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, as well as adding or removing a write-intent bitmap.
129
.B "Incremental Assembly"
130
Add a single device to an appropriate array. If the addition of the
131
device makes the array runnable, the array will be started.
132
This provides a convenient interface to a
134
system. As each device is detected,
136
has a chance to include it in some array as appropriate.
142
in this mode, then any arrays within that container will be assembled
147
This is for doing things to specific components of an array such as
148
adding new spares and removing faulty devices.
152
This is an 'everything else' mode that supports operations on active
153
arrays, operations on component devices such as erasing old superblocks, and
154
information gathering operations.
155
.\"This mode allows operations on independent devices such as examine MD
156
.\"superblocks, erasing old superblocks and stopping active arrays.
160
This mode does not act on a specific device or array, but rather it
161
requests the Linux Kernel to activate any auto-detected arrays.
164
.SH Options for selecting a mode are:
167
.BR \-A ", " \-\-assemble
168
Assemble a pre-existing array.
171
.BR \-B ", " \-\-build
172
Build a legacy array without superblocks.
175
.BR \-C ", " \-\-create
179
.BR \-F ", " \-\-follow ", " \-\-monitor
185
.BR \-G ", " \-\-grow
186
Change the size or shape of an active array.
189
.BR \-I ", " \-\-incremental
190
Add a single device into an appropriate array, and possibly start the array.
194
Request that the kernel starts any auto-detected arrays. This can only
197
is compiled into the kernel \(em not if it is a module.
198
Arrays can be auto-detected by the kernel if all the components are in
199
primary MS-DOS partitions with partition type
201
and all use v0.90 metadata.
202
In-kernel autodetect is not recommended for new installations. Using
204
to detect and assemble arrays \(em possibly in an
206
\(em is substantially more flexible and should be preferred.
209
If a device is given before any options, or if the first option is
214
then the MANAGE mode is assumed.
215
Anything other than these will cause the
219
.SH Options that are not mode-specific are:
222
.BR \-h ", " \-\-help
223
Display general help message or, after one of the above options, a
224
mode-specific help message.
228
Display more detailed help about command line parsing and some commonly
232
.BR \-V ", " \-\-version
233
Print version information for mdadm.
236
.BR \-v ", " \-\-verbose
237
Be more verbose about what is happening. This can be used twice to be
239
The extra verbosity currently only affects
240
.B \-\-detail \-\-scan
242
.BR "\-\-examine \-\-scan" .
245
.BR \-q ", " \-\-quiet
246
Avoid printing purely informative messages. With this,
248
will be silent unless there is something really important to report.
251
.BR \-b ", " \-\-brief
252
Be less verbose. This is used with
260
gives an intermediate level of verbosity.
263
.BR \-f ", " \-\-force
264
Be more forceful about certain operations. See the various modes for
265
the exact meaning of this option in different contexts.
268
.BR \-c ", " \-\-config=
269
Specify the config file. Default is to use
270
.BR /etc/mdadm.conf ,
271
or if that is missing then
272
.BR /etc/mdadm/mdadm.conf .
273
If the config file given is
275
then nothing will be read, but
277
will act as though the config file contained exactly
278
.B "DEVICE partitions containers"
281
to find a list of devices to scan, and
283
to find a list of containers to examine.
286
is given for the config file, then
288
will act as though the config file were empty.
291
.BR \-s ", " \-\-scan
294
for missing information.
295
In general, this option gives
297
permission to get any missing information (like component devices,
298
array devices, array identities, and alert destination) from the
299
configuration file (see previous option);
300
one exception is MISC mode when using
306
says to get a list of array devices from
310
.BR \-e ", " \-\-metadata=
311
Declare the style of RAID metadata (superblock) to be used. The
314
and to guess for other operations.
315
The default can be overridden by setting the
324
.IP "0, 0.90, default"
325
Use the original 0.90 format superblock. This format limits arrays to
326
28 component devices and limits component devices of levels 1 and
327
greater to 2 terabytes.
328
.IP "1, 1.0, 1.1, 1.2"
329
Use the new version-1 format superblock. This has few restrictions.
330
The different sub-versions store the superblock at different locations
331
on the device, either at the end (for 1.0), at the start (for 1.1) or
332
4K from the start (for 1.2).
334
Use the "Industry Standard" DDF (Disk Data Format) format defined by
336
When creating a DDF array a
338
will be created, and normal arrays can be created in that container.
340
Use the Intel(R) Matrix Storage Manager metadata format. This creates a
342
which is managed in a similar manner to DDF, and is supported by an
343
option-rom on some platforms:
345
.B http://www.intel.com/design/chipsets/matrixstorage_sb.htm
351
This will override any
353
setting in the config file and provides the identity of the host which
354
should be considered the home for any arrays.
356
When creating an array, the
358
will be recorded in the metadata. For version-1 superblocks, it will
359
be prefixed to the array name. For version-0.90 superblocks, part of
360
the SHA1 hash of the hostname will be stored in the later half of the
363
When reporting information about an array, any array which is tagged
364
for the given homehost will be reported as such.
366
When using Auto-Assemble, only arrays tagged for the given homehost
367
will be allowed to use 'local' names (i.e. not ending in '_' followed
368
by a digit string). See below under
369
.BR "Auto Assembly" .
371
.SH For create, build, or grow:
374
.BR \-n ", " \-\-raid\-devices=
375
Specify the number of active devices in the array. This, plus the
376
number of spare devices (see below) must equal the number of
378
(including "\fBmissing\fP" devices)
379
that are listed on the command line for
381
Setting a value of 1 is probably
382
a mistake and so requires that
384
be specified first. A value of 1 will then be allowed for linear,
385
multipath, RAID0 and RAID1. It is never allowed for RAID4, RAID5 or RAID6.
387
This number can only be changed using
389
for RAID1, RAID4, RAID5 and RAID6 arrays, and only on kernels which provide
390
the necessary support.
393
.BR \-x ", " \-\-spare\-devices=
394
Specify the number of spare (eXtra) devices in the initial array.
395
Spares can also be added
396
and removed later. The number of component devices listed
397
on the command line must equal the number of RAID devices plus the
398
number of spare devices.
401
.BR \-z ", " \-\-size=
402
Amount (in Kibibytes) of space to use from each drive in RAID levels 1/4/5/6.
403
This must be a multiple of the chunk size, and must leave about 128Kb
404
of space at the end of the drive for the RAID superblock.
405
If this is not specified
406
(as it normally is not) the smallest drive (or partition) sets the
407
size, though if there is a variance among the drives of greater than 1%, a warning is
410
This value can be set with
412
for RAID level 1/4/5/6. If the array was created with a size smaller
413
than the currently active drives, the extra space can be accessed
416
The size can be given as
418
which means to choose the largest size that fits on all current drives.
420
This value can not be used with
422
metadata such as DDF and IMSM.
425
.BR \-c ", " \-\-chunk=
426
Specify chunk size of kibibytes. The default is 64.
427
This is only meaningful for RAID0, RAID4, RAID5, RAID6, and RAID10.
431
Specify rounding factor for a Linear array. The size of each
432
component will be rounded down to a multiple of this size.
433
This is a synonym for
435
but highlights the different meaning for Linear as compared to other
439
.BR \-l ", " \-\-level=
440
Set RAID level. When used with
442
options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
443
raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty, container.
444
Obviously some of these are synonymous.
448
metadata type is requested, only the
450
level is permitted, and it does not need to be explicitly given.
454
only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid.
456
Not yet supported with
460
.BR \-p ", " \-\-layout=
461
This option configures the fine details of data layout for RAID5, RAID6,
462
and RAID10 arrays, and controls the failure modes for
465
The layout of the RAID5 parity block can be one of
466
.BR left\-asymmetric ,
467
.BR left\-symmetric ,
468
.BR right\-asymmetric ,
469
.BR right\-symmetric ,
470
.BR la ", " ra ", " ls ", " rs .
472
.BR left\-symmetric .
474
It is also possibly to cause RAID5 to use a RAID4-like layout by
480
Finally for RAID5 there are DDF\-compatible layouts,
481
.BR ddf\-zero\-restart ,
482
.BR ddf\-N\-restart ,
484
.BR ddf\-N\-continue .
486
These same layouts are available for RAID6. There are also 4 layouts
487
that will provide an intermediate stage for converting between RAID5
488
and RAID6. These provide a layout which is identical to the
489
corresponding RAID5 layout on the first N\-1 devices, and has the 'Q'
490
syndrome (the second 'parity' block used by RAID6) on the last device.
492
.BR left\-symmetric\-6 ,
493
.BR right\-symmetric\-6 ,
494
.BR left\-asymmetric\-6 ,
495
.BR right\-asymmetric\-6 ,
497
.BR pairty\-first\-6 .
499
When setting the failure mode for level
502
.BR write\-transient ", " wt ,
503
.BR read\-transient ", " rt ,
504
.BR write\-persistent ", " wp ,
505
.BR read\-persistent ", " rp ,
507
.BR read\-fixable ", " rf ,
508
.BR clear ", " flush ", " none .
510
Each failure mode can be followed by a number, which is used as a period
511
between fault generation. Without a number, the fault is generated
512
once on the first relevant request. With a number, the fault will be
513
generated after that many requests, and will continue to be generated
514
every time the period elapses.
516
Multiple failure modes can be current simultaneously by using the
518
option to set subsequent failure modes.
520
"clear" or "none" will remove any pending or periodic failure modes,
521
and "flush" will clear any persistent faults.
523
To set the parity with
525
the level of the array ("faulty")
526
must be specified before the fault mode is specified.
528
Finally, the layout options for RAID10 are one of 'n', 'o' or 'f' followed
529
by a small number. The default is 'n2'. The supported options are:
532
signals 'near' copies. Multiple copies of one data block are at
533
similar offsets in different devices.
536
signals 'offset' copies. Rather than the chunks being duplicated
537
within a stripe, whole stripes are duplicated but are rotated by one
538
device so duplicate blocks are on different devices. Thus subsequent
539
copies of a block are in the next drive, and are one chunk further
544
(multiple copies have very different offsets).
545
See md(4) for more detail about 'near', 'offset', and 'far'.
547
The number is the number of copies of each datablock. 2 is normal, 3
548
can be useful. This number can be at most equal to the number of
549
devices in the array. It does not need to divide evenly into that
550
number (e.g. it is perfectly legal to have an 'n2' layout for an array
551
with an odd number of devices).
557
(thus explaining the p of
561
.BR \-b ", " \-\-bitmap=
562
Specify a file to store a write-intent bitmap in. The file should not
565
is also given. The same file should be provided
566
when assembling the array. If the word
568
is given, then the bitmap is stored with the metadata on the array,
569
and so is replicated on all devices. If the word
573
mode, then any bitmap that is present is removed.
575
To help catch typing errors, the filename must contain at least one
576
slash ('/') if it is a real file (not 'internal' or 'none').
578
Note: external bitmaps are only known to work on ext2 and ext3.
579
Storing bitmap files on other filesystems may result in serious problems.
582
.BR \-\-bitmap\-chunk=
583
Set the chunksize of the bitmap. Each bit corresponds to that many
584
Kilobytes of storage.
585
When using a file based bitmap, the default is to use the smallest
586
size that is at-least 4 and requires no more than 2^21 chunks.
589
bitmap, the chunksize is automatically determined to make best use of
593
.BR \-W ", " \-\-write\-mostly
594
subsequent devices listed in a
599
command will be flagged as 'write-mostly'. This is valid for RAID1
600
only and means that the 'md' driver will avoid reading from these
601
devices if at all possible. This can be useful if mirroring over a
605
.BR \-\-write\-behind=
606
Specify that write-behind mode should be enabled (valid for RAID1
607
only). If an argument is specified, it will set the maximum number
608
of outstanding writes allowed. The default value is 256.
609
A write-intent bitmap is required in order to use write-behind
610
mode, and write-behind is only attempted on drives marked as
614
.BR \-\-assume\-clean
617
that the array pre-existed and is known to be clean. It can be useful
618
when trying to recover from a major failure as you can be sure that no
619
data will be affected unless you actually write to the array. It can
620
also be used when creating a RAID1 or RAID10 if you want to avoid the
621
initial resync, however this practice \(em while normally safe \(em is not
622
recommended. Use this only if you really know what you are doing.
625
.BR \-\-backup\-file=
628
is used to increase the number of
629
raid-devices in a RAID5 if there are no spare devices available.
630
See the GROW MODE section below on RAID\-DEVICES CHANGES. The file
631
should be stored on a separate device, not on the RAID array being
635
.BR \-N ", " \-\-name=
638
for the array. This is currently only effective when creating an
639
array with a version-1 superblock, or an array in a DDF container.
640
The name is a simple textual string that can be used to identify array
641
components when assembling. If name is needed but not specified, it
642
is taken from the basename of the device that is being created.
654
run the array, even if some of the components
655
appear to be active in another array or filesystem. Normally
657
will ask for confirmation before including such components in an
658
array. This option causes that question to be suppressed.
661
.BR \-f ", " \-\-force
664
accept the geometry and layout specified without question. Normally
666
will not allow creation of an array with only one device, and will try
667
to create a RAID5 array with one missing drive (as this makes the
668
initial resync work faster). With
671
will not try to be so clever.
674
.BR \-a ", " "\-\-auto{=yes,md,mdp,part,p}{NN}"
675
Instruct mdadm how to create the device file if needed, possibly allocating
676
an unused minor number. "md" causes a non-partitionable array
677
to be used (though since Linux 2.6.28, these array devices are in fact
678
partitionable). "mdp", "part" or "p" causes a partitionable array (2.6 and
679
later) to be used. "yes" requires the named md device to have
680
a 'standard' format, and the type and minor number will be determined
681
from this. With mdadm 3.0, device creation is normally left up to
683
so this option is unlikely to be needed.
684
See DEVICE NAMES below.
686
The argument can also come immediately after
691
is not given on the command line or in the config file, then
697
is also given, then any
699
entries in the config file will override the
701
instruction given on the command line.
703
For partitionable arrays,
705
will create the device file for the whole array and for the first 4
706
partitions. A different number of partitions can be specified at the
707
end of this option (e.g.
709
If the device name ends with a digit, the partition names add a 'p',
711
.IR /dev/md/home1p3 .
712
If there is no trailing digit, then the partition names just have a
714
.IR /dev/md/scratch3 .
716
If the md device name is in a 'standard' format as described in DEVICE
717
NAMES, then it will be created, if necessary, with the appropriate
718
device number based on that name. If the device name is not in one of these
719
formats, then a unused device number will be allocated. The device
720
number will be considered unused if there is no active array for that
721
number, and there is no entry in /dev for that number and with a
722
non-standard name. Names that are not in 'standard' format are only
723
allowed in "/dev/md/".
726
.\".BR \-\-symlink = no
731
.\"to create devices in
733
.\"it will also create symlinks from
735
.\"with names starting with
741
.\"to suppress this, or
742
.\".B \-\-symlink=yes
743
.\"to enforce this even if it is suppressing
750
.BR \-u ", " \-\-uuid=
751
uuid of array to assemble. Devices which don't have this uuid are
755
.BR \-m ", " \-\-super\-minor=
756
Minor number of device that array was created for. Devices which
757
don't have this minor number are excluded. If you create an array as
758
/dev/md1, then all superblocks will contain the minor number 1, even if
759
the array is later assembled as /dev/md2.
761
Giving the literal word "dev" for
765
to use the minor number of the md device that is being assembled.
768
.B \-\-super\-minor=dev
769
will look for super blocks with a minor number of 0.
772
is only relevant for v0.90 metadata, and should not normally be used.
778
.BR \-N ", " \-\-name=
779
Specify the name of the array to assemble. This must be the name
780
that was specified when creating the array. It must either match
781
the name stored in the superblock exactly, or it must match
784
prefixed to the start of the given name.
787
.BR \-f ", " \-\-force
788
Assemble the array even if the metadata on some devices appears to be
791
cannot find enough working devices to start the array, but can find
792
some devices that are recorded as having failed, then it will mark
793
those devices as working so that the array can be started.
794
An array which requires
796
to be started may contain data corruption. Use it carefully.
800
Attempt to start the array even if fewer drives were given than were
801
present last time the array was active. Normally if not all the
802
expected drives are found and
804
is not used, then the array will be assembled but not started.
807
an attempt will be made to start it anyway.
811
This is the reverse of
813
in that it inhibits the startup of array unless all expected drives
814
are present. This is only needed with
816
and can be used if the physical connections to devices are
817
not as reliable as you would like.
820
.BR \-a ", " "\-\-auto{=no,yes,md,mdp,part}"
821
See this option under Create and Build options.
824
.BR \-b ", " \-\-bitmap=
825
Specify the bitmap file that was given when the array was created. If
828
bitmap, there is no need to specify this when assembling the array.
831
.BR \-\-backup\-file=
834
was used to grow the number of raid-devices in a RAID5, and the system
835
crashed during the critical section, then the same
839
to allow possibly corrupted data to be restored.
842
.BR \-U ", " \-\-update=
843
Update the superblock on each device while assembling the array. The
844
argument given to this flag can be one of
858
option will adjust the superblock of an array what was created on a Sparc
859
machine running a patched 2.2 Linux kernel. This kernel got the
860
alignment of part of the superblock wrong. You can use the
861
.B "\-\-examine \-\-sparc2.2"
864
to see what effect this would have.
868
option will update the
870
field on each superblock to match the minor number of the array being
872
This can be useful if
874
reports a different "Preferred Minor" to
876
In some cases this update will be performed automatically
877
by the kernel driver. In particular the update happens automatically
878
at the first write to an array with redundancy (RAID level 1 or
879
greater) on a 2.6 (or later) kernel.
883
option will change the uuid of the array. If a UUID is given with the
885
option that UUID will be used as a new UUID and will
887
be used to help identify the devices in the array.
890
is given, a random UUID is chosen.
894
option will change the
896
of the array as stored in the superblock. This is only supported for
897
version-1 superblocks.
901
option will change the
903
as recorded in the superblock. For version-0 superblocks, this is the
904
same as updating the UUID.
905
For version-1 superblocks, this involves updating the name.
909
option will cause the array to be marked
911
meaning that any redundancy in the array (e.g. parity for RAID5,
912
copies for RAID1) may be incorrect. This will cause the RAID system
913
to perform a "resync" pass to make sure that all redundant information
918
option allows arrays to be moved between machines with different
920
When assembling such an array for the first time after a move, giving
921
.B "\-\-update=byteorder"
924
to expect superblocks to have their byteorder reversed, and will
925
correct that order before assembling the array. This is only valid
926
with original (Version 0.90) superblocks.
930
option will correct the summaries in the superblock. That is the
931
counts of total, working, active, failed, and spare devices.
935
will rarely be of use. It applies to version 1.1 and 1.2 metadata
936
only (where the metadata is at the start of the device) and is only
937
useful when the component device has changed size (typically become
938
larger). The version 1 metadata records the amount of the device that
939
can be used to store data, so if a device in a version 1.1 or 1.2
940
array becomes larger, the metadata will still be visible, but the
941
extra space will not. In this case it might be useful to assemble the
943
.BR \-\-update=devicesize .
946
to determine the maximum usable amount of space on each device and
947
update the relevant field in the metadata.
951
.B \-\-auto\-update\-homehost
952
This flag is only meaningful with auto-assembly (see discussion below).
953
In that situation, if no suitable arrays are found for this homehost,
955
will rescan for any arrays at all and will assemble them and update the
956
homehost to match the current host.
963
hot-add listed devices. For arrays with redundancy, the listed
964
devices become available as spares. If the array is degraded, it will
965
immediately start recovering data on to one of these spares.
969
re-add a device that was recently removed from an array. This is only
970
needed for arrays that have be built (i.e. with
972
For created arrays, devices are always re-added if that is possible.
973
When re-adding a device, if nothing has changed on the array since the
974
device was removed, no recovery is performed. Also, if the array has
975
a write-intent bitmap, then the recovery performed after a re-add will
976
be limited to those blocks which, according to the bitmap, might have
977
changed since the device was removed.
980
.BR \-r ", " \-\-remove
981
remove listed devices. They must not be active. i.e. they should
982
be failed or spare devices. As well as the name of a device file
991
The first causes all failed device to be removed. The second causes
992
any device which is no longer connected to the system (i.e an 'open'
995
to be removed. This will only succeed for devices that are spares or
996
have already been marked as failed.
999
.BR \-f ", " \-\-fail
1000
mark listed devices as faulty.
1001
As well as the name of a device file, the word
1003
can be given. This will cause any device that has been detached from
1004
the system to be marked as failed. It can then be removed.
1012
.BR \-\-write\-mostly
1013
Subsequent devices that are added or re-added will have the 'write-mostly'
1014
flag set. This is only valid for RAID1 and means that the 'md' driver
1015
will avoid reading from these devices if possible.
1018
Subsequent devices that are added or re-added will have the 'write-mostly'
1022
Each of these options requires that the first device listed is the array
1023
to be acted upon, and the remainder are component devices to be added,
1024
removed, marked as faulty, etc. Several different operations can be
1025
specified for different devices, e.g.
1027
mdadm /dev/md0 \-\-add /dev/sda1 \-\-fail /dev/sdb1 \-\-remove /dev/sdb1
1029
Each operation applies to all devices listed until the next
1032
If an array is using a write-intent bitmap, then devices which have
1033
been removed can be re-added in a way that avoids a full
1034
reconstruction but instead just updates the blocks that have changed
1035
since the device was removed. For arrays with persistent metadata
1036
(superblocks) this is done automatically. For arrays created with
1038
mdadm needs to be told that this device we removed recently with
1041
Devices can only be removed from an array if they are not in active
1042
use, i.e. that must be spares or failed devices. To remove an active
1043
device, it must first be marked as
1049
.BR \-Q ", " \-\-query
1050
Examine a device to see
1051
(1) if it is an md device and (2) if it is a component of an md
1053
Information about what is discovered is presented.
1056
.BR \-D ", " \-\-detail
1057
Print details of one or more md devices.
1060
.BR \-\-detail\-platform
1061
Print details of the platform's RAID capabilities (firmware / hardware
1062
topology) for a given metadata format.
1065
.BR \-Y ", " \-\-export
1070
output will be formatted as
1072
pairs for easy import into the environment.
1075
.BR \-E ", " \-\-examine
1076
Print contents of the metadata stored on the named device(s).
1077
Note the contrast between
1082
applies to devices which are components of an array, while
1084
applies to a whole array which is currently active.
1087
If an array was created on a SPARC machine with a 2.2 Linux kernel
1088
patched with RAID support, the superblock will have been created
1089
incorrectly, or at least incompatibly with 2.4 and later kernels.
1094
will fix the superblock before displaying it. If this appears to do
1095
the right thing, then the array can be successfully assembled using
1096
.BR "\-\-assemble \-\-update=sparc2.2" .
1099
.BR \-X ", " \-\-examine\-bitmap
1100
Report information about a bitmap file.
1101
The argument is either an external bitmap file or an array component
1102
in case of an internal bitmap. Note that running this on an array
1105
does not report the bitmap for that array.
1108
.BR \-R ", " \-\-run
1109
start a partially assembled array. If
1111
did not find enough devices to fully start the array, it might leaving
1112
it partially assembled. If you wish, you can then use
1114
to start the array in degraded mode.
1117
.BR \-S ", " \-\-stop
1118
deactivate array, releasing all resources.
1121
.BR \-o ", " \-\-readonly
1122
mark array as readonly.
1125
.BR \-w ", " \-\-readwrite
1126
mark array as readwrite.
1129
.B \-\-zero\-superblock
1130
If the device contains a valid md superblock, the block is
1131
overwritten with zeros. With
1133
the block where the superblock would be is overwritten even if it
1134
doesn't appear to be valid.
1137
.BR \-t ", " \-\-test
1142
is set to reflect the status of the device. See below in
1147
.BR \-W ", " \-\-wait
1148
For each md device given, wait for any resync, recovery, or reshape
1149
activity to finish before returning.
1151
will return with success if it actually waited for every device
1152
listed, otherwise it will return failure.
1156
For each md device given, or each device in /proc/mdstat if
1158
is given, arrange for the array to be marked clean as soon as possible.
1159
Also, quiesce resync so that the monitor for external metadata arrays
1160
(mdmon) has an opportunity to checkpoint the resync position.
1162
will return with success if the array uses external metadata and we
1163
successfully waited. For native arrays this returns immediately as the
1164
kernel handles both dirty-clean transitions and resync checkpointing in
1165
the kernel at shutdown. No action is taken if safe-mode handling is
1168
.SH For Incremental Assembly mode:
1170
.BR \-\-rebuild\-map ", " \-r
1171
Rebuild the map file
1172
.RB ( /var/run/mdadm/map )
1175
uses to help track which arrays are currently being assembled.
1178
.BR \-\-run ", " \-R
1179
Run any array assembled as soon as a minimal number of devices are
1180
available, rather than waiting until all expected devices are present.
1184
This allows the hot-plug system to prevent arrays from running when it knows
1185
that more disks may arrive later in the discovery process.
1188
.BR \-\-scan ", " \-s
1189
Only meaningful with
1193
file for arrays that are being incrementally assembled and will try to
1194
start any that are not already started. If any such array is listed
1197
as requiring an external bitmap, that bitmap will be attached first.
1199
.SH For Monitor mode:
1201
.BR \-m ", " \-\-mail
1202
Give a mail address to send alerts to.
1205
.BR \-p ", " \-\-program ", " \-\-alert
1206
Give a program to be run whenever an event is detected.
1209
.BR \-y ", " \-\-syslog
1210
Cause all events to be reported through 'syslog'. The messages have
1211
facility of 'daemon' and varying priorities.
1214
.BR \-d ", " \-\-delay
1215
Give a delay in seconds.
1217
polls the md arrays and then waits this many seconds before polling
1218
again. The default is 60 seconds. Since 2.6.16, there is no need to
1219
reduce this as the kernel alerts
1221
immediately when there is any change.
1224
.BR \-r ", " \-\-increment
1225
Give a percentage increment.
1227
will generate RebuildNN events with the given percentage increment.
1230
.BR \-f ", " \-\-daemonise
1233
to run as a background daemon if it decides to monitor anything. This
1234
causes it to fork and run in the child, and to disconnect from the
1235
terminal. The process id of the child is written to stdout.
1238
which will only continue monitoring if a mail address or alert program
1239
is found in the config file.
1242
.BR \-i ", " \-\-pid\-file
1245
is running in daemon mode, write the pid of the daemon process to
1246
the specified file, instead of printing it on standard output.
1249
.BR \-1 ", " \-\-oneshot
1250
Check arrays only once. This will generate
1252
events and more significantly
1258
.B " mdadm \-\-monitor \-\-scan \-1"
1260
from a cron script will ensure regular notification of any degraded arrays.
1263
.BR \-t ", " \-\-test
1266
alert for every array found at startup. This alert gets mailed and
1267
passed to the alert program. This can be used for testing that alert
1268
message do get through successfully.
1274
.B mdadm \-\-assemble
1275
.I md-device options-and-component-devices...
1278
.B mdadm \-\-assemble \-\-scan
1279
.I md-devices-and-options...
1282
.B mdadm \-\-assemble \-\-scan
1286
This usage assembles one or more RAID arrays from pre-existing components.
1287
For each array, mdadm needs to know the md device, the identity of the
1288
array, and a number of component-devices. These can be found in a number of ways.
1290
In the first usage example (without the
1292
the first device given is the md device.
1293
In the second usage example, all devices listed are treated as md
1294
devices and assembly is attempted.
1295
In the third (where no devices are listed) all md devices that are
1296
listed in the configuration file are assembled. If not arrays are
1297
described by the configuration file, then any arrays that
1298
can be found on unused devices will be assembled.
1300
If precisely one device is listed, but
1306
was given and identity information is extracted from the configuration file.
1308
The identity can be given with the
1314
option, will be taken from the md-device record in the config file, or
1315
will be taken from the super block of the first component-device
1316
listed on the command line.
1318
Devices can be given on the
1320
command line or in the config file. Only devices which have an md
1321
superblock which contains the right identity will be considered for
1324
The config file is only used if explicitly named with
1326
or requested with (a possibly implicit)
1331
.B /etc/mdadm/mdadm.conf
1336
is not given, then the config file will only be used to find the
1337
identity of md arrays.
1339
Normally the array will be started after it is assembled. However if
1341
is not given and not all expected drives were listed, then the array
1342
is not started (to guard against usage errors). To insist that the
1343
array be started in this case (as may work for RAID1, 4, 5, 6, or 10),
1352
does not create any entries in
1356
It does record information in
1357
.B /var/run/mdadm/map
1360
to choose the correct name.
1364
detects that udev is not configured, it will create the devices in
1368
In Linux kernels prior to version 2.6.28 there were two distinctly
1369
different types of md devices that could be created: one that could be
1370
partitioned using standard partitioning tools and one that could not.
1371
Since 2.6.28 that distinction is no longer relevant as both type of
1372
devices can be partitioned.
1374
will normally create the type that originally could not be partitioned
1375
as it has a well defined major number (9).
1377
Prior to 2.6.28, it is important that mdadm chooses the correct type
1378
of array device to use. This can be controlled with the
1380
option. In particular, a value of "mdp" or "part" or "p" tells mdadm
1381
to use a partitionable device rather than the default.
1383
In the no-udev case, the value given to
1385
can be suffixed by a number. This tells
1387
to create that number of partition devices rather than the default of 4.
1391
can also be given in the configuration file as a word starting
1393
on the ARRAY line for the relevant array.
1400
and no devices are listed,
1402
will first attempt to assemble all the arrays listed in the config
1405
In no array at listed in the config (other than those marked
1407
it will look through the available devices for possible arrays and
1408
will try to assemble anything that it finds. Arrays which are tagged
1409
as belonging to the given homehost will be assembled and started
1410
normally. Arrays which do not obviously belong to this host are given
1411
names that are expected not to conflict with anything local, and are
1412
started "read-auto" so that nothing is written to any device until the
1413
array is written to. i.e. automatic resync etc is delayed.
1417
finds a consistent set of devices that look like they should comprise
1418
an array, and if the superblock is tagged as belonging to the given
1419
home host, it will automatically choose a device name and try to
1420
assemble the array. If the array uses version-0.90 metadata, then the
1422
number as recorded in the superblock is used to create a name in
1426
If the array uses version-1 metadata, then the
1428
from the superblock is used to similarly create a name in
1430
(the name will have any 'host' prefix stripped first).
1435
cannot find any array for the given host at all, and if
1436
.B \-\-auto\-update\-homehost
1439
will search again for any array (not just an array created for this
1440
host) and will assemble each assuming
1441
.BR \-\-update=homehost .
1442
This will change the host tag in the superblock so that on the next run,
1443
these arrays will be found without the second pass. The intention of
1444
this feature is to support transitioning a set of md arrays to using
1447
The reason for requiring arrays to be tagged with the homehost for
1448
auto assembly is to guard against problems that can arise when moving
1449
devices from one host to another.
1460
.BI \-\-raid\-devices= Z
1464
This usage is similar to
1466
The difference is that it creates an array without a superblock. With
1467
these arrays there is no difference between initially creating the array and
1468
subsequently assembling the array, except that hopefully there is useful
1469
data there in the second case.
1471
The level may raid0, linear, raid1, raid10, multipath, or faulty, or
1472
one of their synonyms. All devices must be listed and the array will
1473
be started once complete. It will often be appropriate to use
1474
.B \-\-assume\-clean
1475
with levels raid1 or raid10.
1486
.BI \-\-raid\-devices= Z
1490
This usage will initialise a new md array, associate some devices with
1491
it, and activate the array.
1493
The named device will normally not exist when
1494
.I "mdadm \-\-create"
1495
is run, but will be created by
1497
once the array becomes active.
1499
As devices are added, they are checked to see if they contain RAID
1500
superblocks or filesystems. They are also checked to see if the variance in
1501
device size exceeds 1%.
1503
If any discrepancy is found, the array will not automatically be run, though
1506
can override this caution.
1508
To create a "degraded" array in which some devices are missing, simply
1509
give the word "\fBmissing\fP"
1510
in place of a device name. This will cause
1512
to leave the corresponding slot in the array empty.
1513
For a RAID4 or RAID5 array at most one slot can be
1514
"\fBmissing\fP"; for a RAID6 array at most two slots.
1515
For a RAID1 array, only one real device needs to be given. All of the
1519
When creating a RAID5 array,
1521
will automatically create a degraded array with an extra spare drive.
1522
This is because building the spare into a degraded array is in general
1523
faster than resyncing the parity on a non-degraded, but not clean,
1524
array. This feature can be overridden with the
1528
When creating an array with version-1 metadata a name for the array is
1530
If this is not given with the
1534
will choose a name based on the last component of the name of the
1535
device being created. So if
1537
is being created, then the name
1542
is being created, then the name
1546
When creating a partition based array, using
1548
with version-1.x metadata, the partition type should be set to
1550
(non fs-data). This type selection allows for greater precision since
1551
using any other [RAID auto-detect (0xFD) or a GNU/Linux partition (0x83)],
1552
might create problems in the event of array recovery through a live cdrom.
1554
A new array will normally get a randomly assigned 128bit UUID which is
1555
very likely to be unique. If you have a specific need, you can choose
1556
a UUID for the array by giving the
1558
option. Be warned that creating two arrays with the same UUID is a
1559
recipe for disaster. Also, using
1561
when creating a v0.90 array will silently override any
1566
.\"option is given, it is not necessary to list any component-devices in this command.
1567
.\"They can be added later, before a
1571
.\"is given, the apparent size of the smallest drive given is used.
1573
When creating an array within a
1576
can be given either the list of devices to use, or simply the name of
1577
the container. The former case gives control over which devices in
1578
the container will be used for the array. The latter case allows
1580
to automatically choose which devices to use based on how much spare
1583
The General Management options that are valid with
1588
insist on running the array even if some devices look like they might
1593
start the array readonly \(em not supported yet.
1600
.I options... devices...
1603
This usage will allow individual devices in an array to be failed,
1604
removed or added. It is possible to perform multiple operations with
1605
on command. For example:
1607
.B " mdadm /dev/md0 \-f /dev/hda1 \-r /dev/hda1 \-a /dev/hda1"
1613
and will then remove it from the array and finally add it back
1614
in as a spare. However only one md array can be affected by a single
1617
When a device is added to an active array, mdadm checks to see if it
1618
has metadata on it which suggests that it was recently a member of the
1619
array. If it does, it tried to "re-add" the device. If there have
1620
been no changes since the device was removed, or if the array has a
1621
write-intent bitmap which has recorded whatever changes there were,
1622
then the device will immediately become a full member of the array and
1623
those differences recorded in the bitmap will be resolved.
1633
MISC mode includes a number of distinct operations that
1634
operate on distinct devices. The operations are:
1637
The device is examined to see if it is
1638
(1) an active md array, or
1639
(2) a component of an md array.
1640
The information discovered is reported.
1644
The device should be an active md device.
1646
will display a detailed description of the array.
1650
will cause the output to be less detailed and the format to be
1651
suitable for inclusion in
1652
.BR /etc/mdadm.conf .
1655
will normally be 0 unless
1657
failed to get useful information about the device(s); however, if the
1659
option is given, then the exit status will be:
1663
The array is functioning normally.
1666
The array has at least one failed device.
1669
The array has multiple failed devices such that it is unusable.
1672
There was an error while trying to get information about the device.
1676
.B \-\-detail\-platform
1677
Print detail of the platform's RAID capabilities (firmware / hardware
1678
topology). If the metadata is specified with
1682
then the return status will be:
1686
metadata successfully enumerated its platform components on this system
1689
metadata is platform independent
1692
metadata failed to find its platform components on this system
1697
The device should be a component of an md array.
1699
will read the md superblock of the device and display the contents.
1704
is given, then multiple devices that are components of the one array
1705
are grouped together and reported in a single entry suitable
1707
.BR /etc/mdadm.conf .
1711
without listing any devices will cause all devices listed in the
1712
config file to be examined.
1716
The devices should be active md arrays which will be deactivated, as
1717
long as they are not currently in use.
1721
This will fully activate a partially assembled md array.
1725
This will mark an active array as read-only, providing that it is
1726
not currently being used.
1732
array back to being read/write.
1736
For all operations except
1739
will cause the operation to be applied to all arrays listed in
1744
causes all devices listed in the config file to be examined.
1750
.B mdadm \-\-monitor
1751
.I options... devices...
1756
to periodically poll a number of md arrays and to report on any events
1759
will never exit once it decides that there are arrays to be checked,
1760
so it should normally be run in the background.
1762
As well as reporting events,
1764
may move a spare drive from one array to another if they are in the
1767
and if the destination array has a failed drive but no spares.
1769
If any devices are listed on the command line,
1771
will only monitor those devices. Otherwise all arrays listed in the
1772
configuration file will be monitored. Further, if
1774
is given, then any other md devices that appear in
1776
will also be monitored.
1778
The result of monitoring the arrays is the generation of events.
1779
These events are passed to a separate program (if specified) and may
1780
be mailed to a given E-mail address.
1782
When passing events to a program, the program is run once for each event,
1783
and is given 2 or 3 command-line arguments: the first is the
1784
name of the event (see below), the second is the name of the
1785
md device which is affected, and the third is the name of a related
1786
device if relevant (such as a component device that has failed).
1790
is given, then a program or an E-mail address must be specified on the
1791
command line or in the config file. If neither are available, then
1793
will not monitor anything.
1797
will continue monitoring as long as something was found to monitor. If
1798
no program or email is given, then each event is reported to
1801
The different events are:
1805
.B DeviceDisappeared
1806
An md array which previously was configured appears to no longer be
1807
configured. (syslog priority: Critical)
1811
was told to monitor an array which is RAID0 or Linear, then it will
1813
.B DeviceDisappeared
1814
with the extra information
1816
This is because RAID0 and Linear do not support the device-failed,
1817
hot-spare and resync operations which are monitored.
1821
An md array started reconstruction. (syslog priority: Warning)
1827
is a two-digit number (ie. 05, 48). This indicates that rebuild
1828
has passed that many percent of the total. The events are generated
1829
with fixed increment since 0. Increment size may be specified with
1830
a commandline option (default is 20). (syslog priority: Warning)
1834
An md array that was rebuilding, isn't any more, either because it
1835
finished normally or was aborted. (syslog priority: Warning)
1839
An active component device of an array has been marked as
1840
faulty. (syslog priority: Critical)
1844
A spare component device which was being rebuilt to replace a faulty
1845
device has failed. (syslog priority: Critical)
1849
A spare component device which was being rebuilt to replace a faulty
1850
device has been successfully rebuilt and has been made active.
1851
(syslog priority: Info)
1855
A new md array has been detected in the
1857
file. (syslog priority: Info)
1861
A newly noticed array appears to be degraded. This message is not
1864
notices a drive failure which causes degradation, but only when
1866
notices that an array is degraded when it first sees the array.
1867
(syslog priority: Critical)
1871
A spare drive has been moved from one array in a
1873
to another to allow a failed drive to be replaced.
1874
(syslog priority: Info)
1880
has been told, via the config file, that an array should have a certain
1881
number of spare devices, and
1883
detects that it has fewer than this number when it first sees the
1884
array, it will report a
1887
(syslog priority: Warning)
1891
An array was found at startup, and the
1894
(syslog priority: Info)
1904
cause Email to be sent. All events cause the program to be run.
1905
The program is run with two or three arguments: the event
1906
name, the array device and possibly a second device.
1908
Each event has an associated array device (e.g.
1910
and possibly a second device. For
1915
the second device is the relevant component device.
1918
the second device is the array that the spare was moved from.
1922
to move spares from one array to another, the different arrays need to
1923
be labeled with the same
1925
in the configuration file. The
1927
name can be any string; it is only necessary that different spare
1928
groups use different names.
1932
detects that an array in a spare group has fewer active
1933
devices than necessary for the complete array, and has no spare
1934
devices, it will look for another array in the same spare group that
1935
has a full complement of working drive and a spare. It will then
1936
attempt to remove the spare from the second drive and add it to the
1938
If the removal succeeds but the adding fails, then it is added back to
1942
The GROW mode is used for changing the size or shape of an active
1944
For this to work, the kernel must support the necessary change.
1945
Various types of growth are being added during 2.6 development,
1946
including restructuring a RAID5 array to have more active devices.
1948
Currently the only support available is to
1950
change the "size" attribute
1951
for RAID1, RAID5 and RAID6.
1953
increase the "raid\-devices" attribute of RAID1, RAID5, and RAID6.
1955
add a write-intent bitmap to any array which supports these bitmaps, or
1956
remove a write-intent bitmap from such an array.
1959
GROW mode is not currently supported for
1961
or arrays inside containers.
1964
Normally when an array is built the "size" it taken from the smallest
1965
of the drives. If all the small drives in an arrays are, one at a
1966
time, removed and replaced with larger drives, then you could have an
1967
array of large drives with only a small amount used. In this
1968
situation, changing the "size" with "GROW" mode will allow the extra
1969
space to start being used. If the size is increased in this way, a
1970
"resync" process will start to make sure the new parts of the array
1973
Note that when an array changes size, any filesystem that may be
1974
stored in the array will not automatically grow to use the space. The
1975
filesystem will need to be explicitly told to use the extra space.
1977
Also the size of an array cannot be changed while it has an active
1978
bitmap. If an array has a bitmap, it must be removed before the size
1979
can be changed. Once the change it complete a new bitmap can be created.
1981
.SS RAID\-DEVICES CHANGES
1983
A RAID1 array can work with any number of devices from 1 upwards
1984
(though 1 is not very useful). There may be times which you want to
1985
increase or decrease the number of active devices. Note that this is
1986
different to hot-add or hot-remove which changes the number of
1989
When reducing the number of devices in a RAID1 array, the slots which
1990
are to be removed from the array must already be vacant. That is, the
1991
devices which were in those slots must be failed and removed.
1993
When the number of devices is increased, any hot spares that are
1994
present will be activated immediately.
1996
Increasing the number of active devices in a RAID5 is much more
1997
effort. Every block in the array will need to be read and written
1998
back to a new location. From 2.6.17, the Linux Kernel is able to do
1999
this safely, including restarting an interrupted "reshape".
2001
When relocating the first few stripes on a RAID5, it is not possible
2002
to keep the data on disk completely consistent and crash-proof. To
2003
provide the required safety, mdadm disables writes to the array while
2004
this "critical section" is reshaped, and takes a backup of the data
2005
that is in that section. This backup is normally stored in any spare
2006
devices that the array has, however it can also be stored in a
2007
separate file specified with the
2009
option. If this option is used, and the system does crash during the
2010
critical period, the same file must be passed to
2012
to restore the backup and reassemble the array.
2016
A write-intent bitmap can be added to, or removed from, an active
2017
array. Either internal bitmaps, or bitmaps stored in a separate file,
2018
can be added. Note that if you add a bitmap stored in a file which is
2019
in a filesystem that is on the RAID array being affected, the system
2020
will deadlock. The bitmap must be on a separate filesystem.
2022
.SH INCREMENTAL MODE
2026
.B mdadm \-\-incremental
2032
.B mdadm \-\-incremental \-\-rebuild
2035
.B mdadm \-\-incremental \-\-run \-\-scan
2038
This mode is designed to be used in conjunction with a device
2039
discovery system. As devices are found in a system, they can be
2041
.B "mdadm \-\-incremental"
2042
to be conditionally added to an appropriate array.
2044
If the device passed is a
2046
device created by a previous call to
2048
then rather than trying to add that device to an array, all the arrays
2049
described by the metadata of the container will be started.
2052
performs a number of tests to determine if the device is part of an
2053
array, and which array it should be part of. If an appropriate array
2054
is found, or can be created,
2056
adds the device to the array and conditionally starts the array.
2060
will only add devices to an array which were previously working
2061
(active or spare) parts of that array. It does not currently support
2062
automatic inclusion of a new drive as a spare in some array.
2066
makes are as follow:
2068
Is the device permitted by
2070
That is, is it listed in a
2072
line in that file. If
2074
is absent then the default it to allow any device. Similar if
2076
contains the special word
2078
then any device is allowed. Otherwise the device name given to
2080
must match one of the names or patterns in a
2085
Does the device have a valid md superblock. If a specific metadata
2086
version is request with
2090
then only that style of metadata is accepted, otherwise
2092
finds any known version of metadata. If no
2094
metadata is found, the device is rejected.
2098
Does the metadata match an expected array?
2099
The metadata can match in two ways. Either there is an array listed
2102
which identifies the array (either by UUID, by name, by device list,
2103
or by minor-number), or the array was created with a
2109
or on the command line.
2112
is not able to positively identify the array as belonging to the
2113
current host, the device will be rejected.
2117
keeps a list of arrays that it has partially assembled in
2118
.B /var/run/mdadm/map
2120
.B /var/run/mdadm.map
2121
if the directory doesn't exist. Or maybe even
2122
.BR /dev/.mdadm.map ).
2123
If no array exists which matches
2124
the metadata on the new device,
2126
must choose a device name and unit number. It does this based on any
2129
or any name information stored in the metadata. If this name
2130
suggests a unit number, that number will be used, otherwise a free
2131
unit number will be chosen. Normally
2133
will prefer to create a partitionable array, however if the
2137
suggests that a non-partitionable array is preferred, that will be
2140
If the array is not found in the config file and its metadata does not
2141
identify it as belonging to the "homehost", then
2143
will choose a name for the array which is certain not to conflict with
2144
any array which does belong to this host. It does this be adding an
2145
underscore and a small number to the name preferred by the metadata.
2147
Once an appropriate array is found or created and the device is added,
2149
must decide if the array is ready to be started. It will
2150
normally compare the number of available (non-spare) devices to the
2151
number of devices that the metadata suggests need to be active. If
2152
there are at least that many, the array will be started. This means
2153
that if any devices are missing the array will not be restarted.
2159
in which case the array will be run as soon as there are enough
2160
devices present for the data to be accessible. For a RAID1, that
2161
means one device will start the array. For a clean RAID5, the array
2162
will be started as soon as all but one drive is present.
2164
Note that neither of these approaches is really ideal. If it can
2165
be known that all device discovery has completed, then
2169
can be run which will try to start all arrays that are being
2170
incrementally assembled. They are started in "read-auto" mode in
2171
which they are read-only until the first write request. This means
2172
that no metadata updates are made and no attempt at resync or recovery
2173
happens. Further devices that are found before the first write can
2174
still be added safely.
2177
This section describes environment variables that affect how mdadm
2182
Setting this value to 1 will prevent mdadm from automatically launching
2183
mdmon. This variable is intended primarily for debugging mdadm/mdmon.
2189
does not create any device nodes in /dev, but leaves that task to
2193
appears not to be configured, or if this environment variable is set
2196
will create and devices that are needed.
2200
.B " mdadm \-\-query /dev/name-of-device"
2202
This will find out if a given device is a RAID array, or is part of
2203
one, and will provide brief information about the device.
2205
.B " mdadm \-\-assemble \-\-scan"
2207
This will assemble and start all arrays listed in the standard config
2208
file. This command will typically go in a system startup file.
2210
.B " mdadm \-\-stop \-\-scan"
2212
This will shut down all arrays that can be shut down (i.e. are not
2213
currently in use). This will typically go in a system shutdown script.
2215
.B " mdadm \-\-follow \-\-scan \-\-delay=120"
2217
If (and only if) there is an Email address or program given in the
2218
standard config file, then
2219
monitor the status of all arrays listed in that file by
2220
polling them ever 2 minutes.
2222
.B " mdadm \-\-create /dev/md0 \-\-level=1 \-\-raid\-devices=2 /dev/hd[ac]1"
2224
Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
2227
.B " echo 'DEVICE /dev/hd*[0\-9] /dev/sd*[0\-9]' > mdadm.conf"
2229
.B " mdadm \-\-detail \-\-scan >> mdadm.conf"
2231
This will create a prototype config file that describes currently
2232
active arrays that are known to be made from partitions of IDE or SCSI drives.
2233
This file should be reviewed before being used as it may
2234
contain unwanted detail.
2236
.B " echo 'DEVICE /dev/hd[a\-z] /dev/sd*[a\-z]' > mdadm.conf"
2238
.B " mdadm \-\-examine \-\-scan \-\-config=mdadm.conf >> mdadm.conf"
2240
This will find arrays which could be assembled from existing IDE and
2241
SCSI whole drives (not partitions), and store the information in the
2242
format of a config file.
2243
This file is very likely to contain unwanted detail, particularly
2246
entries. It should be reviewed and edited before being used as an
2249
.B " mdadm \-\-examine \-\-brief \-\-scan \-\-config=partitions"
2251
.B " mdadm \-Ebsc partitions"
2253
Create a list of devices by reading
2254
.BR /proc/partitions ,
2255
scan these for RAID superblocks, and printout a brief listing of all
2258
.B " mdadm \-Ac partitions \-m 0 /dev/md0"
2260
Scan all partitions and devices listed in
2261
.BR /proc/partitions
2264
out of all such devices with a RAID superblock with a minor number of 0.
2266
.B " mdadm \-\-monitor \-\-scan \-\-daemonise > /var/run/mdadm"
2268
If config file contains a mail address or alert program, run mdadm in
2269
the background in monitor mode monitoring all md devices. Also write
2270
pid of mdadm daemon to
2271
.BR /var/run/mdadm .
2273
.B " mdadm \-Iq /dev/somedevice"
2275
Try to incorporate newly discovered device into some array as
2278
.B " mdadm \-\-incremental \-\-rebuild \-\-run \-\-scan"
2280
Rebuild the array map from any current arrays, and then start any that
2283
.B " mdadm /dev/md4 --fail detached --remove detached"
2285
Any devices which are components of /dev/md4 will be marked as faulty
2286
and then remove from the array.
2288
.B " mdadm --create /dev/md/ddf --metadata=ddf --raid-disks 6 /dev/sd[a-f]"
2290
Create a DDF array over 6 devices.
2292
.B " mdadm --create /dev/md/home -n3 -l5 -z 30000000 /dev/md/ddf"
2294
Create a RAID5 array over any 3 devices in the given DDF set. Use
2295
only 30 gigabytes of each device.
2297
.B " mdadm -A /dev/md/ddf1 /dev/sd[a-f]"
2299
Assemble a pre-exist ddf array.
2301
.B " mdadm -I /dev/md/ddf1"
2303
Assemble all arrays contained in the ddf array, assigning names as
2306
.B " mdadm \-\-create \-\-help"
2308
Provide help about the Create mode.
2310
.B " mdadm \-\-config \-\-help"
2312
Provide help about the format of the config file.
2314
.B " mdadm \-\-help"
2316
Provide general help.
2326
lists all active md devices with information about them.
2328
uses this to find arrays when
2330
is given in Misc mode, and to monitor array reconstruction
2335
The config file lists which devices may be scanned to see if
2336
they contain MD super block, and gives identifying information
2337
(e.g. UUID) about known MD arrays. See
2341
.SS /var/run/mdadm/map
2344
mode is used, this file gets a list of arrays currently being created.
2347
does not exist as a directory, then
2348
.B /var/run/mdadm.map
2351
is not available (as may be the case during early boot),
2353
is used on the basis that
2355
is usually available very early in boot.
2360
understand two sorts of names for array devices.
2362
The first is the so-called 'standard' format name, which matches the
2363
names used by the kernel and which appear in
2366
The second sort can be freely chosen, but must reside in
2368
When giving a device name to
2370
to create or assemble an array, either full path name such as
2374
can be given, or just the suffix of the second sort of name, such as
2380
chooses device names during auto-assembly or incremental assembly, it
2381
will sometimes add a small sequence number to the end of the name to
2382
avoid conflicted between multiple arrays that have the same name. If
2384
can reasonably determine that the array really is meant for this host,
2385
either by a hostname in the metadata, or by the presence of the array
2386
in /etc/mdadm.conf, then it will leave off the suffix if possible.
2387
Also if the homehost is specified as
2390
will only use a suffix if a different array of the same name already
2391
exists or is listed in the config file.
2393
The standard names for non-partitioned arrays (the only sort of md
2394
array available in 2.4 and earlier) are of the form
2398
where NN is a number.
2399
The standard names for partitionable arrays (as available from 2.6
2400
onwards) are of the form
2404
Partition numbers should be indicated by added "pMM" to these, thus "/dev/md/d1p2".
2406
From kernel version, 2.6.28 the "non-partitioned array" can actually
2407
be partitioned. So the "md_dNN" names are no longer needed, and
2408
partitions such as "/dev/mdNNpXX" are possible.
2412
was previously known as
2416
is completely separate from the
2418
package, and does not use the
2420
configuration file at all.
2423
For further information on mdadm usage, MD and the various levels of
2426
.B http://linux\-raid.osdl.org/
2428
(based upon Jakob \(/Ostergaard's Software\-RAID.HOWTO)
2430
.\"for new releases of the RAID driver check out:
2433
.\".UR ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2434
.\"ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2439
.\".UR http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2440
.\"http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2443
The latest version of
2445
should always be available from
2447
.B http://www.kernel.org/pub/linux/utils/raid/mdadm/