrays with persistent metadata
992 (superblocks) this is done automatically. For arrays created with
994 mdadm needs to be told that this device we removed recently with
997 Devices can only be removed from an array if they are not in active
998 use, i.e. that must be spares or failed devices. To remove an active
999 device, it must first be marked as
1005 .BR \-Q ", " \-\-query
1006 Examine a device to see
1007 (1) if it is an md device and (2) if it is a component of an md
1009 Information about what is discovered is presented.
1012 .BR \-D ", " \-\-detail
1013 Print detail of one or more md devices.
1016 .BR \-\-detail\-platform
1017 Print detail of the platform's raid capabilities (firmware / hardware
1018 topology) for a given metadata format.
1021 .BR \-Y ", " \-\-export
1026 output will be formatted as
1028 pairs for easy import into the environment.
1031 .BR \-E ", " \-\-examine
1032 Print content of md superblock on device(s).
1035 If an array was created on a 2.2 Linux kernel patched with RAID
1036 support, the superblock will have been created incorrectly, or at
1037 least incompatibly with 2.4 and later kernels. Using the
1041 will fix the superblock before displaying it. If this appears to do
1042 the right thing, then the array can be successfully assembled using
1043 .BR "\-\-assemble \-\-update=sparc2.2" .
1046 .BR \-X ", " \-\-examine\-bitmap
1047 Report information about a bitmap file.
1048 The argument is either an external bitmap file or an array component
1049 in case of an internal bitmap.
1052 .BR \-R ", " \-\-run
1053 start a partially built array.
1056 .BR \-S ", " \-\-stop
1057 deactivate array, releasing all resources.
1060 .BR \-o ", " \-\-readonly
1061 mark array as readonly.
1064 .BR \-w ", " \-\-readwrite
1065 mark array as readwrite.
1068 .B \-\-zero\-superblock
1069 If the device contains a valid md superblock, the block is
1070 overwritten with zeros. With
1072 the block where the superblock would be is overwritten even if it
1073 doesn't appear to be valid.
1076 .BR \-t ", " \-\-test
1081 is set to reflect the status of the device.
1084 .BR \-W ", " \-\-wait
1085 For each md device given, wait for any resync, recovery, or reshape
1086 activity to finish before returning.
1088 will return with success if it actually waited for every device
1089 listed, otherwise it will return failure.
1093 For each md device given, or each device in /proc/mdstat if
1095 is given, arrange for the array to be marked clean as soon as possible.
1096 Also, quiesce resync so that the monitor for external metadata arrays
1097 (mdmon) has an opportunity to checkpoint the resync position.
1099 will return with success if the array uses external metadata and we
1100 successfully waited. For native arrays this returns immediately as the
1101 kernel handles both dirty-clean transitions and resync checkpointing in
1102 the kernel at shutdown. No action is taken if safe-mode handling is
1105 .SH For Incremental Assembly mode:
1107 .BR \-\-rebuild\-map ", " \-r
1108 Rebuild the map file
1109 .RB ( /var/run/mdadm/map )
1112 uses to help track which arrays are currently being assembled.
1115 .BR \-\-run ", " \-R
1116 Run any array assembled as soon as a minimal number of devices are
1117 available, rather than waiting until all expected devices are present.
1121 This allows the hot-plug system to prevent arrays from running when it knows
1122 that more disks may arrive later in the discovery process.
1125 .BR \-\-scan ", " \-s
1126 Only meaningful with
1130 file for arrays that are being incrementally assembled and will try to
1131 start any that are not already started. If any such array is listed
1134 as requiring an external bitmap, that bitmap will be attached first.
1136 .SH For Monitor mode:
1138 .BR \-m ", " \-\-mail
1139 Give a mail address to send alerts to.
1142 .BR \-p ", " \-\-program ", " \-\-alert
1143 Give a program to be run whenever an event is detected.
1146 .BR \-y ", " \-\-syslog
1147 Cause all events to be reported through 'syslog'. The messages have
1148 facility of 'daemon' and varying priorities.
1151 .BR \-d ", " \-\-delay
1152 Give a delay in seconds.
1154 polls the md arrays and then waits this many seconds before polling
1155 again. The default is 60 seconds.
1158 .BR \-f ", " \-\-daemonise
1161 to run as a background daemon if it decides to monitor anything. This
1162 causes it to fork and run in the child, and to disconnect form the
1163 terminal. The process id of the child is written to stdout.
1166 which will only continue monitoring if a mail address or alert program
1167 is found in the config file.
1170 .BR \-i ", " \-\-pid\-file
1173 is running in daemon mode, write the pid of the daemon process to
1174 the specified file, instead of printing it on standard output.
1177 .BR \-1 ", " \-\-oneshot
1178 Check arrays only once. This will generate
1180 events and more significantly
1186 .B " mdadm \-\-monitor \-\-scan \-1"
1188 from a cron script will ensure regular notification of any degraded arrays.
1191 .BR \-t ", " \-\-test
1194 alert for every array found at startup. This alert gets mailed and
1195 passed to the alert program. This can be used for testing that alert
1196 message do get through successfully.
1202 .B mdadm \-\-assemble
1203 .I md-device options-and-component-devices...
1206 .B mdadm \-\-assemble \-\-scan
1207 .I md-devices-and-options...
1210 .B mdadm \-\-assemble \-\-scan
1214 This usage assembles one or more raid arrays from pre-existing components.
1215 For each array, mdadm needs to know the md device, the identity of the
1216 array, and a number of component-devices. These can be found in a number of ways.
1218 In the first usage example (without the
1220 the first device given is the md device.
1221 In the second usage example, all devices listed are treated as md
1222 devices and assembly is attempted.
1223 In the third (where no devices are listed) all md devices that are
1224 listed in the configuration file are assembled. Then any arrays that
1225 can be found on unused devices will also be assembled.
1227 If precisely one device is listed, but
1233 was given and identity information is extracted from the configuration file.
1235 The identity can be given with the
1239 option, will be taken from the md-device record in the config file, or
1240 will be taken from the super block of the first component-device
1241 listed on the command line.
1243 Devices can be given on the
1245 command line or in the config file. Only devices which have an md
1246 superblock which contains the right identity will be considered for
1249 The config file is only used if explicitly named with
1251 or requested with (a possibly implicit)
1256 .B /etc/mdadm/mdadm.conf
1261 is not given, then the config file will only be used to find the
1262 identity of md arrays.
1264 Normally the array will be started after it is assembled. However if
1266 is not given and insufficient drives were listed to start a complete
1267 (non-degraded) array, then the array is not started (to guard against
1268 usage errors). To insist that the array be started in this case (as
1269 may work for RAID1, 4, 5, 6, or 10), give the
1273 If the md device does not exist, then it will be created providing the
1274 intent is clear. i.e. the name must be in a standard form, or the
1276 option must be given to clarify how and whether the device should be
1278 This can be useful for handling partitioned devices (which don't have
1279 a stable device number \(em it can change after a reboot) and when using
1280 "udev" to manage your
1282 tree (udev cannot handle md devices because of the unusual device
1283 initialisation conventions).
1285 If the option to "auto" is "mdp" or "part" or (on the command line
1286 only) "p", then mdadm will create a partitionable array, using the
1287 first free one that is not in use and does not already have an entry
1288 in /dev (apart from numeric /dev/md* entries).
1290 If the option to "auto" is "yes" or "md" or (on the command line)
1291 nothing, then mdadm will create a traditional, non-partitionable md
1294 It is expected that the "auto" functionality will be used to create
1295 device entries with meaningful names such as "/dev/md/home" or
1296 "/dev/md/root", rather than names based on the numerical array number.
1298 When using option "auto" to create a partitionable array, the device
1299 files for the first 4 partitions are also created. If a different
1300 number is required it can be simply appended to the auto option.
1301 e.g. "auto=part8". Partition names are created by appending a digit
1302 string to the device name, with an intervening "p" if the device name
1307 option is also available in Build and Create modes. As those modes do
1308 not use a config file, the "auto=" config option does not apply to
1316 and no devices are listed,
1318 will first attempt to assemble all the arrays listed in the config
1321 It will then look further for possible arrays and will try to assemble
1322 anything that it finds. Arrays which are tagged as belonging to the given
1323 homehost will be assembled and started normally. Arrays which do not
1324 obviously belong to this host are given names that are expected not to
1325 conflict with anything local, and are started "read-auto" so that
1326 nothing is written to any device until the array is written to. i.e.
1327 automatic resync etc is delayed.
1331 finds a consistent set of devices that look like they should comprise
1332 an array, and if the superblock is tagged as belonging to the given
1333 home host, it will automatically choose a device name and try to
1334 assemble the array. If the array uses version-0.90 metadata, then the
1336 number as recorded in the superblock is used to create a name in
1340 If the array uses version-1 metadata, then the
1342 from the superblock is used to similarly create a name in
1344 (the name will have any 'host' prefix stripped first).
1348 cannot find any array for the given host at all, and if
1349 .B \-\-auto\-update\-homehost
1352 will search again for any array (not just an array created for this
1353 host) and will assemble each assuming
1354 .BR \-\-update=homehost .
1355 This will change the host tag in the superblock so that on the next run,
1356 these arrays will be found without the second pass. The intention of
1357 this feature is to support transitioning a set of md arrays to using
1360 The reason for requiring arrays to be tagged with the homehost for
1361 auto assembly is to guard against problems that can arise when moving
1362 devices from one host to another.
1372 .BI \-\-raid\-devices= Z
1376 This usage is similar to
1378 The difference is that it creates an array without a superblock. With
1379 these arrays there is no difference between initially creating the array and
1380 subsequently assembling the array, except that hopefully there is useful
1381 data there in the second case.
1383 The level may raid0, linear, multipath, or faulty, or one of their
1384 synonyms. All devices must be listed and the array will be started
1396 .BI \-\-raid\-devices= Z
1400 This usage will initialise a new md array, associate some devices with
1401 it, and activate the array.
1405 option is given (as described in more detail in the section on
1406 Assemble mode), then the md device will be created with a suitable
1407 device number if necessary.
1409 As devices are added, they are checked to see if they contain raid
1410 superblocks or filesystems. They are also checked to see if the variance in
1411 device size exceeds 1%.
1413 If any discrepancy is found, the array will not automatically be run, though
1416 can override this caution.
1418 To create a "degraded" array in which some devices are missing, simply
1419 give the word "\fBmissing\fP"
1420 in place of a device name. This will cause
1422 to leave the corresponding slot in the array empty.
1423 For a RAID4 or RAID5 array at most one slot can be
1424 "\fBmissing\fP"; for a RAID6 array at most two slots.
1425 For a RAID1 array, only one real device needs to be given. All of the
1429 When creating a RAID5 array,
1431 will automatically create a degraded array with an extra spare drive.
1432 This is because building the spare into a degraded array is in general faster than resyncing
1433 the parity on a non-degraded, but not clean, array. This feature can
1434 be overridden with the
1438 When creating an array with version-1 metadata a name for the array is
1440 If this is not given with the
1444 will choose a name based on the last component of the name of the
1445 device being created. So if
1447 is being created, then the name
1452 is being created, then the name
1456 When creating a partition based array, using
1458 with version-1.x metadata, the partition type should be set to
1460 (non fs-data). This type selection allows for greater precision since
1461 using any other [RAID auto-detect (0xFD) or a GNU/Linux partition (0x83)],
1462 might create problems in the event of array recovery through a live cdrom.
1464 A new array will normally get a randomly assigned 128bit UUID which is
1465 very likely to be unique. If you have a specific need, you can choose
1466 a UUID for the array by giving the
1468 option. Be warned that creating two arrays with the same UUID is a
1469 recipe for disaster. Also, using
1471 when creating a v0.90 array will silently override any
1476 .\"option is given, it is not necessary to list any component-devices in this command.
1477 .\"They can be added later, before a
1481 .\"is given, the apparent size of the smallest drive given is used.
1483 When creating an array within a
1486 can be given either the list of devices to use, or simply the name of
1487 the container. The former case gives control over which devices in
1488 the container will be used for the array. The latter case allows
1490 to automatically choose which devices to use based on how much spare
1493 The General Management options that are valid with
1498 insist on running the array even if some devices look like they might
1503 start the array readonly \(em not supported yet.
1511 .I options... devices...
1514 This usage will allow individual devices in an array to be failed,
1515 removed or added. It is possible to perform multiple operations with
1516 on command. For example:
1518 .B " mdadm /dev/md0 \-f /dev/hda1 \-r /dev/hda1 \-a /dev/hda1"
1524 and will then remove it from the array and finally add it back
1525 in as a spare. However only one md array can be affected by a single
1536 MISC mode includes a number of distinct operations that
1537 operate on distinct devices. The operations are:
1540 The device is examined to see if it is
1541 (1) an active md array, or
1542 (2) a component of an md array.
1543 The information discovered is reported.
1547 The device should be an active md device.
1549 will display a detailed description of the array.
1553 will cause the output to be less detailed and the format to be
1554 suitable for inclusion in
1555 .BR /etc/mdadm.conf .
1558 will normally be 0 unless
1560 failed to get useful information about the device(s); however, if the
1562 option is given, then the exit status will be:
1566 The array is functioning normally.
1569 The array has at least one failed device.
1572 The array has multiple failed devices such that it is unusable.
1575 There was an error while trying to get information about the device.
1579 .B \-\-detail\-platform
1580 Print detail of the platform's raid capabilities (firmware / hardware
1581 topology). If the metadata is specified with
1585 then the return status will be:
1589 metadata successfully enumerated its platform components on this system
1592 metadata is platform independent
1595 metadata failed to find its platform components on this system
1600 The device should be a component of an md array.
1602 will read the md superblock of the device and display the contents.
1607 is given, then multiple devices that are components of the one array
1608 are grouped together and reported in a single entry suitable
1610 .BR /etc/mdadm.conf .
1614 without listing any devices will cause all devices listed in the
1615 config file to be examined.
1619 The devices should be active md arrays which will be deactivated, as
1620 long as they are not currently in use.
1624 This will fully activate a partially assembled md array.
1628 This will mark an active array as read-only, providing that it is
1629 not currently being used.
1635 array back to being read/write.
1639 For all operations except
1642 will cause the operation to be applied to all arrays listed in
1647 causes all devices listed in the config file to be examined.
1654 .B mdadm \-\-monitor
1655 .I options... devices...
1660 to periodically poll a number of md arrays and to report on any events
1663 will never exit once it decides that there are arrays to be checked,
1664 so it should normally be run in the background.
1666 As well as reporting events,
1668 may move a spare drive from one array to another if they are in the
1671 and if the destination array has a failed drive but no spares.
1673 If any devices are listed on the command line,
1675 will only monitor those devices. Otherwise all arrays listed in the
1676 configuration file will be monitored. Further, if
1678 is given, then any other md devices that appear in
1680 will also be monitored.
1682 The result of monitoring the arrays is the generation of events.
1683 These events are passed to a separate program (if specified) and may
1684 be mailed to a given E-mail address.
1686 When passing events to a program, the program is run once for each event,
1687 and is given 2 or 3 command-line arguments: the first is the
1688 name of the event (see below), the second is the name of the
1689 md device which is affected, and the third is the name of a related
1690 device if relevant (such as a component device that has failed).
1694 is given, then a program or an E-mail address must be specified on the
1695 command line or in the config file. If neither are available, then
1697 will not monitor anything.
1701 will continue monitoring as long as something was found to monitor. If
1702 no program or email is given, then each event is reported to
1705 The different events are:
1709 .B DeviceDisappeared
1710 An md array which previously was configured appears to no longer be
1711 configured. (syslog priority: Critical)
1715 was told to monitor an array which is RAID0 or Linear, then it will
1717 .B DeviceDisappeared
1718 with the extra information
1720 This is because RAID0 and Linear do not support the device-failed,
1721 hot-spare and resync operations which are monitored.
1725 An md array started reconstruction. (syslog priority: Warning)
1731 is 20, 40, 60, or 80, this indicates that rebuild has passed that many
1732 percentage of the total. (syslog priority: Warning)
1736 An md array that was rebuilding, isn't any more, either because it
1737 finished normally or was aborted. (syslog priority: Warning)
1741 An active component device of an array has been marked as
1742 faulty. (syslog priority: Critical)
1746 A spare component device which was being rebuilt to replace a faulty
1747 device has failed. (syslog priority: Critical)
1751 A spare component device which was being rebuilt to replace a faulty
1752 device has been successfully rebuilt and has been made active.
1753 (syslog priority: Info)
1757 A new md array has been detected in the
1759 file. (syslog priority: Info)
1763 A newly noticed array appears to be degraded. This message is not
1766 notices a drive failure which causes degradation, but only when
1768 notices that an array is degraded when it first sees the array.
1769 (syslog priority: Critical)
1773 A spare drive has been moved from one array in a
1775 to another to allow a failed drive to be replaced.
1776 (syslog priority: Info)
1782 has been told, via the config file, that an array should have a certain
1783 number of spare devices, and
1785 detects that it has fewer than this number when it first sees the
1786 array, it will report a
1789 (syslog priority: Warning)
1793 An array was found at startup, and the
1796 (syslog priority: Info)
1806 cause Email to be sent. All events cause the program to be run.
1807 The program is run with two or three arguments: the event
1808 name, the array device and possibly a second device.
1810 Each event has an associated array device (e.g.
1812 and possibly a second device. For
1817 the second device is the relevant component device.
1820 the second device is the array that the spare was moved from.
1824 to move spares from one array to another, the different arrays need to
1825 be labeled with the same
1827 in the configuration file. The
1829 name can be any string; it is only necessary that different spare
1830 groups use different names.
1834 detects that an array in a spare group has fewer active
1835 devices than necessary for the complete array, and has no spare
1836 devices, it will look for another array in the same spare group that
1837 has a full complement of working drive and a spare. It will then
1838 attempt to remove the spare from the second drive and add it to the
1840 If the removal succeeds but the adding fails, then it is added back to
1844 The GROW mode is used for changing the size or shape of an active
1846 For this to work, the kernel must support the necessary change.
1847 Various types of growth are being added during 2.6 development,
1848 including restructuring a raid5 array to have more active devices.
1850 Currently the only support available is to
1852 change the "size" attribute
1853 for RAID1, RAID5 and RAID6.
1855 increase the "raid\-devices" attribute of RAID1, RAID5, and RAID6.
1857 add a write-intent bitmap to any array which supports these bitmaps, or
1858 remove a write-intent bitmap from such an array.
1861 GROW mode is not currently supported for
1863 or arrays inside containers.
1866 Normally when an array is built the "size" it taken from the smallest
1867 of the drives. If all the small drives in an arrays are, one at a
1868 time, removed and replaced with larger drives, then you could have an
1869 array of large drives with only a small amount used. In this
1870 situation, changing the "size" with "GROW" mode will allow the extra
1871 space to start being used. If the size is increased in this way, a
1872 "resync" process will start to make sure the new parts of the array
1875 Note that when an array changes size, any filesystem that may be
1876 stored in the array will not automatically grow to use the space. The
1877 filesystem will need to be explicitly told to use the extra space.
1879 .SS RAID-DEVICES CHANGES
1881 A RAID1 array can work with any number of devices from 1 upwards
1882 (though 1 is not very useful). There may be times which you want to
1883 increase or decrease the number of active devices. Note that this is
1884 different to hot-add or hot-remove which changes the number of
1887 When reducing the number of devices in a RAID1 array, the slots which
1888 are to be removed from the array must already be vacant. That is, the
1889 devices which were in those slots must be failed and removed.
1891 When the number of devices is increased, any hot spares that are
1892 present will be activated immediately.
1894 Increasing the number of active devices in a RAID5 is much more
1895 effort. Every block in the array will need to be read and written
1896 back to a new location. From 2.6.17, the Linux Kernel is able to do
1897 this safely, including restart and interrupted "reshape".
1899 When relocating the first few stripes on a raid5, it is not possible
1900 to keep the data on disk completely consistent and crash-proof. To
1901 provide the required safety, mdadm disables writes to the array while
1902 this "critical section" is reshaped, and takes a backup of the data
1903 that is in that section. This backup is normally stored in any spare
1904 devices that the array has, however it can also be stored in a
1905 separate file specified with the
1907 option. If this option is used, and the system does crash during the
1908 critical period, the same file must be passed to
1910 to restore the backup and reassemble the array.
1914 A write-intent bitmap can be added to, or removed from, an active
1915 array. Either internal bitmaps, or bitmaps stored in a separate file,
1916 can be added. Note that if you add a bitmap stored in a file which is
1917 in a filesystem that is on the raid array being affected, the system
1918 will deadlock. The bitmap must be on a separate filesystem.
1920 .SH INCREMENTAL MODE
1924 .B mdadm \-\-incremental
1930 .B mdadm \-\-incremental \-\-rebuild
1933 .B mdadm \-\-incremental \-\-run \-\-scan
1937 This mode is designed to be used in conjunction with a device
1938 discovery system. As devices are found in a system, they can be
1940 .B "mdadm \-\-incremental"
1941 to be conditionally added to an appropriate array.
1943 If the device passed is a
1945 device created by a previous call to
1947 then rather than trying to add that device to an array, all the arrays
1948 described by the metadata of the container will be started.
1951 performs a number of tests to determine if the device is part of an
1952 array, and which array it should be part of. If an appropriate array
1953 is found, or can be created,
1955 adds the device to the array and conditionally starts the array.
1959 will only add devices to an array which were previously working
1960 (active or spare) parts of that array. It does not currently support
1961 automatic inclusion of a new drive as a spare in some array.
1965 makes are as follow:
1967 Is the device permitted by
1969 That is, is it listed in a
1971 line in that file. If
1973 is absent then the default it to allow any device. Similar if
1975 contains the special word
1977 then any device is allowed. Otherwise the device name given to
1979 must match one of the names or patterns in a
1984 Does the device have a valid md superblock. If a specific metadata
1985 version is request with
1989 then only that style of metadata is accepted, otherwise
1991 finds any known version of metadata. If no
1993 metadata is found, the device is rejected.
1996 Does the metadata match an expected array?
1997 The metadata can match in two ways. Either there is an array listed
2000 which identifies the array (either by UUID, by name, by device list,
2001 or by minor-number), or the array was created with a
2007 or on the command line.
2010 is not able to positively identify the array as belonging to the
2011 current host, the device will be rejected.
2015 keeps a list of arrays that it has partially assembled in
2016 .B /var/run/mdadm/map
2018 .B /var/run/mdadm.map
2019 if the directory doesn't exist). If no array exists which matches
2020 the metadata on the new device,
2022 must choose a device name and unit number. It does this based on any
2025 or any name information stored in the metadata. If this name
2026 suggests a unit number, that number will be used, otherwise a free
2027 unit number will be chosen. Normally
2029 will prefer to create a partitionable array, however if the
2033 suggests that a non-partitionable array is preferred, that will be
2037 Once an appropriate array is found or created and the device is added,
2039 must decide if the array is ready to be started. It will
2040 normally compare the number of available (non-spare) devices to the
2041 number of devices that the metadata suggests need to be active. If
2042 there are at least that many, the array will be started. This means
2043 that if any devices are missing the array will not be restarted.
2049 in which case the array will be run as soon as there are enough
2050 devices present for the data to be accessible. For a raid1, that
2051 means one device will start the array. For a clean raid5, the array
2052 will be started as soon as all but one drive is present.
2054 Note that neither of these approaches is really ideal. If it can
2055 be known that all device discovery has completed, then
2059 can be run which will try to start all arrays that are being
2060 incrementally assembled. They are started in "read-auto" mode in
2061 which they are read-only until the first write request. This means
2062 that no metadata updates are made and no attempt at resync or recovery
2063 happens. Further devices that are found before the first write can
2064 still be added safely.
2068 This section describes environment variables that affect how mdadm
2073 Setting this value to 1 will prevent mdadm from automatically launching
2074 mdmon. This variable is intended primarily for debugging mdadm/mdmon.
2080 does not create any device nodes in /dev, but leaves that task to
2084 appears not to be configured, or if this environment variable is set
2087 will create and devices that are needed.
2091 .B " mdadm \-\-query /dev/name-of-device"
2093 This will find out if a given device is a raid array, or is part of
2094 one, and will provide brief information about the device.
2096 .B " mdadm \-\-assemble \-\-scan"
2098 This will assemble and start all arrays listed in the standard config
2099 file. This command will typically go in a system startup file.
2101 .B " mdadm \-\-stop \-\-scan"
2103 This will shut down all arrays that can be shut down (i.e. are not
2104 currently in use). This will typically go in a system shutdown script.
2106 .B " mdadm \-\-follow \-\-scan \-\-delay=120"
2108 If (and only if) there is an Email address or program given in the
2109 standard config file, then
2110 monitor the status of all arrays listed in that file by
2111 polling them ever 2 minutes.
2113 .B " mdadm \-\-create /dev/md0 \-\-level=1 \-\-raid\-devices=2 /dev/hd[ac]1"
2115 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
2118 .B " echo 'DEVICE /dev/hd*[0\-9] /dev/sd*[0\-9]' > mdadm.conf"
2120 .B " mdadm \-\-detail \-\-scan >> mdadm.conf"
2122 This will create a prototype config file that describes currently
2123 active arrays that are known to be made from partitions of IDE or SCSI drives.
2124 This file should be reviewed before being used as it may
2125 contain unwanted detail.
2127 .B " echo 'DEVICE /dev/hd[a\-z] /dev/sd*[a\-z]' > mdadm.conf"
2129 .B " mdadm \-\-examine \-\-scan \-\-config=mdadm.conf >> mdadm.conf"
2131 This will find arrays which could be assembled from existing IDE and
2132 SCSI whole drives (not partitions), and store the information in the
2133 format of a config file.
2134 This file is very likely to contain unwanted detail, particularly
2137 entries. It should be reviewed and edited before being used as an
2140 .B " mdadm \-\-examine \-\-brief \-\-scan \-\-config=partitions"
2142 .B " mdadm \-Ebsc partitions"
2144 Create a list of devices by reading
2145 .BR /proc/partitions ,
2146 scan these for RAID superblocks, and printout a brief listing of all
2149 .B " mdadm \-Ac partitions \-m 0 /dev/md0"
2151 Scan all partitions and devices listed in
2152 .BR /proc/partitions
2155 out of all such devices with a RAID superblock with a minor number of 0.
2157 .B " mdadm \-\-monitor \-\-scan \-\-daemonise > /var/run/mdadm"
2159 If config file contains a mail address or alert program, run mdadm in
2160 the background in monitor mode monitoring all md devices. Also write
2161 pid of mdadm daemon to
2162 .BR /var/run/mdadm .
2164 .B " mdadm \-Iq /dev/somedevice"
2166 Try to incorporate newly discovered device into some array as
2169 .B " mdadm \-\-incremental \-\-rebuild \-\-run \-\-scan"
2171 Rebuild the array map from any current arrays, and then start any that
2174 .B " mdadm /dev/md4 --fail detached --remove detached"
2176 Any devices which are components of /dev/md4 will be marked as faulty
2177 and then remove from the array.
2179 .B " mdadm --create /dev/md/ddf --metadata=ddf --raid-disks 6 /dev/sd[a-f]"
2181 Create a DDF array over 6 devices.
2183 .B " mdadm --create /dev/md/home -n3 -l5 -z 30000000 /dev/md/ddf"
2185 Create a raid5 array over any 3 devices in the given DDF set. Use
2186 only 30 gigabytes of each device.
2188 .B " mdadm -A /dev/md/ddf1 /dev/sd[a-f]"
2190 Assemble a pre-exist ddf array.
2192 .B " mdadm -I /dev/md/ddf1"
2194 Assemble all arrays contained in the ddf array, assigning names as
2197 .B " mdadm \-\-create \-\-help"
2199 Provide help about the Create mode.
2201 .B " mdadm \-\-config \-\-help"
2203 Provide help about the format of the config file.
2205 .B " mdadm \-\-help"
2207 Provide general help.
2218 lists all active md devices with information about them.
2220 uses this to find arrays when
2222 is given in Misc mode, and to monitor array reconstruction
2228 The config file lists which devices may be scanned to see if
2229 they contain MD super block, and gives identifying information
2230 (e.g. UUID) about known MD arrays. See
2234 .SS /var/run/mdadm/map
2237 mode is used, this file gets a list of arrays currently being created.
2240 does not exist as a directory, then
2241 .B /var/run/mdadm.map
2247 understand two sorts of names for array devices.
2249 The first is the so-called 'standard' format name, which matches the
2250 names used by the kernel and which appear in
2253 The second sort can be freely chosen, but must reside in
2255 When giving a device name to
2257 to create or assemble an array, either full path name such as
2261 can be given, or just the suffix of the second sort of name, such as
2267 chooses device names during auto-assembly, it will normally add a
2268 small sequence number to the end of the name to avoid conflicted
2269 between multiple arrays that have the same name. If
2271 can reasonably determine that the array really is meant for this host,
2272 either by a hostname in the metadata, or by the presence of the array
2273 in /etc/mdadm.conf, then it will leave of the suffix if possible.
2275 The standard names for non-partitioned arrays (the only sort of md
2276 array available in 2.4 and earlier) are of the form
2280 where NN is a number.
2281 The standard names for partitionable arrays (as available from 2.6
2282 onwards) are of the form
2286 Partition numbers should be indicated by added "pMM" to these, thus "/dev/md/d1p2".
2288 From kernel version, 2.6.28 the "non-partitioned array" can actually
2289 be partitioned. So the "md_dNN" names are no longer needed, and
2290 partitions such as "/dev/mdNNpXX" are possible.
2294 was previously known as
2298 is completely separate from the
2300 package, and does not use the
2302 configuration file at all.
2305 For further information on mdadm usage, MD and the various levels of
2308 .B http://linux\-raid.osdl.org/
2310 (based upon Jakob \(/Ostergaard's Software\-RAID.HOWTO)
2312 .\"for new releases of the RAID driver check out:
2315 .\".UR ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2316 .\"ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2321 .\".UR http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2322 .\"http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2325 The latest version of
2327 should always be available from
2329 .B http://www.kernel.org/pub/linux/utils/raid/mdadm/