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.
144 in this mode, then any arrays within that container will be assembled
149 This is for doing things to specific components of an array such as
150 adding new spares and removing faulty devices.
154 This is an 'everything else' mode that supports operations on active
155 arrays, operations on component devices such as erasing old superblocks, and
156 information gathering operations.
157 .\"This mode allows operations on independent devices such as examine MD
158 .\"superblocks, erasing old superblocks and stopping active arrays.
162 This mode does not act on a specific device or array, but rather it
163 requests the Linux Kernel to activate any auto-detected arrays.
166 .SH Options for selecting a mode are:
169 .BR \-A ", " \-\-assemble
170 Assemble a pre-existing array.
173 .BR \-B ", " \-\-build
174 Build a legacy array without superblocks.
177 .BR \-C ", " \-\-create
181 .BR \-F ", " \-\-follow ", " \-\-monitor
187 .BR \-G ", " \-\-grow
188 Change the size or shape of an active array.
191 .BR \-I ", " \-\-incremental
192 Add a single device into an appropriate array, and possibly start the array.
196 Request that the kernel starts any auto-detected arrays. This can only
199 is compiled into the kernel \(em not if it is a module.
200 Arrays can be auto-detected by the kernel if all the components are in
201 primary MS-DOS partitions with partition type
203 and all use v0.90 metadata.
204 In-kernel autodetect is not recommended for new installations. Using
206 to detect and assemble arrays \(em possibly in an
208 \(em is substantially more flexible and should be preferred.
211 If a device is given before any options, or if the first option is
216 then the MANAGE mode is assumed.
217 Anything other than these will cause the
221 .SH Options that are not mode-specific are:
224 .BR \-h ", " \-\-help
225 Display general help message or, after one of the above options, a
226 mode-specific help message.
230 Display more detailed help about command line parsing and some commonly
234 .BR \-V ", " \-\-version
235 Print version information for mdadm.
238 .BR \-v ", " \-\-verbose
239 Be more verbose about what is happening. This can be used twice to be
241 The extra verbosity currently only affects
242 .B \-\-detail \-\-scan
244 .BR "\-\-examine \-\-scan" .
247 .BR \-q ", " \-\-quiet
248 Avoid printing purely informative messages. With this,
250 will be silent unless there is something really important to report.
253 .BR \-f ", " \-\-force
254 Be more forceful about certain operations. See the various modes for
255 the exact meaning of this option in different contexts.
258 .BR \-c ", " \-\-config=
259 Specify the config file. Default is to use
260 .BR /etc/mdadm.conf ,
261 or if that is missing then
262 .BR /etc/mdadm/mdadm.conf .
263 If the config file given is
265 then nothing will be read, but
267 will act as though the config file contained exactly
268 .B "DEVICE partitions containers"
271 to find a list of devices to scan, and
273 to find a list of containers to examine.
276 is given for the config file, then
278 will act as though the config file were empty.
281 .BR \-s ", " \-\-scan
284 for missing information.
285 In general, this option gives
287 permission to get any missing information (like component devices,
288 array devices, array identities, and alert destination) from the
289 configuration file (see previous option);
290 one exception is MISC mode when using
296 says to get a list of array devices from
300 .BR \-e ", " \-\-metadata=
301 Declare the style of RAID metadata (superblock) to be used. The
302 default is {DEFAULT_METADATA} for
304 and to guess for other operations.
305 The default can be overridden by setting the
314 .ie '{DEFAULT_METADATA}'0.90'
315 .IP "0, 0.90, default"
319 Use the original 0.90 format superblock. This format limits arrays to
320 28 component devices and limits component devices of levels 1 and
321 greater to 2 terabytes.
322 .ie '{DEFAULT_METADATA}'0.90'
323 .IP "1, 1.0, 1.1, 1.2"
325 .IP "1, 1.0, 1.1, 1.2 default"
327 Use the new version-1 format superblock. This has few restrictions.
328 The different sub-versions store the superblock at different locations
329 on the device, either at the end (for 1.0), at the start (for 1.1) or
330 4K from the start (for 1.2). "1" is equivalent to "1.0".
331 'if '{DEFAULT_METADATA}'1.2' "default" is equivalent to "1.2".
333 Use the "Industry Standard" DDF (Disk Data Format) format defined by
335 When creating a DDF array a
337 will be created, and normal arrays can be created in that container.
339 Use the Intel(R) Matrix Storage Manager metadata format. This creates a
341 which is managed in a similar manner to DDF, and is supported by an
342 option-rom on some platforms:
344 .B http://www.intel.com/design/chipsets/matrixstorage_sb.htm
350 This will override any
352 setting in the config file and provides the identity of the host which
353 should be considered the home for any arrays.
355 When creating an array, the
357 will be recorded in the metadata. For version-1 superblocks, it will
358 be prefixed to the array name. For version-0.90 superblocks, part of
359 the SHA1 hash of the hostname will be stored in the later half of the
362 When reporting information about an array, any array which is tagged
363 for the given homehost will be reported as such.
365 When using Auto-Assemble, only arrays tagged for the given homehost
366 will be allowed to use 'local' names (i.e. not ending in '_' followed
367 by a digit string). See below under
368 .BR "Auto Assembly" .
370 .SH For create, build, or grow:
373 .BR \-n ", " \-\-raid\-devices=
374 Specify the number of active devices in the array. This, plus the
375 number of spare devices (see below) must equal the number of
377 (including "\fBmissing\fP" devices)
378 that are listed on the command line for
380 Setting a value of 1 is probably
381 a mistake and so requires that
383 be specified first. A value of 1 will then be allowed for linear,
384 multipath, RAID0 and RAID1. It is never allowed for RAID4, RAID5 or RAID6.
386 This number can only be changed using
388 for RAID1, RAID4, RAID5 and RAID6 arrays, and only on kernels which provide
389 the necessary support.
392 .BR \-x ", " \-\-spare\-devices=
393 Specify the number of spare (eXtra) devices in the initial array.
394 Spares can also be added
395 and removed later. The number of component devices listed
396 on the command line must equal the number of RAID devices plus the
397 number of spare devices.
400 .BR \-z ", " \-\-size=
401 Amount (in Kibibytes) of space to use from each drive in RAID levels 1/4/5/6.
402 This must be a multiple of the chunk size, and must leave about 128Kb
403 of space at the end of the drive for the RAID superblock.
404 If this is not specified
405 (as it normally is not) the smallest drive (or partition) sets the
406 size, though if there is a variance among the drives of greater than 1%, a warning is
409 This value can be set with
411 for RAID level 1/4/5/6. If the array was created with a size smaller
412 than the currently active drives, the extra space can be accessed
415 The size can be given as
417 which means to choose the largest size that fits on all current drives.
419 This value can not be used with
421 metadata such as DDF and IMSM.
424 .BR \-Z ", " \-\-array-size=
425 This is only meaningful with
427 and its effect is not persistent: when the array is stopped an
428 restarted the default array size will be restored.
430 Setting the array-size causes the array to appear smaller to programs
431 that access the data. This is particularly needed before reshaping an
432 array so that it will be smaller. As the reshape is not reversible,
433 but setting the size with
435 is, it is required that the array size is reduced as appropriate
436 before the number of devices in the array is reduced.
439 .BR \-c ", " \-\-chunk=
440 Specify chunk size of kibibytes. The default when creating an
441 array is 512KB. To ensure compatibility with earlier versions, the
442 default when Building and array with no persistent metadata is 64KB.
443 This is only meaningful for RAID0, RAID4, RAID5, RAID6, and RAID10.
447 Specify rounding factor for a Linear array. The size of each
448 component will be rounded down to a multiple of this size.
449 This is a synonym for
451 but highlights the different meaning for Linear as compared to other
452 RAID levels. The default is 64K if a kernel earlier than 2.6.16 is in
453 use, and is 0K (i.e. no rounding) in later kernels.
456 .BR \-l ", " \-\-level=
457 Set RAID level. When used with
459 options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
460 raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty, container.
461 Obviously some of these are synonymous.
465 metadata type is requested, only the
467 level is permitted, and it does not need to be explicitly given.
471 only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid.
475 to change the RAID level in some cases. See LEVEL CHANGES below.
478 .BR \-p ", " \-\-layout=
479 This option configures the fine details of data layout for RAID5, RAID6,
480 and RAID10 arrays, and controls the failure modes for
483 The layout of the RAID5 parity block can be one of
484 .BR left\-asymmetric ,
485 .BR left\-symmetric ,
486 .BR right\-asymmetric ,
487 .BR right\-symmetric ,
488 .BR la ", " ra ", " ls ", " rs .
490 .BR left\-symmetric .
492 It is also possibly to cause RAID5 to use a RAID4-like layout by
498 Finally for RAID5 there are DDF\-compatible layouts,
499 .BR ddf\-zero\-restart ,
500 .BR ddf\-N\-restart ,
502 .BR ddf\-N\-continue .
504 These same layouts are available for RAID6. There are also 4 layouts
505 that will provide an intermediate stage for converting between RAID5
506 and RAID6. These provide a layout which is identical to the
507 corresponding RAID5 layout on the first N\-1 devices, and has the 'Q'
508 syndrome (the second 'parity' block used by RAID6) on the last device.
510 .BR left\-symmetric\-6 ,
511 .BR right\-symmetric\-6 ,
512 .BR left\-asymmetric\-6 ,
513 .BR right\-asymmetric\-6 ,
515 .BR parity\-first\-6 .
517 When setting the failure mode for level
520 .BR write\-transient ", " wt ,
521 .BR read\-transient ", " rt ,
522 .BR write\-persistent ", " wp ,
523 .BR read\-persistent ", " rp ,
525 .BR read\-fixable ", " rf ,
526 .BR clear ", " flush ", " none .
528 Each failure mode can be followed by a number, which is used as a period
529 between fault generation. Without a number, the fault is generated
530 once on the first relevant request. With a number, the fault will be
531 generated after that many requests, and will continue to be generated
532 every time the period elapses.
534 Multiple failure modes can be current simultaneously by using the
536 option to set subsequent failure modes.
538 "clear" or "none" will remove any pending or periodic failure modes,
539 and "flush" will clear any persistent faults.
541 Finally, the layout options for RAID10 are one of 'n', 'o' or 'f' followed
542 by a small number. The default is 'n2'. The supported options are:
545 signals 'near' copies. Multiple copies of one data block are at
546 similar offsets in different devices.
549 signals 'offset' copies. Rather than the chunks being duplicated
550 within a stripe, whole stripes are duplicated but are rotated by one
551 device so duplicate blocks are on different devices. Thus subsequent
552 copies of a block are in the next drive, and are one chunk further
557 (multiple copies have very different offsets).
558 See md(4) for more detail about 'near', 'offset', and 'far'.
560 The number is the number of copies of each datablock. 2 is normal, 3
561 can be useful. This number can be at most equal to the number of
562 devices in the array. It does not need to divide evenly into that
563 number (e.g. it is perfectly legal to have an 'n2' layout for an array
564 with an odd number of devices).
566 When an array is converted between RAID5 and RAID6 an intermediate
567 RAID6 layout is used in which the second parity block (Q) is always on
568 the last device. To convert a RAID5 to RAID6 and leave it in this new
569 layout (which does not require re-striping) use
570 .BR \-\-layout=preserve .
571 This will try to avoid any restriping.
573 The converse of this is
574 .B \-\-layout=normalise
575 which will change a non-standard RAID6 layout into a more standard
582 (thus explaining the p of
586 .BR \-b ", " \-\-bitmap=
587 Specify a file to store a write-intent bitmap in. The file should not
590 is also given. The same file should be provided
591 when assembling the array. If the word
593 is given, then the bitmap is stored with the metadata on the array,
594 and so is replicated on all devices. If the word
598 mode, then any bitmap that is present is removed.
600 To help catch typing errors, the filename must contain at least one
601 slash ('/') if it is a real file (not 'internal' or 'none').
603 Note: external bitmaps are only known to work on ext2 and ext3.
604 Storing bitmap files on other filesystems may result in serious problems.
607 .BR \-\-bitmap\-chunk=
608 Set the chunksize of the bitmap. Each bit corresponds to that many
609 Kilobytes of storage.
610 When using a file based bitmap, the default is to use the smallest
611 size that is at-least 4 and requires no more than 2^21 chunks.
614 bitmap, the chunksize defaults to 64Meg, or larger if necessary to
615 fit the bitmap into the available space.
618 .BR \-W ", " \-\-write\-mostly
619 subsequent devices listed in a
624 command will be flagged as 'write-mostly'. This is valid for RAID1
625 only and means that the 'md' driver will avoid reading from these
626 devices if at all possible. This can be useful if mirroring over a
630 .BR \-\-write\-behind=
631 Specify that write-behind mode should be enabled (valid for RAID1
632 only). If an argument is specified, it will set the maximum number
633 of outstanding writes allowed. The default value is 256.
634 A write-intent bitmap is required in order to use write-behind
635 mode, and write-behind is only attempted on drives marked as
639 .BR \-\-assume\-clean
642 that the array pre-existed and is known to be clean. It can be useful
643 when trying to recover from a major failure as you can be sure that no
644 data will be affected unless you actually write to the array. It can
645 also be used when creating a RAID1 or RAID10 if you want to avoid the
646 initial resync, however this practice \(em while normally safe \(em is not
647 recommended. Use this only if you really know what you are doing.
649 When the devices that will be part of a new array were filled
650 with zeros before creation the operator knows the array is
651 actually clean. If that is the case, such as after running
652 badblocks, this argument can be used to tell mdadm the
653 facts the operator knows.
656 .BR \-\-backup\-file=
659 is used to increase the number of
660 raid-devices in a RAID5 if there are no spare devices available.
661 See the GROW MODE section below on RAID\-DEVICES CHANGES. The file
662 should be stored on a separate device, not on the RAID array being
666 .BR \-\-array-size= ", " \-Z
667 Set the size of the array which is seen by users of the device such as
668 filesystems. This can be less that the real size, but never greater.
669 The size set this way does not persist across restarts of the array.
671 This is most useful when reducing the number of devices in a RAID5 or
672 RAID6. Such arrays require the array-size to be reduced before a
673 reshape can be performed that reduces the real size.
677 restores the apparent size of the array to be whatever the real
678 amount of available space is.
681 .BR \-N ", " \-\-name=
684 for the array. This is currently only effective when creating an
685 array with a version-1 superblock, or an array in a DDF container.
686 The name is a simple textual string that can be used to identify array
687 components when assembling. If name is needed but not specified, it
688 is taken from the basename of the device that is being created.
700 run the array, even if some of the components
701 appear to be active in another array or filesystem. Normally
703 will ask for confirmation before including such components in an
704 array. This option causes that question to be suppressed.
707 .BR \-f ", " \-\-force
710 accept the geometry and layout specified without question. Normally
712 will not allow creation of an array with only one device, and will try
713 to create a RAID5 array with one missing drive (as this makes the
714 initial resync work faster). With
717 will not try to be so clever.
720 .BR \-a ", " "\-\-auto{=yes,md,mdp,part,p}{NN}"
721 Instruct mdadm how to create the device file if needed, possibly allocating
722 an unused minor number. "md" causes a non-partitionable array
723 to be used (though since Linux 2.6.28, these array devices are in fact
724 partitionable). "mdp", "part" or "p" causes a partitionable array (2.6 and
725 later) to be used. "yes" requires the named md device to have
726 a 'standard' format, and the type and minor number will be determined
727 from this. With mdadm 3.0, device creation is normally left up to
729 so this option is unlikely to be needed.
730 See DEVICE NAMES below.
732 The argument can also come immediately after
737 is not given on the command line or in the config file, then
743 is also given, then any
745 entries in the config file will override the
747 instruction given on the command line.
749 For partitionable arrays,
751 will create the device file for the whole array and for the first 4
752 partitions. A different number of partitions can be specified at the
753 end of this option (e.g.
755 If the device name ends with a digit, the partition names add a 'p',
757 .IR /dev/md/home1p3 .
758 If there is no trailing digit, then the partition names just have a
760 .IR /dev/md/scratch3 .
762 If the md device name is in a 'standard' format as described in DEVICE
763 NAMES, then it will be created, if necessary, with the appropriate
764 device number based on that name. If the device name is not in one of these
765 formats, then a unused device number will be allocated. The device
766 number will be considered unused if there is no active array for that
767 number, and there is no entry in /dev for that number and with a
768 non-standard name. Names that are not in 'standard' format are only
769 allowed in "/dev/md/".
773 .\".BR \-\-symlink = no
778 .\"to create devices in
780 .\"it will also create symlinks from
782 .\"with names starting with
788 .\"to suppress this, or
789 .\".B \-\-symlink=yes
790 .\"to enforce this even if it is suppressing
798 .BR \-u ", " \-\-uuid=
799 uuid of array to assemble. Devices which don't have this uuid are
803 .BR \-m ", " \-\-super\-minor=
804 Minor number of device that array was created for. Devices which
805 don't have this minor number are excluded. If you create an array as
806 /dev/md1, then all superblocks will contain the minor number 1, even if
807 the array is later assembled as /dev/md2.
809 Giving the literal word "dev" for
813 to use the minor number of the md device that is being assembled.
816 .B \-\-super\-minor=dev
817 will look for super blocks with a minor number of 0.
820 is only relevant for v0.90 metadata, and should not normally be used.
826 .BR \-N ", " \-\-name=
827 Specify the name of the array to assemble. This must be the name
828 that was specified when creating the array. It must either match
829 the name stored in the superblock exactly, or it must match
832 prefixed to the start of the given name.
835 .BR \-f ", " \-\-force
836 Assemble the array even if the metadata on some devices appears to be
839 cannot find enough working devices to start the array, but can find
840 some devices that are recorded as having failed, then it will mark
841 those devices as working so that the array can be started.
842 An array which requires
844 to be started may contain data corruption. Use it carefully.
848 Attempt to start the array even if fewer drives were given than were
849 present last time the array was active. Normally if not all the
850 expected drives are found and
852 is not used, then the array will be assembled but not started.
855 an attempt will be made to start it anyway.
859 This is the reverse of
861 in that it inhibits the startup of array unless all expected drives
862 are present. This is only needed with
864 and can be used if the physical connections to devices are
865 not as reliable as you would like.
868 .BR \-a ", " "\-\-auto{=no,yes,md,mdp,part}"
869 See this option under Create and Build options.
872 .BR \-b ", " \-\-bitmap=
873 Specify the bitmap file that was given when the array was created. If
876 bitmap, there is no need to specify this when assembling the array.
879 .BR \-\-backup\-file=
882 was used to grow the number of raid-devices in a RAID5, and the system
883 crashed during the critical section, then the same
887 to allow possibly corrupted data to be restored.
890 .BR \-U ", " \-\-update=
891 Update the superblock on each device while assembling the array. The
892 argument given to this flag can be one of
906 option will adjust the superblock of an array what was created on a Sparc
907 machine running a patched 2.2 Linux kernel. This kernel got the
908 alignment of part of the superblock wrong. You can use the
909 .B "\-\-examine \-\-sparc2.2"
912 to see what effect this would have.
916 option will update the
918 field on each superblock to match the minor number of the array being
920 This can be useful if
922 reports a different "Preferred Minor" to
924 In some cases this update will be performed automatically
925 by the kernel driver. In particular the update happens automatically
926 at the first write to an array with redundancy (RAID level 1 or
927 greater) on a 2.6 (or later) kernel.
931 option will change the uuid of the array. If a UUID is given with the
933 option that UUID will be used as a new UUID and will
935 be used to help identify the devices in the array.
938 is given, a random UUID is chosen.
942 option will change the
944 of the array as stored in the superblock. This is only supported for
945 version-1 superblocks.
949 option will change the
951 as recorded in the superblock. For version-0 superblocks, this is the
952 same as updating the UUID.
953 For version-1 superblocks, this involves updating the name.
957 option will cause the array to be marked
959 meaning that any redundancy in the array (e.g. parity for RAID5,
960 copies for RAID1) may be incorrect. This will cause the RAID system
961 to perform a "resync" pass to make sure that all redundant information
966 option allows arrays to be moved between machines with different
968 When assembling such an array for the first time after a move, giving
969 .B "\-\-update=byteorder"
972 to expect superblocks to have their byteorder reversed, and will
973 correct that order before assembling the array. This is only valid
974 with original (Version 0.90) superblocks.
978 option will correct the summaries in the superblock. That is the
979 counts of total, working, active, failed, and spare devices.
983 will rarely be of use. It applies to version 1.1 and 1.2 metadata
984 only (where the metadata is at the start of the device) and is only
985 useful when the component device has changed size (typically become
986 larger). The version 1 metadata records the amount of the device that
987 can be used to store data, so if a device in a version 1.1 or 1.2
988 array becomes larger, the metadata will still be visible, but the
989 extra space will not. In this case it might be useful to assemble the
991 .BR \-\-update=devicesize .
994 to determine the maximum usable amount of space on each device and
995 update the relevant field in the metadata.
999 .B \-\-auto\-update\-homehost
1000 This flag is only meaningful with auto-assembly (see discussion below).
1001 In that situation, if no suitable arrays are found for this homehost,
1003 will rescan for any arrays at all and will assemble them and update the
1004 homehost to match the current host.
1007 .SH For Manage mode:
1010 .BR \-a ", " \-\-add
1011 hot-add listed devices. For arrays with redundancy, the listed
1012 devices become available as spares. If the array is degraded, it will
1013 immediately start recovering data on to one of these spares.
1017 re-add a device that was recently removed from an array. This is only
1018 needed for arrays that have be built (i.e. with
1020 For created arrays, devices are always re-added if that is possible.
1021 When re-adding a device, if nothing has changed on the array since the
1022 device was removed, no recovery is performed. Also, if the array has
1023 a write-intent bitmap, then the recovery performed after a re-add will
1024 be limited to those blocks which, according to the bitmap, might have
1025 changed since the device was removed.
1028 .BR \-r ", " \-\-remove
1029 remove listed devices. They must not be active. i.e. they should
1030 be failed or spare devices. As well as the name of a device file
1039 The first causes all failed device to be removed. The second causes
1040 any device which is no longer connected to the system (i.e an 'open'
1043 to be removed. This will only succeed for devices that are spares or
1044 have already been marked as failed.
1047 .BR \-f ", " \-\-fail
1048 mark listed devices as faulty.
1049 As well as the name of a device file, the word
1051 can be given. This will cause any device that has been detached from
1052 the system to be marked as failed. It can then be removed.
1060 .BR \-\-write\-mostly
1061 Subsequent devices that are added or re-added will have the 'write-mostly'
1062 flag set. This is only valid for RAID1 and means that the 'md' driver
1063 will avoid reading from these devices if possible.
1066 Subsequent devices that are added or re-added will have the 'write-mostly'
1070 Each of these options requires that the first device listed is the array
1071 to be acted upon, and the remainder are component devices to be added,
1072 removed, marked as faulty, etc. Several different operations can be
1073 specified for different devices, e.g.
1075 mdadm /dev/md0 \-\-add /dev/sda1 \-\-fail /dev/sdb1 \-\-remove /dev/sdb1
1077 Each operation applies to all devices listed until the next
1080 If an array is using a write-intent bitmap, then devices which have
1081 been removed can be re-added in a way that avoids a full
1082 reconstruction but instead just updates the blocks that have changed
1083 since the device was removed. For arrays with persistent metadata
1084 (superblocks) this is done automatically. For arrays created with
1086 mdadm needs to be told that this device we removed recently with
1089 Devices can only be removed from an array if they are not in active
1090 use, i.e. that must be spares or failed devices. To remove an active
1091 device, it must first be marked as
1097 .BR \-Q ", " \-\-query
1098 Examine a device to see
1099 (1) if it is an md device and (2) if it is a component of an md
1101 Information about what is discovered is presented.
1104 .BR \-D ", " \-\-detail
1105 Print details of one or more md devices.
1108 .BR \-\-detail\-platform
1109 Print details of the platform's RAID capabilities (firmware / hardware
1110 topology) for a given metadata format.
1113 .BR \-Y ", " \-\-export
1118 output will be formatted as
1120 pairs for easy import into the environment.
1123 .BR \-E ", " \-\-examine
1124 Print contents of the metadata stored on the named device(s).
1125 Note the contrast between
1130 applies to devices which are components of an array, while
1132 applies to a whole array which is currently active.
1135 If an array was created on a SPARC machine with a 2.2 Linux kernel
1136 patched with RAID support, the superblock will have been created
1137 incorrectly, or at least incompatibly with 2.4 and later kernels.
1142 will fix the superblock before displaying it. If this appears to do
1143 the right thing, then the array can be successfully assembled using
1144 .BR "\-\-assemble \-\-update=sparc2.2" .
1147 .BR \-X ", " \-\-examine\-bitmap
1148 Report information about a bitmap file.
1149 The argument is either an external bitmap file or an array component
1150 in case of an internal bitmap. Note that running this on an array
1153 does not report the bitmap for that array.
1156 .BR \-R ", " \-\-run
1157 start a partially assembled array. If
1159 did not find enough devices to fully start the array, it might leaving
1160 it partially assembled. If you wish, you can then use
1162 to start the array in degraded mode.
1165 .BR \-S ", " \-\-stop
1166 deactivate array, releasing all resources.
1169 .BR \-o ", " \-\-readonly
1170 mark array as readonly.
1173 .BR \-w ", " \-\-readwrite
1174 mark array as readwrite.
1177 .B \-\-zero\-superblock
1178 If the device contains a valid md superblock, the block is
1179 overwritten with zeros. With
1181 the block where the superblock would be is overwritten even if it
1182 doesn't appear to be valid.
1185 .BR \-t ", " \-\-test
1190 is set to reflect the status of the device. See below in
1195 .BR \-W ", " \-\-wait
1196 For each md device given, wait for any resync, recovery, or reshape
1197 activity to finish before returning.
1199 will return with success if it actually waited for every device
1200 listed, otherwise it will return failure.
1204 For each md device given, or each device in /proc/mdstat if
1206 is given, arrange for the array to be marked clean as soon as possible.
1207 Also, quiesce resync so that the monitor for external metadata arrays
1208 (mdmon) has an opportunity to checkpoint the resync position.
1210 will return with success if the array uses external metadata and we
1211 successfully waited. For native arrays this returns immediately as the
1212 kernel handles both dirty-clean transitions and resync checkpointing in
1213 the kernel at shutdown. No action is taken if safe-mode handling is
1216 .SH For Incremental Assembly mode:
1218 .BR \-\-rebuild\-map ", " \-r
1219 Rebuild the map file
1220 .RB ( /var/run/mdadm/map )
1223 uses to help track which arrays are currently being assembled.
1226 .BR \-\-run ", " \-R
1227 Run any array assembled as soon as a minimal number of devices are
1228 available, rather than waiting until all expected devices are present.
1231 .BR \-\-scan ", " \-s
1232 Only meaningful with
1236 file for arrays that are being incrementally assembled and will try to
1237 start any that are not already started. If any such array is listed
1240 as requiring an external bitmap, that bitmap will be attached first.
1242 .SH For Monitor mode:
1244 .BR \-m ", " \-\-mail
1245 Give a mail address to send alerts to.
1248 .BR \-p ", " \-\-program ", " \-\-alert
1249 Give a program to be run whenever an event is detected.
1252 .BR \-y ", " \-\-syslog
1253 Cause all events to be reported through 'syslog'. The messages have
1254 facility of 'daemon' and varying priorities.
1257 .BR \-d ", " \-\-delay
1258 Give a delay in seconds.
1260 polls the md arrays and then waits this many seconds before polling
1261 again. The default is 60 seconds. Since 2.6.16, there is no need to
1262 reduce this as the kernel alerts
1264 immediately when there is any change.
1267 .BR \-r ", " \-\-increment
1268 Give a percentage increment.
1270 will generate RebuildNN events with the given percentage increment.
1273 .BR \-f ", " \-\-daemonise
1276 to run as a background daemon if it decides to monitor anything. This
1277 causes it to fork and run in the child, and to disconnect from the
1278 terminal. The process id of the child is written to stdout.
1281 which will only continue monitoring if a mail address or alert program
1282 is found in the config file.
1285 .BR \-i ", " \-\-pid\-file
1288 is running in daemon mode, write the pid of the daemon process to
1289 the specified file, instead of printing it on standard output.
1292 .BR \-1 ", " \-\-oneshot
1293 Check arrays only once. This will generate
1295 events and more significantly
1301 .B " mdadm \-\-monitor \-\-scan \-1"
1303 from a cron script will ensure regular notification of any degraded arrays.
1306 .BR \-t ", " \-\-test
1309 alert for every array found at startup. This alert gets mailed and
1310 passed to the alert program. This can be used for testing that alert
1311 message do get through successfully.
1317 .B mdadm \-\-assemble
1318 .I md-device options-and-component-devices...
1321 .B mdadm \-\-assemble \-\-scan
1322 .I md-devices-and-options...
1325 .B mdadm \-\-assemble \-\-scan
1329 This usage assembles one or more RAID arrays from pre-existing components.
1330 For each array, mdadm needs to know the md device, the identity of the
1331 array, and a number of component-devices. These can be found in a number of ways.
1333 In the first usage example (without the
1335 the first device given is the md device.
1336 In the second usage example, all devices listed are treated as md
1337 devices and assembly is attempted.
1338 In the third (where no devices are listed) all md devices that are
1339 listed in the configuration file are assembled. If not arrays are
1340 described by the configuration file, then any arrays that
1341 can be found on unused devices will be assembled.
1343 If precisely one device is listed, but
1349 was given and identity information is extracted from the configuration file.
1351 The identity can be given with the
1357 option, will be taken from the md-device record in the config file, or
1358 will be taken from the super block of the first component-device
1359 listed on the command line.
1361 Devices can be given on the
1363 command line or in the config file. Only devices which have an md
1364 superblock which contains the right identity will be considered for
1367 The config file is only used if explicitly named with
1369 or requested with (a possibly implicit)
1374 .B /etc/mdadm/mdadm.conf
1379 is not given, then the config file will only be used to find the
1380 identity of md arrays.
1382 Normally the array will be started after it is assembled. However if
1384 is not given and not all expected drives were listed, then the array
1385 is not started (to guard against usage errors). To insist that the
1386 array be started in this case (as may work for RAID1, 4, 5, 6, or 10),
1395 does not create any entries in
1399 It does record information in
1400 .B /var/run/mdadm/map
1403 to choose the correct name.
1407 detects that udev is not configured, it will create the devices in
1411 In Linux kernels prior to version 2.6.28 there were two distinctly
1412 different types of md devices that could be created: one that could be
1413 partitioned using standard partitioning tools and one that could not.
1414 Since 2.6.28 that distinction is no longer relevant as both type of
1415 devices can be partitioned.
1417 will normally create the type that originally could not be partitioned
1418 as it has a well defined major number (9).
1420 Prior to 2.6.28, it is important that mdadm chooses the correct type
1421 of array device to use. This can be controlled with the
1423 option. In particular, a value of "mdp" or "part" or "p" tells mdadm
1424 to use a partitionable device rather than the default.
1426 In the no-udev case, the value given to
1428 can be suffixed by a number. This tells
1430 to create that number of partition devices rather than the default of 4.
1434 can also be given in the configuration file as a word starting
1436 on the ARRAY line for the relevant array.
1443 and no devices are listed,
1445 will first attempt to assemble all the arrays listed in the config
1448 In no array at listed in the config (other than those marked
1450 it will look through the available devices for possible arrays and
1451 will try to assemble anything that it finds. Arrays which are tagged
1452 as belonging to the given homehost will be assembled and started
1453 normally. Arrays which do not obviously belong to this host are given
1454 names that are expected not to conflict with anything local, and are
1455 started "read-auto" so that nothing is written to any device until the
1456 array is written to. i.e. automatic resync etc is delayed.
1460 finds a consistent set of devices that look like they should comprise
1461 an array, and if the superblock is tagged as belonging to the given
1462 home host, it will automatically choose a device name and try to
1463 assemble the array. If the array uses version-0.90 metadata, then the
1465 number as recorded in the superblock is used to create a name in
1469 If the array uses version-1 metadata, then the
1471 from the superblock is used to similarly create a name in
1473 (the name will have any 'host' prefix stripped first).
1475 This behaviour can be modified by the
1479 configuration file. This line can indicate that specific metadata
1480 type should, or should not, be automatically assembled. If an array
1481 is found which is not listed in
1483 and has a metadata format that is denied by the
1485 line, then it will not be assembled.
1488 line can also request that all arrays identified as being for this
1489 homehost should be assembled regardless of their metadata type.
1492 for further details.
1497 cannot find any array for the given host at all, and if
1498 .B \-\-auto\-update\-homehost
1501 will search again for any array (not just an array created for this
1502 host) and will assemble each assuming
1503 .BR \-\-update=homehost .
1504 This will change the host tag in the superblock so that on the next run,
1505 these arrays will be found without the second pass. The intention of
1506 this feature is to support transitioning a set of md arrays to using
1509 The reason for requiring arrays to be tagged with the homehost for
1510 auto assembly is to guard against problems that can arise when moving
1511 devices from one host to another.
1522 .BI \-\-raid\-devices= Z
1526 This usage is similar to
1528 The difference is that it creates an array without a superblock. With
1529 these arrays there is no difference between initially creating the array and
1530 subsequently assembling the array, except that hopefully there is useful
1531 data there in the second case.
1533 The level may raid0, linear, raid1, raid10, multipath, or faulty, or
1534 one of their synonyms. All devices must be listed and the array will
1535 be started once complete. It will often be appropriate to use
1536 .B \-\-assume\-clean
1537 with levels raid1 or raid10.
1548 .BI \-\-raid\-devices= Z
1552 This usage will initialise a new md array, associate some devices with
1553 it, and activate the array.
1555 The named device will normally not exist when
1556 .I "mdadm \-\-create"
1557 is run, but will be created by
1559 once the array becomes active.
1561 As devices are added, they are checked to see if they contain RAID
1562 superblocks or filesystems. They are also checked to see if the variance in
1563 device size exceeds 1%.
1565 If any discrepancy is found, the array will not automatically be run, though
1568 can override this caution.
1570 To create a "degraded" array in which some devices are missing, simply
1571 give the word "\fBmissing\fP"
1572 in place of a device name. This will cause
1574 to leave the corresponding slot in the array empty.
1575 For a RAID4 or RAID5 array at most one slot can be
1576 "\fBmissing\fP"; for a RAID6 array at most two slots.
1577 For a RAID1 array, only one real device needs to be given. All of the
1581 When creating a RAID5 array,
1583 will automatically create a degraded array with an extra spare drive.
1584 This is because building the spare into a degraded array is in general
1585 faster than resyncing the parity on a non-degraded, but not clean,
1586 array. This feature can be overridden with the
1590 When creating an array with version-1 metadata a name for the array is
1592 If this is not given with the
1596 will choose a name based on the last component of the name of the
1597 device being created. So if
1599 is being created, then the name
1604 is being created, then the name
1608 When creating a partition based array, using
1610 with version-1.x metadata, the partition type should be set to
1612 (non fs-data). This type selection allows for greater precision since
1613 using any other [RAID auto-detect (0xFD) or a GNU/Linux partition (0x83)],
1614 might create problems in the event of array recovery through a live cdrom.
1616 A new array will normally get a randomly assigned 128bit UUID which is
1617 very likely to be unique. If you have a specific need, you can choose
1618 a UUID for the array by giving the
1620 option. Be warned that creating two arrays with the same UUID is a
1621 recipe for disaster. Also, using
1623 when creating a v0.90 array will silently override any
1628 .\"option is given, it is not necessary to list any component-devices in this command.
1629 .\"They can be added later, before a
1633 .\"is given, the apparent size of the smallest drive given is used.
1635 When creating an array within a
1638 can be given either the list of devices to use, or simply the name of
1639 the container. The former case gives control over which devices in
1640 the container will be used for the array. The latter case allows
1642 to automatically choose which devices to use based on how much spare
1645 The General Management options that are valid with
1650 insist on running the array even if some devices look like they might
1655 start the array readonly \(em not supported yet.
1662 .I options... devices...
1665 This usage will allow individual devices in an array to be failed,
1666 removed or added. It is possible to perform multiple operations with
1667 on command. For example:
1669 .B " mdadm /dev/md0 \-f /dev/hda1 \-r /dev/hda1 \-a /dev/hda1"
1675 and will then remove it from the array and finally add it back
1676 in as a spare. However only one md array can be affected by a single
1679 When a device is added to an active array, mdadm checks to see if it
1680 has metadata on it which suggests that it was recently a member of the
1681 array. If it does, it tried to "re-add" the device. If there have
1682 been no changes since the device was removed, or if the array has a
1683 write-intent bitmap which has recorded whatever changes there were,
1684 then the device will immediately become a full member of the array and
1685 those differences recorded in the bitmap will be resolved.
1695 MISC mode includes a number of distinct operations that
1696 operate on distinct devices. The operations are:
1699 The device is examined to see if it is
1700 (1) an active md array, or
1701 (2) a component of an md array.
1702 The information discovered is reported.
1706 The device should be an active md device.
1708 will display a detailed description of the array.
1712 will cause the output to be less detailed and the format to be
1713 suitable for inclusion in
1714 .BR /etc/mdadm.conf .
1717 will normally be 0 unless
1719 failed to get useful information about the device(s); however, if the
1721 option is given, then the exit status will be:
1725 The array is functioning normally.
1728 The array has at least one failed device.
1731 The array has multiple failed devices such that it is unusable.
1734 There was an error while trying to get information about the device.
1738 .B \-\-detail\-platform
1739 Print detail of the platform's RAID capabilities (firmware / hardware
1740 topology). If the metadata is specified with
1744 then the return status will be:
1748 metadata successfully enumerated its platform components on this system
1751 metadata is platform independent
1754 metadata failed to find its platform components on this system
1759 The device should be a component of an md array.
1761 will read the md superblock of the device and display the contents.
1766 is given, then multiple devices that are components of the one array
1767 are grouped together and reported in a single entry suitable
1769 .BR /etc/mdadm.conf .
1773 without listing any devices will cause all devices listed in the
1774 config file to be examined.
1778 The devices should be active md arrays which will be deactivated, as
1779 long as they are not currently in use.
1783 This will fully activate a partially assembled md array.
1787 This will mark an active array as read-only, providing that it is
1788 not currently being used.
1794 array back to being read/write.
1798 For all operations except
1801 will cause the operation to be applied to all arrays listed in
1806 causes all devices listed in the config file to be examined.
1809 .BR \-b ", " \-\-brief
1810 Be less verbose. This is used with
1818 gives an intermediate level of verbosity.
1824 .B mdadm \-\-monitor
1825 .I options... devices...
1830 to periodically poll a number of md arrays and to report on any events
1833 will never exit once it decides that there are arrays to be checked,
1834 so it should normally be run in the background.
1836 As well as reporting events,
1838 may move a spare drive from one array to another if they are in the
1841 and if the destination array has a failed drive but no spares.
1843 If any devices are listed on the command line,
1845 will only monitor those devices. Otherwise all arrays listed in the
1846 configuration file will be monitored. Further, if
1848 is given, then any other md devices that appear in
1850 will also be monitored.
1852 The result of monitoring the arrays is the generation of events.
1853 These events are passed to a separate program (if specified) and may
1854 be mailed to a given E-mail address.
1856 When passing events to a program, the program is run once for each event,
1857 and is given 2 or 3 command-line arguments: the first is the
1858 name of the event (see below), the second is the name of the
1859 md device which is affected, and the third is the name of a related
1860 device if relevant (such as a component device that has failed).
1864 is given, then a program or an E-mail address must be specified on the
1865 command line or in the config file. If neither are available, then
1867 will not monitor anything.
1871 will continue monitoring as long as something was found to monitor. If
1872 no program or email is given, then each event is reported to
1875 The different events are:
1879 .B DeviceDisappeared
1880 An md array which previously was configured appears to no longer be
1881 configured. (syslog priority: Critical)
1885 was told to monitor an array which is RAID0 or Linear, then it will
1887 .B DeviceDisappeared
1888 with the extra information
1890 This is because RAID0 and Linear do not support the device-failed,
1891 hot-spare and resync operations which are monitored.
1895 An md array started reconstruction. (syslog priority: Warning)
1901 is a two-digit number (ie. 05, 48). This indicates that rebuild
1902 has passed that many percent of the total. The events are generated
1903 with fixed increment since 0. Increment size may be specified with
1904 a commandline option (default is 20). (syslog priority: Warning)
1908 An md array that was rebuilding, isn't any more, either because it
1909 finished normally or was aborted. (syslog priority: Warning)
1913 An active component device of an array has been marked as
1914 faulty. (syslog priority: Critical)
1918 A spare component device which was being rebuilt to replace a faulty
1919 device has failed. (syslog priority: Critical)
1923 A spare component device which was being rebuilt to replace a faulty
1924 device has been successfully rebuilt and has been made active.
1925 (syslog priority: Info)
1929 A new md array has been detected in the
1931 file. (syslog priority: Info)
1935 A newly noticed array appears to be degraded. This message is not
1938 notices a drive failure which causes degradation, but only when
1940 notices that an array is degraded when it first sees the array.
1941 (syslog priority: Critical)
1945 A spare drive has been moved from one array in a
1947 to another to allow a failed drive to be replaced.
1948 (syslog priority: Info)
1954 has been told, via the config file, that an array should have a certain
1955 number of spare devices, and
1957 detects that it has fewer than this number when it first sees the
1958 array, it will report a
1961 (syslog priority: Warning)
1965 An array was found at startup, and the
1968 (syslog priority: Info)
1978 cause Email to be sent. All events cause the program to be run.
1979 The program is run with two or three arguments: the event
1980 name, the array device and possibly a second device.
1982 Each event has an associated array device (e.g.
1984 and possibly a second device. For
1989 the second device is the relevant component device.
1992 the second device is the array that the spare was moved from.
1996 to move spares from one array to another, the different arrays need to
1997 be labeled with the same
1999 in the configuration file. The
2001 name can be any string; it is only necessary that different spare
2002 groups use different names.
2006 detects that an array in a spare group has fewer active
2007 devices than necessary for the complete array, and has no spare
2008 devices, it will look for another array in the same spare group that
2009 has a full complement of working drive and a spare. It will then
2010 attempt to remove the spare from the second drive and add it to the
2012 If the removal succeeds but the adding fails, then it is added back to
2016 The GROW mode is used for changing the size or shape of an active
2018 For this to work, the kernel must support the necessary change.
2019 Various types of growth are being added during 2.6 development,
2020 including restructuring a RAID5 array to have more active devices.
2022 Currently the only support available is to
2024 change the "size" attribute
2025 for RAID1, RAID5 and RAID6.
2027 increase or decrease the "raid\-devices" attribute of RAID1, RAID5,
2030 change the chunk-size and layout of RAID5 and RAID6.
2032 convert between RAID1 and RAID5, and between RAID5 and RAID6.
2034 add a write-intent bitmap to any array which supports these bitmaps, or
2035 remove a write-intent bitmap from such an array.
2038 GROW mode is not currently supported for
2040 or arrays inside containers.
2043 Normally when an array is built the "size" it taken from the smallest
2044 of the drives. If all the small drives in an arrays are, one at a
2045 time, removed and replaced with larger drives, then you could have an
2046 array of large drives with only a small amount used. In this
2047 situation, changing the "size" with "GROW" mode will allow the extra
2048 space to start being used. If the size is increased in this way, a
2049 "resync" process will start to make sure the new parts of the array
2052 Note that when an array changes size, any filesystem that may be
2053 stored in the array will not automatically grow to use the space. The
2054 filesystem will need to be explicitly told to use the extra space.
2056 Also the size of an array cannot be changed while it has an active
2057 bitmap. If an array has a bitmap, it must be removed before the size
2058 can be changed. Once the change it complete a new bitmap can be created.
2060 .SS RAID\-DEVICES CHANGES
2062 A RAID1 array can work with any number of devices from 1 upwards
2063 (though 1 is not very useful). There may be times which you want to
2064 increase or decrease the number of active devices. Note that this is
2065 different to hot-add or hot-remove which changes the number of
2068 When reducing the number of devices in a RAID1 array, the slots which
2069 are to be removed from the array must already be vacant. That is, the
2070 devices which were in those slots must be failed and removed.
2072 When the number of devices is increased, any hot spares that are
2073 present will be activated immediately.
2075 Changing the number of active devices in a RAID5 or RAID6 is much more
2076 effort. Every block in the array will need to be read and written
2077 back to a new location. From 2.6.17, the Linux Kernel is able to
2078 increase the number of devices in a RAID5 safely, including restarting
2079 an interrupted "reshape". From 2.6.31, the Linux Kernel is able to
2080 increase or decrease the number of devices in a RAID5 or RAID6.
2082 When decreasing the number of devices, the size of the array will also
2083 decrease. If there was data in the array, it could get destroyed and
2084 this is not reversible. To help prevent accidents,
2086 requires that the size of the array be decreased first with
2087 .BR "mdadm --grow --array-size" .
2088 This is a reversible change which simply makes the end of the array
2089 inaccessible. The integrity of any data can then be checked before
2090 the non-reversible reduction in the number of devices is request.
2092 When relocating the first few stripes on a RAID5, it is not possible
2093 to keep the data on disk completely consistent and crash-proof. To
2094 provide the required safety, mdadm disables writes to the array while
2095 this "critical section" is reshaped, and takes a backup of the data
2096 that is in that section. This backup is normally stored in any spare
2097 devices that the array has, however it can also be stored in a
2098 separate file specified with the
2100 option. If this option is used, and the system does crash during the
2101 critical period, the same file must be passed to
2103 to restore the backup and reassemble the array.
2107 Changing the RAID level of any array happens instantaneously. However
2108 in the RAID to RAID6 case this requires a non-standard layout of the
2109 RAID6 data, and in the RAID6 to RAID5 case that non-standard layout is
2110 required before the change can be accomplish. So while the level
2111 change is instant, the accompanying layout change can take quite a
2114 .SS CHUNK-SIZE AND LAYOUT CHANGES
2116 Changing the chunk-size of layout without also changing the number of
2117 devices as the same time will involve re-writing all blocks in-place.
2118 To ensure against data loss in the case of a crash, a
2120 must be provided for these changes. Small sections of the array will
2121 be copied to the backup file while they are being rearranged.
2123 If the reshape is interrupted for any reason, this backup file must be
2125 .B "mdadm --assemble"
2126 so the array can be reassembled. Consequently the file cannot be
2127 stored on the device being reshaped.
2132 A write-intent bitmap can be added to, or removed from, an active
2133 array. Either internal bitmaps, or bitmaps stored in a separate file,
2134 can be added. Note that if you add a bitmap stored in a file which is
2135 in a filesystem that is on the RAID array being affected, the system
2136 will deadlock. The bitmap must be on a separate filesystem.
2138 .SH INCREMENTAL MODE
2142 .B mdadm \-\-incremental
2148 .B mdadm \-\-incremental \-\-rebuild
2151 .B mdadm \-\-incremental \-\-run \-\-scan
2154 This mode is designed to be used in conjunction with a device
2155 discovery system. As devices are found in a system, they can be
2157 .B "mdadm \-\-incremental"
2158 to be conditionally added to an appropriate array.
2160 If the device passed is a
2162 device created by a previous call to
2164 then rather than trying to add that device to an array, all the arrays
2165 described by the metadata of the container will be started.
2168 performs a number of tests to determine if the device is part of an
2169 array, and which array it should be part of. If an appropriate array
2170 is found, or can be created,
2172 adds the device to the array and conditionally starts the array.
2176 will only add devices to an array which were previously working
2177 (active or spare) parts of that array. It does not currently support
2178 automatic inclusion of a new drive as a spare in some array.
2182 makes are as follow:
2184 Is the device permitted by
2186 That is, is it listed in a
2188 line in that file. If
2190 is absent then the default it to allow any device. Similar if
2192 contains the special word
2194 then any device is allowed. Otherwise the device name given to
2196 must match one of the names or patterns in a
2201 Does the device have a valid md superblock. If a specific metadata
2202 version is request with
2206 then only that style of metadata is accepted, otherwise
2208 finds any known version of metadata. If no
2210 metadata is found, the device is rejected.
2214 Does the metadata match an expected array?
2215 The metadata can match in two ways. Either there is an array listed
2218 which identifies the array (either by UUID, by name, by device list,
2219 or by minor-number), or the array was created with a
2225 or on the command line.
2228 is not able to positively identify the array as belonging to the
2229 current host, the device will be rejected.
2233 keeps a list of arrays that it has partially assembled in
2234 .B /var/run/mdadm/map
2236 .B /var/run/mdadm.map
2237 if the directory doesn't exist. Or maybe even
2238 .BR /dev/.mdadm.map ).
2239 If no array exists which matches
2240 the metadata on the new device,
2242 must choose a device name and unit number. It does this based on any
2245 or any name information stored in the metadata. If this name
2246 suggests a unit number, that number will be used, otherwise a free
2247 unit number will be chosen. Normally
2249 will prefer to create a partitionable array, however if the
2253 suggests that a non-partitionable array is preferred, that will be
2256 If the array is not found in the config file and its metadata does not
2257 identify it as belonging to the "homehost", then
2259 will choose a name for the array which is certain not to conflict with
2260 any array which does belong to this host. It does this be adding an
2261 underscore and a small number to the name preferred by the metadata.
2263 Once an appropriate array is found or created and the device is added,
2265 must decide if the array is ready to be started. It will
2266 normally compare the number of available (non-spare) devices to the
2267 number of devices that the metadata suggests need to be active. If
2268 there are at least that many, the array will be started. This means
2269 that if any devices are missing the array will not be restarted.
2275 in which case the array will be run as soon as there are enough
2276 devices present for the data to be accessible. For a RAID1, that
2277 means one device will start the array. For a clean RAID5, the array
2278 will be started as soon as all but one drive is present.
2280 Note that neither of these approaches is really ideal. If it can
2281 be known that all device discovery has completed, then
2285 can be run which will try to start all arrays that are being
2286 incrementally assembled. They are started in "read-auto" mode in
2287 which they are read-only until the first write request. This means
2288 that no metadata updates are made and no attempt at resync or recovery
2289 happens. Further devices that are found before the first write can
2290 still be added safely.
2293 This section describes environment variables that affect how mdadm
2298 Setting this value to 1 will prevent mdadm from automatically launching
2299 mdmon. This variable is intended primarily for debugging mdadm/mdmon.
2305 does not create any device nodes in /dev, but leaves that task to
2309 appears not to be configured, or if this environment variable is set
2312 will create and devices that are needed.
2316 .B " mdadm \-\-query /dev/name-of-device"
2318 This will find out if a given device is a RAID array, or is part of
2319 one, and will provide brief information about the device.
2321 .B " mdadm \-\-assemble \-\-scan"
2323 This will assemble and start all arrays listed in the standard config
2324 file. This command will typically go in a system startup file.
2326 .B " mdadm \-\-stop \-\-scan"
2328 This will shut down all arrays that can be shut down (i.e. are not
2329 currently in use). This will typically go in a system shutdown script.
2331 .B " mdadm \-\-follow \-\-scan \-\-delay=120"
2333 If (and only if) there is an Email address or program given in the
2334 standard config file, then
2335 monitor the status of all arrays listed in that file by
2336 polling them ever 2 minutes.
2338 .B " mdadm \-\-create /dev/md0 \-\-level=1 \-\-raid\-devices=2 /dev/hd[ac]1"
2340 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
2343 .B " echo 'DEVICE /dev/hd*[0\-9] /dev/sd*[0\-9]' > mdadm.conf"
2345 .B " mdadm \-\-detail \-\-scan >> mdadm.conf"
2347 This will create a prototype config file that describes currently
2348 active arrays that are known to be made from partitions of IDE or SCSI drives.
2349 This file should be reviewed before being used as it may
2350 contain unwanted detail.
2352 .B " echo 'DEVICE /dev/hd[a\-z] /dev/sd*[a\-z]' > mdadm.conf"
2354 .B " mdadm \-\-examine \-\-scan \-\-config=mdadm.conf >> mdadm.conf"
2356 This will find arrays which could be assembled from existing IDE and
2357 SCSI whole drives (not partitions), and store the information in the
2358 format of a config file.
2359 This file is very likely to contain unwanted detail, particularly
2362 entries. It should be reviewed and edited before being used as an
2365 .B " mdadm \-\-examine \-\-brief \-\-scan \-\-config=partitions"
2367 .B " mdadm \-Ebsc partitions"
2369 Create a list of devices by reading
2370 .BR /proc/partitions ,
2371 scan these for RAID superblocks, and printout a brief listing of all
2374 .B " mdadm \-Ac partitions \-m 0 /dev/md0"
2376 Scan all partitions and devices listed in
2377 .BR /proc/partitions
2380 out of all such devices with a RAID superblock with a minor number of 0.
2382 .B " mdadm \-\-monitor \-\-scan \-\-daemonise > /var/run/mdadm"
2384 If config file contains a mail address or alert program, run mdadm in
2385 the background in monitor mode monitoring all md devices. Also write
2386 pid of mdadm daemon to
2387 .BR /var/run/mdadm .
2389 .B " mdadm \-Iq /dev/somedevice"
2391 Try to incorporate newly discovered device into some array as
2394 .B " mdadm \-\-incremental \-\-rebuild \-\-run \-\-scan"
2396 Rebuild the array map from any current arrays, and then start any that
2399 .B " mdadm /dev/md4 --fail detached --remove detached"
2401 Any devices which are components of /dev/md4 will be marked as faulty
2402 and then remove from the array.
2404 .B " mdadm --grow /dev/md4 --level=6 --backup-file=/root/backup-md4
2408 which is currently a RAID5 array will be converted to RAID6. There
2409 should normally already be a spare drive attached to the array as a
2410 RAID6 needs one more drive than a matching RAID5.
2412 .B " mdadm --create /dev/md/ddf --metadata=ddf --raid-disks 6 /dev/sd[a-f]"
2414 Create a DDF array over 6 devices.
2416 .B " mdadm --create /dev/md/home -n3 -l5 -z 30000000 /dev/md/ddf"
2418 Create a RAID5 array over any 3 devices in the given DDF set. Use
2419 only 30 gigabytes of each device.
2421 .B " mdadm -A /dev/md/ddf1 /dev/sd[a-f]"
2423 Assemble a pre-exist ddf array.
2425 .B " mdadm -I /dev/md/ddf1"
2427 Assemble all arrays contained in the ddf array, assigning names as
2430 .B " mdadm \-\-create \-\-help"
2432 Provide help about the Create mode.
2434 .B " mdadm \-\-config \-\-help"
2436 Provide help about the format of the config file.
2438 .B " mdadm \-\-help"
2440 Provide general help.
2450 lists all active md devices with information about them.
2452 uses this to find arrays when
2454 is given in Misc mode, and to monitor array reconstruction
2459 The config file lists which devices may be scanned to see if
2460 they contain MD super block, and gives identifying information
2461 (e.g. UUID) about known MD arrays. See
2465 .SS /var/run/mdadm/map
2468 mode is used, this file gets a list of arrays currently being created.
2471 does not exist as a directory, then
2472 .B /var/run/mdadm.map
2475 is not available (as may be the case during early boot),
2477 is used on the basis that
2479 is usually available very early in boot.
2484 understand two sorts of names for array devices.
2486 The first is the so-called 'standard' format name, which matches the
2487 names used by the kernel and which appear in
2490 The second sort can be freely chosen, but must reside in
2492 When giving a device name to
2494 to create or assemble an array, either full path name such as
2498 can be given, or just the suffix of the second sort of name, such as
2504 chooses device names during auto-assembly or incremental assembly, it
2505 will sometimes add a small sequence number to the end of the name to
2506 avoid conflicted between multiple arrays that have the same name. If
2508 can reasonably determine that the array really is meant for this host,
2509 either by a hostname in the metadata, or by the presence of the array
2510 in /etc/mdadm.conf, then it will leave off the suffix if possible.
2511 Also if the homehost is specified as
2514 will only use a suffix if a different array of the same name already
2515 exists or is listed in the config file.
2517 The standard names for non-partitioned arrays (the only sort of md
2518 array available in 2.4 and earlier) are of the form
2522 where NN is a number.
2523 The standard names for partitionable arrays (as available from 2.6
2524 onwards) are of the form
2528 Partition numbers should be indicated by added "pMM" to these, thus "/dev/md/d1p2".
2530 From kernel version, 2.6.28 the "non-partitioned array" can actually
2531 be partitioned. So the "md_dNN" names are no longer needed, and
2532 partitions such as "/dev/mdNNpXX" are possible.
2536 was previously known as
2540 is completely separate from the
2542 package, and does not use the
2544 configuration file at all.
2547 For further information on mdadm usage, MD and the various levels of
2550 .B http://linux\-raid.osdl.org/
2552 (based upon Jakob \(/Ostergaard's Software\-RAID.HOWTO)
2554 .\"for new releases of the RAID driver check out:
2557 .\".UR ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2558 .\"ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2563 .\".UR http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2564 .\"http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2567 The latest version of
2569 should always be available from
2571 .B http://www.kernel.org/pub/linux/utils/raid/mdadm/
projects / thirdparty / mdadm.git / blob