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
126 Linear and RAID levels 0/1/4/5/6,
127 changing the RAID level between 0, 1, 5, and 6, and between 0 and 10,
128 changing the chunk size and layout for RAID 0,4,5,6, as well as adding or
129 removing a write-intent bitmap.
132 .B "Incremental Assembly"
133 Add a single device to an appropriate array. If the addition of the
134 device makes the array runnable, the array will be started.
135 This provides a convenient interface to a
137 system. As each device is detected,
139 has a chance to include it in some array as appropriate.
142 flag is passed in we will remove the device from any active array
143 instead of adding it.
149 in this mode, then any arrays within that container will be assembled
154 This is for doing things to specific components of an array such as
155 adding new spares and removing faulty devices.
159 This is an 'everything else' mode that supports operations on active
160 arrays, operations on component devices such as erasing old superblocks, and
161 information gathering operations.
162 .\"This mode allows operations on independent devices such as examine MD
163 .\"superblocks, erasing old superblocks and stopping active arrays.
167 This mode does not act on a specific device or array, but rather it
168 requests the Linux Kernel to activate any auto-detected arrays.
171 .SH Options for selecting a mode are:
174 .BR \-A ", " \-\-assemble
175 Assemble a pre-existing array.
178 .BR \-B ", " \-\-build
179 Build a legacy array without superblocks.
182 .BR \-C ", " \-\-create
186 .BR \-F ", " \-\-follow ", " \-\-monitor
192 .BR \-G ", " \-\-grow
193 Change the size or shape of an active array.
196 .BR \-I ", " \-\-incremental
197 Add/remove a single device to/from an appropriate array, and possibly start the array.
201 Request that the kernel starts any auto-detected arrays. This can only
204 is compiled into the kernel \(em not if it is a module.
205 Arrays can be auto-detected by the kernel if all the components are in
206 primary MS-DOS partitions with partition type
208 and all use v0.90 metadata.
209 In-kernel autodetect is not recommended for new installations. Using
211 to detect and assemble arrays \(em possibly in an
213 \(em is substantially more flexible and should be preferred.
216 If a device is given before any options, or if the first option is
221 then the MANAGE mode is assumed.
222 Anything other than these will cause the
226 .SH Options that are not mode-specific are:
229 .BR \-h ", " \-\-help
230 Display general help message or, after one of the above options, a
231 mode-specific help message.
235 Display more detailed help about command line parsing and some commonly
239 .BR \-V ", " \-\-version
240 Print version information for mdadm.
243 .BR \-v ", " \-\-verbose
244 Be more verbose about what is happening. This can be used twice to be
246 The extra verbosity currently only affects
247 .B \-\-detail \-\-scan
249 .BR "\-\-examine \-\-scan" .
252 .BR \-q ", " \-\-quiet
253 Avoid printing purely informative messages. With this,
255 will be silent unless there is something really important to report.
258 .BR \-f ", " \-\-force
259 Be more forceful about certain operations. See the various modes for
260 the exact meaning of this option in different contexts.
263 .BR \-c ", " \-\-config=
264 Specify the config file. Default is to use
265 .BR /etc/mdadm.conf ,
266 or if that is missing then
267 .BR /etc/mdadm/mdadm.conf .
268 If the config file given is
270 then nothing will be read, but
272 will act as though the config file contained exactly
273 .B "DEVICE partitions containers"
276 to find a list of devices to scan, and
278 to find a list of containers to examine.
281 is given for the config file, then
283 will act as though the config file were empty.
286 .BR \-s ", " \-\-scan
289 for missing information.
290 In general, this option gives
292 permission to get any missing information (like component devices,
293 array devices, array identities, and alert destination) from the
294 configuration file (see previous option);
295 one exception is MISC mode when using
301 says to get a list of array devices from
305 .BR \-e ", " \-\-metadata=
306 Declare the style of RAID metadata (superblock) to be used. The
307 default is {DEFAULT_METADATA} for
309 and to guess for other operations.
310 The default can be overridden by setting the
319 .ie '{DEFAULT_METADATA}'0.90'
320 .IP "0, 0.90, default"
324 Use the original 0.90 format superblock. This format limits arrays to
325 28 component devices and limits component devices of levels 1 and
326 greater to 2 terabytes. It is also possible for there to be confusion
327 about whether the superblock applies to a whole device or just the
328 last partition, if that partition starts on a 64K boundary.
329 .ie '{DEFAULT_METADATA}'0.90'
330 .IP "1, 1.0, 1.1, 1.2"
332 .IP "1, 1.0, 1.1, 1.2 default"
334 Use the new version-1 format superblock. This has fewer restrictions.
335 It can easily be moved between hosts with different endian-ness, and a
336 recovery operation can be checkpointed and restarted. The different
337 sub-versions store the superblock at different locations on the
338 device, either at the end (for 1.0), at the start (for 1.1) or 4K from
339 the start (for 1.2). "1" is equivalent to "1.0".
340 'if '{DEFAULT_METADATA}'1.2' "default" is equivalent to "1.2".
342 Use the "Industry Standard" DDF (Disk Data Format) format defined by
344 When creating a DDF array a
346 will be created, and normal arrays can be created in that container.
348 Use the Intel(R) Matrix Storage Manager metadata format. This creates a
350 which is managed in a similar manner to DDF, and is supported by an
351 option-rom on some platforms:
353 .B http://www.intel.com/design/chipsets/matrixstorage_sb.htm
359 This will override any
361 setting in the config file and provides the identity of the host which
362 should be considered the home for any arrays.
364 When creating an array, the
366 will be recorded in the metadata. For version-1 superblocks, it will
367 be prefixed to the array name. For version-0.90 superblocks, part of
368 the SHA1 hash of the hostname will be stored in the later half of the
371 When reporting information about an array, any array which is tagged
372 for the given homehost will be reported as such.
374 When using Auto-Assemble, only arrays tagged for the given homehost
375 will be allowed to use 'local' names (i.e. not ending in '_' followed
376 by a digit string). See below under
377 .BR "Auto Assembly" .
379 .SH For create, build, or grow:
382 .BR \-n ", " \-\-raid\-devices=
383 Specify the number of active devices in the array. This, plus the
384 number of spare devices (see below) must equal the number of
386 (including "\fBmissing\fP" devices)
387 that are listed on the command line for
389 Setting a value of 1 is probably
390 a mistake and so requires that
392 be specified first. A value of 1 will then be allowed for linear,
393 multipath, RAID0 and RAID1. It is never allowed for RAID4, RAID5 or RAID6.
395 This number can only be changed using
397 for RAID1, RAID4, RAID5 and RAID6 arrays, and only on kernels which provide
398 the necessary support.
401 .BR \-x ", " \-\-spare\-devices=
402 Specify the number of spare (eXtra) devices in the initial array.
403 Spares can also be added
404 and removed later. The number of component devices listed
405 on the command line must equal the number of RAID devices plus the
406 number of spare devices.
409 .BR \-z ", " \-\-size=
410 Amount (in Kibibytes) of space to use from each drive in RAID levels 1/4/5/6.
411 This must be a multiple of the chunk size, and must leave about 128Kb
412 of space at the end of the drive for the RAID superblock.
413 If this is not specified
414 (as it normally is not) the smallest drive (or partition) sets the
415 size, though if there is a variance among the drives of greater than 1%, a warning is
418 A suffix of 'M' or 'G' can be given to indicate Megabytes or
419 Gigabytes respectively.
421 This value can be set with
423 for RAID level 1/4/5/6. If the array was created with a size smaller
424 than the currently active drives, the extra space can be accessed
427 The size can be given as
429 which means to choose the largest size that fits on all current drives.
431 Before reducing the size of the array (with
432 .BR "\-\-grow \-\-size=" )
433 you should make sure that space isn't needed. If the device holds a
434 filesystem, you would need to resize the filesystem to use less space.
436 After reducing the array size you should check that the data stored in
437 the device is still available. If the device holds a filesystem, then
438 an 'fsck' of the filesystem is a minimum requirement. If there are
439 problems the array can be made bigger again with no loss with another
440 .B "\-\-grow \-\-size="
443 This value can not be used with
445 metadata such as DDF and IMSM.
448 .BR \-Z ", " \-\-array\-size=
449 This is only meaningful with
451 and its effect is not persistent: when the array is stopped and
452 restarted the default array size will be restored.
454 Setting the array-size causes the array to appear smaller to programs
455 that access the data. This is particularly needed before reshaping an
456 array so that it will be smaller. As the reshape is not reversible,
457 but setting the size with
459 is, it is required that the array size is reduced as appropriate
460 before the number of devices in the array is reduced.
462 Before reducing the size of the array you should make sure that space
463 isn't needed. If the device holds a filesystem, you would need to
464 resize the filesystem to use less space.
466 After reducing the array size you should check that the data stored in
467 the device is still available. If the device holds a filesystem, then
468 an 'fsck' of the filesystem is a minimum requirement. If there are
469 problems the array can be made bigger again with no loss with another
470 .B "\-\-grow \-\-array\-size="
473 A suffix of 'M' or 'G' can be given to indicate Megabytes or
474 Gigabytes respectively.
477 restores the apparent size of the array to be whatever the real
478 amount of available space is.
481 .BR \-c ", " \-\-chunk=
482 Specify chunk size of kibibytes. The default when creating an
483 array is 512KB. To ensure compatibility with earlier versions, the
484 default when Building and array with no persistent metadata is 64KB.
485 This is only meaningful for RAID0, RAID4, RAID5, RAID6, and RAID10.
487 RAID4, RAID5, RAID6, and RAID10 require the chunk size to be a power
488 of 2. In any case it must be a multiple of 4KB.
490 A suffix of 'M' or 'G' can be given to indicate Megabytes or
491 Gigabytes respectively.
495 Specify rounding factor for a Linear array. The size of each
496 component will be rounded down to a multiple of this size.
497 This is a synonym for
499 but highlights the different meaning for Linear as compared to other
500 RAID levels. The default is 64K if a kernel earlier than 2.6.16 is in
501 use, and is 0K (i.e. no rounding) in later kernels.
504 .BR \-l ", " \-\-level=
505 Set RAID level. When used with
507 options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
508 raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty, container.
509 Obviously some of these are synonymous.
513 metadata type is requested, only the
515 level is permitted, and it does not need to be explicitly given.
519 only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid.
523 to change the RAID level in some cases. See LEVEL CHANGES below.
526 .BR \-p ", " \-\-layout=
527 This option configures the fine details of data layout for RAID5, RAID6,
528 and RAID10 arrays, and controls the failure modes for
531 The layout of the RAID5 parity block can be one of
532 .BR left\-asymmetric ,
533 .BR left\-symmetric ,
534 .BR right\-asymmetric ,
535 .BR right\-symmetric ,
536 .BR la ", " ra ", " ls ", " rs .
538 .BR left\-symmetric .
540 It is also possible to cause RAID5 to use a RAID4-like layout by
546 Finally for RAID5 there are DDF\-compatible layouts,
547 .BR ddf\-zero\-restart ,
548 .BR ddf\-N\-restart ,
550 .BR ddf\-N\-continue .
552 These same layouts are available for RAID6. There are also 4 layouts
553 that will provide an intermediate stage for converting between RAID5
554 and RAID6. These provide a layout which is identical to the
555 corresponding RAID5 layout on the first N\-1 devices, and has the 'Q'
556 syndrome (the second 'parity' block used by RAID6) on the last device.
558 .BR left\-symmetric\-6 ,
559 .BR right\-symmetric\-6 ,
560 .BR left\-asymmetric\-6 ,
561 .BR right\-asymmetric\-6 ,
563 .BR parity\-first\-6 .
565 When setting the failure mode for level
568 .BR write\-transient ", " wt ,
569 .BR read\-transient ", " rt ,
570 .BR write\-persistent ", " wp ,
571 .BR read\-persistent ", " rp ,
573 .BR read\-fixable ", " rf ,
574 .BR clear ", " flush ", " none .
576 Each failure mode can be followed by a number, which is used as a period
577 between fault generation. Without a number, the fault is generated
578 once on the first relevant request. With a number, the fault will be
579 generated after that many requests, and will continue to be generated
580 every time the period elapses.
582 Multiple failure modes can be current simultaneously by using the
584 option to set subsequent failure modes.
586 "clear" or "none" will remove any pending or periodic failure modes,
587 and "flush" will clear any persistent faults.
589 Finally, the layout options for RAID10 are one of 'n', 'o' or 'f' followed
590 by a small number. The default is 'n2'. The supported options are:
593 signals 'near' copies. Multiple copies of one data block are at
594 similar offsets in different devices.
597 signals 'offset' copies. Rather than the chunks being duplicated
598 within a stripe, whole stripes are duplicated but are rotated by one
599 device so duplicate blocks are on different devices. Thus subsequent
600 copies of a block are in the next drive, and are one chunk further
605 (multiple copies have very different offsets).
606 See md(4) for more detail about 'near', 'offset', and 'far'.
608 The number is the number of copies of each datablock. 2 is normal, 3
609 can be useful. This number can be at most equal to the number of
610 devices in the array. It does not need to divide evenly into that
611 number (e.g. it is perfectly legal to have an 'n2' layout for an array
612 with an odd number of devices).
614 When an array is converted between RAID5 and RAID6 an intermediate
615 RAID6 layout is used in which the second parity block (Q) is always on
616 the last device. To convert a RAID5 to RAID6 and leave it in this new
617 layout (which does not require re-striping) use
618 .BR \-\-layout=preserve .
619 This will try to avoid any restriping.
621 The converse of this is
622 .B \-\-layout=normalise
623 which will change a non-standard RAID6 layout into a more standard
630 (thus explaining the p of
634 .BR \-b ", " \-\-bitmap=
635 Specify a file to store a write-intent bitmap in. The file should not
638 is also given. The same file should be provided
639 when assembling the array. If the word
641 is given, then the bitmap is stored with the metadata on the array,
642 and so is replicated on all devices. If the word
646 mode, then any bitmap that is present is removed.
648 To help catch typing errors, the filename must contain at least one
649 slash ('/') if it is a real file (not 'internal' or 'none').
651 Note: external bitmaps are only known to work on ext2 and ext3.
652 Storing bitmap files on other filesystems may result in serious problems.
655 .BR \-\-bitmap\-chunk=
656 Set the chunksize of the bitmap. Each bit corresponds to that many
657 Kilobytes of storage.
658 When using a file based bitmap, the default is to use the smallest
659 size that is at-least 4 and requires no more than 2^21 chunks.
662 bitmap, the chunksize defaults to 64Meg, or larger if necessary to
663 fit the bitmap into the available space.
665 A suffix of 'M' or 'G' can be given to indicate Megabytes or
666 Gigabytes respectively.
669 .BR \-W ", " \-\-write\-mostly
670 subsequent devices listed in a
675 command will be flagged as 'write-mostly'. This is valid for RAID1
676 only and means that the 'md' driver will avoid reading from these
677 devices if at all possible. This can be useful if mirroring over a
681 .BR \-\-write\-behind=
682 Specify that write-behind mode should be enabled (valid for RAID1
683 only). If an argument is specified, it will set the maximum number
684 of outstanding writes allowed. The default value is 256.
685 A write-intent bitmap is required in order to use write-behind
686 mode, and write-behind is only attempted on drives marked as
690 .BR \-\-assume\-clean
693 that the array pre-existed and is known to be clean. It can be useful
694 when trying to recover from a major failure as you can be sure that no
695 data will be affected unless you actually write to the array. It can
696 also be used when creating a RAID1 or RAID10 if you want to avoid the
697 initial resync, however this practice \(em while normally safe \(em is not
698 recommended. Use this only if you really know what you are doing.
700 When the devices that will be part of a new array were filled
701 with zeros before creation the operator knows the array is
702 actually clean. If that is the case, such as after running
703 badblocks, this argument can be used to tell mdadm the
704 facts the operator knows.
706 When an array is resized to a larger size with
707 .B "\-\-grow \-\-size="
708 the new space is normally resynced in that same way that the whole
709 array is resynced at creation. From Linux version 2.6.40,
711 can be used with that command to avoid the automatic resync.
714 .BR \-\-backup\-file=
717 is used to increase the number of raid-devices in a RAID5 or RAID6 if
718 there are no spare devices available, or to shrink, change RAID level
719 or layout. See the GROW MODE section below on RAID\-DEVICES CHANGES.
720 The file must be stored on a separate device, not on the RAID array
724 .BR \-N ", " \-\-name=
727 for the array. This is currently only effective when creating an
728 array with a version-1 superblock, or an array in a DDF container.
729 The name is a simple textual string that can be used to identify array
730 components when assembling. If name is needed but not specified, it
731 is taken from the basename of the device that is being created.
743 run the array, even if some of the components
744 appear to be active in another array or filesystem. Normally
746 will ask for confirmation before including such components in an
747 array. This option causes that question to be suppressed.
750 .BR \-f ", " \-\-force
753 accept the geometry and layout specified without question. Normally
755 will not allow creation of an array with only one device, and will try
756 to create a RAID5 array with one missing drive (as this makes the
757 initial resync work faster). With
760 will not try to be so clever.
763 .BR \-a ", " "\-\-auto{=yes,md,mdp,part,p}{NN}"
764 Instruct mdadm how to create the device file if needed, possibly allocating
765 an unused minor number. "md" causes a non-partitionable array
766 to be used (though since Linux 2.6.28, these array devices are in fact
767 partitionable). "mdp", "part" or "p" causes a partitionable array (2.6 and
768 later) to be used. "yes" requires the named md device to have
769 a 'standard' format, and the type and minor number will be determined
770 from this. With mdadm 3.0, device creation is normally left up to
772 so this option is unlikely to be needed.
773 See DEVICE NAMES below.
775 The argument can also come immediately after
780 is not given on the command line or in the config file, then
786 is also given, then any
788 entries in the config file will override the
790 instruction given on the command line.
792 For partitionable arrays,
794 will create the device file for the whole array and for the first 4
795 partitions. A different number of partitions can be specified at the
796 end of this option (e.g.
798 If the device name ends with a digit, the partition names add a 'p',
800 .IR /dev/md/home1p3 .
801 If there is no trailing digit, then the partition names just have a
803 .IR /dev/md/scratch3 .
805 If the md device name is in a 'standard' format as described in DEVICE
806 NAMES, then it will be created, if necessary, with the appropriate
807 device number based on that name. If the device name is not in one of these
808 formats, then a unused device number will be allocated. The device
809 number will be considered unused if there is no active array for that
810 number, and there is no entry in /dev for that number and with a
811 non-standard name. Names that are not in 'standard' format are only
812 allowed in "/dev/md/".
816 .\".BR \-\-symlink = no
821 .\"to create devices in
823 .\"it will also create symlinks from
825 .\"with names starting with
831 .\"to suppress this, or
832 .\".B \-\-symlink=yes
833 .\"to enforce this even if it is suppressing
841 .BR \-u ", " \-\-uuid=
842 uuid of array to assemble. Devices which don't have this uuid are
846 .BR \-m ", " \-\-super\-minor=
847 Minor number of device that array was created for. Devices which
848 don't have this minor number are excluded. If you create an array as
849 /dev/md1, then all superblocks will contain the minor number 1, even if
850 the array is later assembled as /dev/md2.
852 Giving the literal word "dev" for
856 to use the minor number of the md device that is being assembled.
859 .B \-\-super\-minor=dev
860 will look for super blocks with a minor number of 0.
863 is only relevant for v0.90 metadata, and should not normally be used.
869 .BR \-N ", " \-\-name=
870 Specify the name of the array to assemble. This must be the name
871 that was specified when creating the array. It must either match
872 the name stored in the superblock exactly, or it must match
875 prefixed to the start of the given name.
878 .BR \-f ", " \-\-force
879 Assemble the array even if the metadata on some devices appears to be
882 cannot find enough working devices to start the array, but can find
883 some devices that are recorded as having failed, then it will mark
884 those devices as working so that the array can be started.
885 An array which requires
887 to be started may contain data corruption. Use it carefully.
891 Attempt to start the array even if fewer drives were given than were
892 present last time the array was active. Normally if not all the
893 expected drives are found and
895 is not used, then the array will be assembled but not started.
898 an attempt will be made to start it anyway.
902 This is the reverse of
904 in that it inhibits the startup of array unless all expected drives
905 are present. This is only needed with
907 and can be used if the physical connections to devices are
908 not as reliable as you would like.
911 .BR \-a ", " "\-\-auto{=no,yes,md,mdp,part}"
912 See this option under Create and Build options.
915 .BR \-a ", " "\-\-add"
916 This option can be used in Grow mode in two cases.
918 If the target array is a Linear array, then
920 can be used to add one or more devices to the array. They
921 are simply catenated on to the end of the array. Once added, the
922 devices cannot be removed.
926 option is being used to increase the number of devices in an array,
929 can be used to add some extra devices to be included in the array.
930 In most cases this is not needed as the extra devices can be added as
931 spares first, and then the number of raid-disks can be changed.
932 However for RAID0, it is not possible to add spares. So to increase
933 the number of devices in a RAID0, it is necessary to set the new
934 number of devices, and to add the new devices, in the same command.
937 .BR \-b ", " \-\-bitmap=
938 Specify the bitmap file that was given when the array was created. If
941 bitmap, there is no need to specify this when assembling the array.
944 .BR \-\-backup\-file=
947 was used while reshaping an array (e.g. changing number of devices or
948 chunk size) and the system crashed during the critical section, then the same
952 to allow possibly corrupted data to be restored, and the reshape
956 .BR \-\-invalid\-backup
957 If the file needed for the above option is not available for any
958 reason an empty file can be given together with this option to
959 indicate that the backup file is invalid. In this case the data that
960 was being rearranged at the time of the crash could be irrecoverably
961 lost, but the rest of the array may still be recoverable. This option
962 should only be used as a last resort if there is no way to recover the
967 .BR \-U ", " \-\-update=
968 Update the superblock on each device while assembling the array. The
969 argument given to this flag can be one of
984 option will adjust the superblock of an array what was created on a Sparc
985 machine running a patched 2.2 Linux kernel. This kernel got the
986 alignment of part of the superblock wrong. You can use the
987 .B "\-\-examine \-\-sparc2.2"
990 to see what effect this would have.
994 option will update the
996 field on each superblock to match the minor number of the array being
998 This can be useful if
1000 reports a different "Preferred Minor" to
1002 In some cases this update will be performed automatically
1003 by the kernel driver. In particular the update happens automatically
1004 at the first write to an array with redundancy (RAID level 1 or
1005 greater) on a 2.6 (or later) kernel.
1009 option will change the uuid of the array. If a UUID is given with the
1011 option that UUID will be used as a new UUID and will
1013 be used to help identify the devices in the array.
1016 is given, a random UUID is chosen.
1020 option will change the
1022 of the array as stored in the superblock. This is only supported for
1023 version-1 superblocks.
1027 option will change the
1029 as recorded in the superblock. For version-0 superblocks, this is the
1030 same as updating the UUID.
1031 For version-1 superblocks, this involves updating the name.
1035 option will cause the array to be marked
1037 meaning that any redundancy in the array (e.g. parity for RAID5,
1038 copies for RAID1) may be incorrect. This will cause the RAID system
1039 to perform a "resync" pass to make sure that all redundant information
1044 option allows arrays to be moved between machines with different
1046 When assembling such an array for the first time after a move, giving
1047 .B "\-\-update=byteorder"
1050 to expect superblocks to have their byteorder reversed, and will
1051 correct that order before assembling the array. This is only valid
1052 with original (Version 0.90) superblocks.
1056 option will correct the summaries in the superblock. That is the
1057 counts of total, working, active, failed, and spare devices.
1061 option will rarely be of use. It applies to version 1.1 and 1.2 metadata
1062 only (where the metadata is at the start of the device) and is only
1063 useful when the component device has changed size (typically become
1064 larger). The version 1 metadata records the amount of the device that
1065 can be used to store data, so if a device in a version 1.1 or 1.2
1066 array becomes larger, the metadata will still be visible, but the
1067 extra space will not. In this case it might be useful to assemble the
1069 .BR \-\-update=devicesize .
1072 to determine the maximum usable amount of space on each device and
1073 update the relevant field in the metadata.
1077 option can be used when an array has an internal bitmap which is
1078 corrupt in some way so that assembling the array normally fails. It
1079 will cause any internal bitmap to be ignored.
1081 .SH For Manage mode:
1084 .BR \-t ", " \-\-test
1085 Unless a more serious error occurred,
1087 will exit with a status of 2 if no changes were made to the array and
1088 0 if at least one change was made.
1089 This can be useful when an indirect specifier such as
1094 is used in requesting an operation on the array.
1096 will report failure if these specifiers didn't find any match.
1099 .BR \-a ", " \-\-add
1100 hot-add listed devices.
1101 If a device appears to have recently been part of the array
1102 (possibly it failed or was removed) the device is re\-added as describe
1104 If that fails or the device was never part of the array, the device is
1105 added as a hot-spare.
1106 If the array is degraded, it will immediately start to rebuild data
1109 Note that this and the following options are only meaningful on array
1110 with redundancy. They don't apply to RAID0 or Linear.
1114 re\-add a device that was previous removed from an array.
1115 If the metadata on the device reports that it is a member of the
1116 array, and the slot that it used is still vacant, then the device will
1117 be added back to the array in the same position. This will normally
1118 cause the data for that device to be recovered. However based on the
1119 event count on the device, the recovery may only require sections that
1120 are flagged a write-intent bitmap to be recovered or may not require
1121 any recovery at all.
1123 When used on an array that has no metadata (i.e. it was built with
1125 it will be assumed that bitmap-based recovery is enough to make the
1126 device fully consistent with the array.
1130 can be accompanied by
1131 .BR \-\-update=devicesize .
1132 See the description of this option when used in Assemble mode for an
1133 explanation of its use.
1135 If the device name given is
1137 then mdadm will try to find any device that looks like it should be
1138 part of the array but isn't and will try to re\-add all such devices.
1141 .BR \-r ", " \-\-remove
1142 remove listed devices. They must not be active. i.e. they should
1143 be failed or spare devices. As well as the name of a device file
1152 The first causes all failed device to be removed. The second causes
1153 any device which is no longer connected to the system (i.e an 'open'
1156 to be removed. This will only succeed for devices that are spares or
1157 have already been marked as failed.
1160 .BR \-f ", " \-\-fail
1161 mark listed devices as faulty.
1162 As well as the name of a device file, the word
1164 can be given. This will cause any device that has been detached from
1165 the system to be marked as failed. It can then be removed.
1173 .BR \-\-write\-mostly
1174 Subsequent devices that are added or re\-added will have the 'write-mostly'
1175 flag set. This is only valid for RAID1 and means that the 'md' driver
1176 will avoid reading from these devices if possible.
1179 Subsequent devices that are added or re\-added will have the 'write-mostly'
1183 Each of these options requires that the first device listed is the array
1184 to be acted upon, and the remainder are component devices to be added,
1185 removed, marked as faulty, etc. Several different operations can be
1186 specified for different devices, e.g.
1188 mdadm /dev/md0 \-\-add /dev/sda1 \-\-fail /dev/sdb1 \-\-remove /dev/sdb1
1190 Each operation applies to all devices listed until the next
1193 If an array is using a write-intent bitmap, then devices which have
1194 been removed can be re\-added in a way that avoids a full
1195 reconstruction but instead just updates the blocks that have changed
1196 since the device was removed. For arrays with persistent metadata
1197 (superblocks) this is done automatically. For arrays created with
1199 mdadm needs to be told that this device we removed recently with
1202 Devices can only be removed from an array if they are not in active
1203 use, i.e. that must be spares or failed devices. To remove an active
1204 device, it must first be marked as
1210 .BR \-Q ", " \-\-query
1211 Examine a device to see
1212 (1) if it is an md device and (2) if it is a component of an md
1214 Information about what is discovered is presented.
1217 .BR \-D ", " \-\-detail
1218 Print details of one or more md devices.
1221 .BR \-\-detail\-platform
1222 Print details of the platform's RAID capabilities (firmware / hardware
1223 topology) for a given metadata format.
1226 .BR \-Y ", " \-\-export
1231 output will be formatted as
1233 pairs for easy import into the environment.
1236 .BR \-E ", " \-\-examine
1237 Print contents of the metadata stored on the named device(s).
1238 Note the contrast between
1243 applies to devices which are components of an array, while
1245 applies to a whole array which is currently active.
1248 If an array was created on a SPARC machine with a 2.2 Linux kernel
1249 patched with RAID support, the superblock will have been created
1250 incorrectly, or at least incompatibly with 2.4 and later kernels.
1255 will fix the superblock before displaying it. If this appears to do
1256 the right thing, then the array can be successfully assembled using
1257 .BR "\-\-assemble \-\-update=sparc2.2" .
1260 .BR \-X ", " \-\-examine\-bitmap
1261 Report information about a bitmap file.
1262 The argument is either an external bitmap file or an array component
1263 in case of an internal bitmap. Note that running this on an array
1266 does not report the bitmap for that array.
1269 .BR \-R ", " \-\-run
1270 start a partially assembled array. If
1272 did not find enough devices to fully start the array, it might leaving
1273 it partially assembled. If you wish, you can then use
1275 to start the array in degraded mode.
1278 .BR \-S ", " \-\-stop
1279 deactivate array, releasing all resources.
1282 .BR \-o ", " \-\-readonly
1283 mark array as readonly.
1286 .BR \-w ", " \-\-readwrite
1287 mark array as readwrite.
1290 .B \-\-zero\-superblock
1291 If the device contains a valid md superblock, the block is
1292 overwritten with zeros. With
1294 the block where the superblock would be is overwritten even if it
1295 doesn't appear to be valid.
1298 .B \-\-kill\-subarray=
1299 If the device is a container and the argument to \-\-kill\-subarray
1300 specifies an inactive subarray in the container, then the subarray is
1301 deleted. Deleting all subarrays will leave an 'empty-container' or
1302 spare superblock on the drives. See \-\-zero\-superblock for completely
1303 removing a superblock. Note that some formats depend on the subarray
1304 index for generating a UUID, this command will fail if it would change
1305 the UUID of an active subarray.
1308 .B \-\-update\-subarray=
1309 If the device is a container and the argument to \-\-update\-subarray
1310 specifies a subarray in the container, then attempt to update the given
1311 superblock field in the subarray. See below in
1316 .BR \-t ", " \-\-test
1321 is set to reflect the status of the device. See below in
1326 .BR \-W ", " \-\-wait
1327 For each md device given, wait for any resync, recovery, or reshape
1328 activity to finish before returning.
1330 will return with success if it actually waited for every device
1331 listed, otherwise it will return failure.
1335 For each md device given, or each device in /proc/mdstat if
1337 is given, arrange for the array to be marked clean as soon as possible.
1339 will return with success if the array uses external metadata and we
1340 successfully waited. For native arrays this returns immediately as the
1341 kernel handles dirty-clean transitions at shutdown. No action is taken
1342 if safe-mode handling is disabled.
1344 .SH For Incremental Assembly mode:
1346 .BR \-\-rebuild\-map ", " \-r
1347 Rebuild the map file
1348 .RB ( /var/run/mdadm/map )
1351 uses to help track which arrays are currently being assembled.
1354 .BR \-\-run ", " \-R
1355 Run any array assembled as soon as a minimal number of devices are
1356 available, rather than waiting until all expected devices are present.
1359 .BR \-\-scan ", " \-s
1360 Only meaningful with
1364 file for arrays that are being incrementally assembled and will try to
1365 start any that are not already started. If any such array is listed
1368 as requiring an external bitmap, that bitmap will be attached first.
1371 .BR \-\-fail ", " \-f
1372 This allows the hot-plug system to remove devices that have fully disappeared
1373 from the kernel. It will first fail and then remove the device from any
1374 array it belongs to.
1375 The device name given should be a kernel device name such as "sda",
1381 Only used with \-\-fail. The 'path' given will be recorded so that if
1382 a new device appears at the same location it can be automatically
1383 added to the same array. This allows the failed device to be
1384 automatically replaced by a new device without metadata if it appears
1385 at specified path. This option is normally only set by a
1389 .SH For Monitor mode:
1391 .BR \-m ", " \-\-mail
1392 Give a mail address to send alerts to.
1395 .BR \-p ", " \-\-program ", " \-\-alert
1396 Give a program to be run whenever an event is detected.
1399 .BR \-y ", " \-\-syslog
1400 Cause all events to be reported through 'syslog'. The messages have
1401 facility of 'daemon' and varying priorities.
1404 .BR \-d ", " \-\-delay
1405 Give a delay in seconds.
1407 polls the md arrays and then waits this many seconds before polling
1408 again. The default is 60 seconds. Since 2.6.16, there is no need to
1409 reduce this as the kernel alerts
1411 immediately when there is any change.
1414 .BR \-r ", " \-\-increment
1415 Give a percentage increment.
1417 will generate RebuildNN events with the given percentage increment.
1420 .BR \-f ", " \-\-daemonise
1423 to run as a background daemon if it decides to monitor anything. This
1424 causes it to fork and run in the child, and to disconnect from the
1425 terminal. The process id of the child is written to stdout.
1428 which will only continue monitoring if a mail address or alert program
1429 is found in the config file.
1432 .BR \-i ", " \-\-pid\-file
1435 is running in daemon mode, write the pid of the daemon process to
1436 the specified file, instead of printing it on standard output.
1439 .BR \-1 ", " \-\-oneshot
1440 Check arrays only once. This will generate
1442 events and more significantly
1448 .B " mdadm \-\-monitor \-\-scan \-1"
1450 from a cron script will ensure regular notification of any degraded arrays.
1453 .BR \-t ", " \-\-test
1456 alert for every array found at startup. This alert gets mailed and
1457 passed to the alert program. This can be used for testing that alert
1458 message do get through successfully.
1462 This inhibits the functionality for moving spares between arrays.
1463 Only one monitoring process started with
1465 but without this flag is allowed, otherwise the two could interfere
1472 .B mdadm \-\-assemble
1473 .I md-device options-and-component-devices...
1476 .B mdadm \-\-assemble \-\-scan
1477 .I md-devices-and-options...
1480 .B mdadm \-\-assemble \-\-scan
1484 This usage assembles one or more RAID arrays from pre-existing components.
1485 For each array, mdadm needs to know the md device, the identity of the
1486 array, and a number of component-devices. These can be found in a number of ways.
1488 In the first usage example (without the
1490 the first device given is the md device.
1491 In the second usage example, all devices listed are treated as md
1492 devices and assembly is attempted.
1493 In the third (where no devices are listed) all md devices that are
1494 listed in the configuration file are assembled. If not arrays are
1495 described by the configuration file, then any arrays that
1496 can be found on unused devices will be assembled.
1498 If precisely one device is listed, but
1504 was given and identity information is extracted from the configuration file.
1506 The identity can be given with the
1512 option, will be taken from the md-device record in the config file, or
1513 will be taken from the super block of the first component-device
1514 listed on the command line.
1516 Devices can be given on the
1518 command line or in the config file. Only devices which have an md
1519 superblock which contains the right identity will be considered for
1522 The config file is only used if explicitly named with
1524 or requested with (a possibly implicit)
1529 .B /etc/mdadm/mdadm.conf
1534 is not given, then the config file will only be used to find the
1535 identity of md arrays.
1537 Normally the array will be started after it is assembled. However if
1539 is not given and not all expected drives were listed, then the array
1540 is not started (to guard against usage errors). To insist that the
1541 array be started in this case (as may work for RAID1, 4, 5, 6, or 10),
1550 does not create any entries in
1554 It does record information in
1555 .B /var/run/mdadm/map
1558 to choose the correct name.
1562 detects that udev is not configured, it will create the devices in
1566 In Linux kernels prior to version 2.6.28 there were two distinctly
1567 different types of md devices that could be created: one that could be
1568 partitioned using standard partitioning tools and one that could not.
1569 Since 2.6.28 that distinction is no longer relevant as both type of
1570 devices can be partitioned.
1572 will normally create the type that originally could not be partitioned
1573 as it has a well defined major number (9).
1575 Prior to 2.6.28, it is important that mdadm chooses the correct type
1576 of array device to use. This can be controlled with the
1578 option. In particular, a value of "mdp" or "part" or "p" tells mdadm
1579 to use a partitionable device rather than the default.
1581 In the no-udev case, the value given to
1583 can be suffixed by a number. This tells
1585 to create that number of partition devices rather than the default of 4.
1589 can also be given in the configuration file as a word starting
1591 on the ARRAY line for the relevant array.
1598 and no devices are listed,
1600 will first attempt to assemble all the arrays listed in the config
1603 In no array at listed in the config (other than those marked
1605 it will look through the available devices for possible arrays and
1606 will try to assemble anything that it finds. Arrays which are tagged
1607 as belonging to the given homehost will be assembled and started
1608 normally. Arrays which do not obviously belong to this host are given
1609 names that are expected not to conflict with anything local, and are
1610 started "read-auto" so that nothing is written to any device until the
1611 array is written to. i.e. automatic resync etc is delayed.
1615 finds a consistent set of devices that look like they should comprise
1616 an array, and if the superblock is tagged as belonging to the given
1617 home host, it will automatically choose a device name and try to
1618 assemble the array. If the array uses version-0.90 metadata, then the
1620 number as recorded in the superblock is used to create a name in
1624 If the array uses version-1 metadata, then the
1626 from the superblock is used to similarly create a name in
1628 (the name will have any 'host' prefix stripped first).
1630 This behaviour can be modified by the
1634 configuration file. This line can indicate that specific metadata
1635 type should, or should not, be automatically assembled. If an array
1636 is found which is not listed in
1638 and has a metadata format that is denied by the
1640 line, then it will not be assembled.
1643 line can also request that all arrays identified as being for this
1644 homehost should be assembled regardless of their metadata type.
1647 for further details.
1649 Note: Auto assembly cannot be used for assembling and activating some
1650 arrays which are undergoing reshape. In particular as the
1652 cannot be given, any reshape which requires a backup-file to continue
1653 cannot be started by auto assembly. An array which is growing to more
1654 devices and has passed the critical section can be assembled using
1665 .BI \-\-raid\-devices= Z
1669 This usage is similar to
1671 The difference is that it creates an array without a superblock. With
1672 these arrays there is no difference between initially creating the array and
1673 subsequently assembling the array, except that hopefully there is useful
1674 data there in the second case.
1676 The level may raid0, linear, raid1, raid10, multipath, or faulty, or
1677 one of their synonyms. All devices must be listed and the array will
1678 be started once complete. It will often be appropriate to use
1679 .B \-\-assume\-clean
1680 with levels raid1 or raid10.
1691 .BI \-\-raid\-devices= Z
1695 This usage will initialise a new md array, associate some devices with
1696 it, and activate the array.
1698 The named device will normally not exist when
1699 .I "mdadm \-\-create"
1700 is run, but will be created by
1702 once the array becomes active.
1704 As devices are added, they are checked to see if they contain RAID
1705 superblocks or filesystems. They are also checked to see if the variance in
1706 device size exceeds 1%.
1708 If any discrepancy is found, the array will not automatically be run, though
1711 can override this caution.
1713 To create a "degraded" array in which some devices are missing, simply
1714 give the word "\fBmissing\fP"
1715 in place of a device name. This will cause
1717 to leave the corresponding slot in the array empty.
1718 For a RAID4 or RAID5 array at most one slot can be
1719 "\fBmissing\fP"; for a RAID6 array at most two slots.
1720 For a RAID1 array, only one real device needs to be given. All of the
1724 When creating a RAID5 array,
1726 will automatically create a degraded array with an extra spare drive.
1727 This is because building the spare into a degraded array is in general
1728 faster than resyncing the parity on a non-degraded, but not clean,
1729 array. This feature can be overridden with the
1733 When creating an array with version-1 metadata a name for the array is
1735 If this is not given with the
1739 will choose a name based on the last component of the name of the
1740 device being created. So if
1742 is being created, then the name
1747 is being created, then the name
1751 When creating a partition based array, using
1753 with version-1.x metadata, the partition type should be set to
1755 (non fs-data). This type selection allows for greater precision since
1756 using any other [RAID auto-detect (0xFD) or a GNU/Linux partition (0x83)],
1757 might create problems in the event of array recovery through a live cdrom.
1759 A new array will normally get a randomly assigned 128bit UUID which is
1760 very likely to be unique. If you have a specific need, you can choose
1761 a UUID for the array by giving the
1763 option. Be warned that creating two arrays with the same UUID is a
1764 recipe for disaster. Also, using
1766 when creating a v0.90 array will silently override any
1771 .\"option is given, it is not necessary to list any component-devices in this command.
1772 .\"They can be added later, before a
1776 .\"is given, the apparent size of the smallest drive given is used.
1778 When creating an array within a
1781 can be given either the list of devices to use, or simply the name of
1782 the container. The former case gives control over which devices in
1783 the container will be used for the array. The latter case allows
1785 to automatically choose which devices to use based on how much spare
1788 The General Management options that are valid with
1793 insist on running the array even if some devices look like they might
1798 start the array readonly \(em not supported yet.
1805 .I options... devices...
1808 This usage will allow individual devices in an array to be failed,
1809 removed or added. It is possible to perform multiple operations with
1810 on command. For example:
1812 .B " mdadm /dev/md0 \-f /dev/hda1 \-r /dev/hda1 \-a /dev/hda1"
1818 and will then remove it from the array and finally add it back
1819 in as a spare. However only one md array can be affected by a single
1822 When a device is added to an active array, mdadm checks to see if it
1823 has metadata on it which suggests that it was recently a member of the
1824 array. If it does, it tries to "re\-add" the device. If there have
1825 been no changes since the device was removed, or if the array has a
1826 write-intent bitmap which has recorded whatever changes there were,
1827 then the device will immediately become a full member of the array and
1828 those differences recorded in the bitmap will be resolved.
1838 MISC mode includes a number of distinct operations that
1839 operate on distinct devices. The operations are:
1842 The device is examined to see if it is
1843 (1) an active md array, or
1844 (2) a component of an md array.
1845 The information discovered is reported.
1849 The device should be an active md device.
1851 will display a detailed description of the array.
1855 will cause the output to be less detailed and the format to be
1856 suitable for inclusion in
1860 will normally be 0 unless
1862 failed to get useful information about the device(s); however, if the
1864 option is given, then the exit status will be:
1868 The array is functioning normally.
1871 The array has at least one failed device.
1874 The array has multiple failed devices such that it is unusable.
1877 There was an error while trying to get information about the device.
1881 .B \-\-detail\-platform
1882 Print detail of the platform's RAID capabilities (firmware / hardware
1883 topology). If the metadata is specified with
1887 then the return status will be:
1891 metadata successfully enumerated its platform components on this system
1894 metadata is platform independent
1897 metadata failed to find its platform components on this system
1901 .B \-\-update\-subarray=
1902 If the device is a container and the argument to \-\-update\-subarray
1903 specifies a subarray in the container, then attempt to update the given
1904 superblock field in the subarray. Similar to updating an array in
1905 "assemble" mode, the field to update is selected by
1909 option. Currently only
1915 option updates the subarray name in the metadata, it may not affect the
1916 device node name or the device node symlink until the subarray is
1917 re\-assembled. If updating
1919 would change the UUID of an active subarray this operation is blocked,
1920 and the command will end in an error.
1924 The device should be a component of an md array.
1926 will read the md superblock of the device and display the contents.
1931 is given, then multiple devices that are components of the one array
1932 are grouped together and reported in a single entry suitable
1938 without listing any devices will cause all devices listed in the
1939 config file to be examined.
1943 The devices should be active md arrays which will be deactivated, as
1944 long as they are not currently in use.
1948 This will fully activate a partially assembled md array.
1952 This will mark an active array as read-only, providing that it is
1953 not currently being used.
1959 array back to being read/write.
1963 For all operations except
1966 will cause the operation to be applied to all arrays listed in
1971 causes all devices listed in the config file to be examined.
1974 .BR \-b ", " \-\-brief
1975 Be less verbose. This is used with
1983 gives an intermediate level of verbosity.
1989 .B mdadm \-\-monitor
1990 .I options... devices...
1995 to periodically poll a number of md arrays and to report on any events
1998 will never exit once it decides that there are arrays to be checked,
1999 so it should normally be run in the background.
2001 As well as reporting events,
2003 may move a spare drive from one array to another if they are in the
2008 and if the destination array has a failed drive but no spares.
2010 If any devices are listed on the command line,
2012 will only monitor those devices. Otherwise all arrays listed in the
2013 configuration file will be monitored. Further, if
2015 is given, then any other md devices that appear in
2017 will also be monitored.
2019 The result of monitoring the arrays is the generation of events.
2020 These events are passed to a separate program (if specified) and may
2021 be mailed to a given E-mail address.
2023 When passing events to a program, the program is run once for each event,
2024 and is given 2 or 3 command-line arguments: the first is the
2025 name of the event (see below), the second is the name of the
2026 md device which is affected, and the third is the name of a related
2027 device if relevant (such as a component device that has failed).
2031 is given, then a program or an E-mail address must be specified on the
2032 command line or in the config file. If neither are available, then
2034 will not monitor anything.
2038 will continue monitoring as long as something was found to monitor. If
2039 no program or email is given, then each event is reported to
2042 The different events are:
2046 .B DeviceDisappeared
2047 An md array which previously was configured appears to no longer be
2048 configured. (syslog priority: Critical)
2052 was told to monitor an array which is RAID0 or Linear, then it will
2054 .B DeviceDisappeared
2055 with the extra information
2057 This is because RAID0 and Linear do not support the device-failed,
2058 hot-spare and resync operations which are monitored.
2062 An md array started reconstruction. (syslog priority: Warning)
2068 is a two-digit number (ie. 05, 48). This indicates that rebuild
2069 has passed that many percent of the total. The events are generated
2070 with fixed increment since 0. Increment size may be specified with
2071 a commandline option (default is 20). (syslog priority: Warning)
2075 An md array that was rebuilding, isn't any more, either because it
2076 finished normally or was aborted. (syslog priority: Warning)
2080 An active component device of an array has been marked as
2081 faulty. (syslog priority: Critical)
2085 A spare component device which was being rebuilt to replace a faulty
2086 device has failed. (syslog priority: Critical)
2090 A spare component device which was being rebuilt to replace a faulty
2091 device has been successfully rebuilt and has been made active.
2092 (syslog priority: Info)
2096 A new md array has been detected in the
2098 file. (syslog priority: Info)
2102 A newly noticed array appears to be degraded. This message is not
2105 notices a drive failure which causes degradation, but only when
2107 notices that an array is degraded when it first sees the array.
2108 (syslog priority: Critical)
2112 A spare drive has been moved from one array in a
2116 to another to allow a failed drive to be replaced.
2117 (syslog priority: Info)
2123 has been told, via the config file, that an array should have a certain
2124 number of spare devices, and
2126 detects that it has fewer than this number when it first sees the
2127 array, it will report a
2130 (syslog priority: Warning)
2134 An array was found at startup, and the
2137 (syslog priority: Info)
2147 cause Email to be sent. All events cause the program to be run.
2148 The program is run with two or three arguments: the event
2149 name, the array device and possibly a second device.
2151 Each event has an associated array device (e.g.
2153 and possibly a second device. For
2158 the second device is the relevant component device.
2161 the second device is the array that the spare was moved from.
2165 to move spares from one array to another, the different arrays need to
2166 be labeled with the same
2168 or the spares must be allowed to migrate through matching POLICY domains
2169 in the configuration file. The
2171 name can be any string; it is only necessary that different spare
2172 groups use different names.
2176 detects that an array in a spare group has fewer active
2177 devices than necessary for the complete array, and has no spare
2178 devices, it will look for another array in the same spare group that
2179 has a full complement of working drive and a spare. It will then
2180 attempt to remove the spare from the second drive and add it to the
2182 If the removal succeeds but the adding fails, then it is added back to
2185 If the spare group for a degraded array is not defined,
2187 will look at the rules of spare migration specified by POLICY lines in
2189 and then follow similar steps as above if a matching spare is found.
2192 The GROW mode is used for changing the size or shape of an active
2194 For this to work, the kernel must support the necessary change.
2195 Various types of growth are being added during 2.6 development.
2197 Currently the supported changes include
2199 change the "size" attribute for RAID1, RAID4, RAID5 and RAID6.
2201 increase or decrease the "raid\-devices" attribute of RAID0, RAID1, RAID4,
2204 change the chunk-size and layout of RAID0, RAID4, RAID5 and RAID6.
2206 convert between RAID1 and RAID5, between RAID5 and RAID6, between
2207 RAID0, RAID5, and RAID5, and between RAID0 and RAID10 (in the near-2 mode).
2209 add a write-intent bitmap to any array which supports these bitmaps, or
2210 remove a write-intent bitmap from such an array.
2213 Using GROW on containers is currently only support for Intel's IMSM
2214 container format. The number of devices in a container can be
2215 increased - which affects all arrays in the container - or an array
2216 in a container can be converted between levels where those levels are
2217 supported by the container, and the conversion is on of those listed
2220 Grow functionality (e.g. expand a number of raid devices) for Intel's
2221 IMSM container format has an experimental status. It is guarded by the
2222 .B MDADM_EXPERIMENTAL
2223 environment variable which must be set to '1' for a GROW command to
2225 This is for the following reasons:
2228 Intel's native IMSM check-pointing is not fully tested yet.
2229 This can causes IMSM incompatibility during the grow process: an array
2230 which is growing cannot roam between Microsoft Windows(R) and Linux
2234 Interrupting a grow operation is not recommended, because it
2235 has not been fully tested for Intel's IMSM container format yet.
2238 Note: Intel's native checkpointing doesn't use
2240 option and it is transparent for assembly feature.
2243 Normally when an array is built the "size" is taken from the smallest
2244 of the drives. If all the small drives in an arrays are, one at a
2245 time, removed and replaced with larger drives, then you could have an
2246 array of large drives with only a small amount used. In this
2247 situation, changing the "size" with "GROW" mode will allow the extra
2248 space to start being used. If the size is increased in this way, a
2249 "resync" process will start to make sure the new parts of the array
2252 Note that when an array changes size, any filesystem that may be
2253 stored in the array will not automatically grow for shrink to use or
2254 vacate the space. The
2255 filesystem will need to be explicitly told to use the extra space
2256 after growing, or to reduce its size
2258 to shrinking the array.
2260 Also the size of an array cannot be changed while it has an active
2261 bitmap. If an array has a bitmap, it must be removed before the size
2262 can be changed. Once the change it complete a new bitmap can be created.
2264 .SS RAID\-DEVICES CHANGES
2266 A RAID1 array can work with any number of devices from 1 upwards
2267 (though 1 is not very useful). There may be times which you want to
2268 increase or decrease the number of active devices. Note that this is
2269 different to hot-add or hot-remove which changes the number of
2272 When reducing the number of devices in a RAID1 array, the slots which
2273 are to be removed from the array must already be vacant. That is, the
2274 devices which were in those slots must be failed and removed.
2276 When the number of devices is increased, any hot spares that are
2277 present will be activated immediately.
2279 Changing the number of active devices in a RAID5 or RAID6 is much more
2280 effort. Every block in the array will need to be read and written
2281 back to a new location. From 2.6.17, the Linux Kernel is able to
2282 increase the number of devices in a RAID5 safely, including restarting
2283 an interrupted "reshape". From 2.6.31, the Linux Kernel is able to
2284 increase or decrease the number of devices in a RAID5 or RAID6.
2286 From 2.6.35, the Linux Kernel is able to convert a RAID0 in to a RAID4
2289 uses this functionality and the ability to add
2290 devices to a RAID4 to allow devices to be added to a RAID0. When
2291 requested to do this,
2293 will convert the RAID0 to a RAID4, add the necessary disks and make
2294 the reshape happen, and then convert the RAID4 back to RAID0.
2296 When decreasing the number of devices, the size of the array will also
2297 decrease. If there was data in the array, it could get destroyed and
2298 this is not reversible, so you should firstly shrink the filesystem on
2299 the array to fit within the new size. To help prevent accidents,
2301 requires that the size of the array be decreased first with
2302 .BR "mdadm --grow --array-size" .
2303 This is a reversible change which simply makes the end of the array
2304 inaccessible. The integrity of any data can then be checked before
2305 the non-reversible reduction in the number of devices is request.
2307 When relocating the first few stripes on a RAID5 or RAID6, it is not
2308 possible to keep the data on disk completely consistent and
2309 crash-proof. To provide the required safety, mdadm disables writes to
2310 the array while this "critical section" is reshaped, and takes a
2311 backup of the data that is in that section. For grows, this backup may be
2312 stored in any spare devices that the array has, however it can also be
2313 stored in a separate file specified with the
2315 option, and is required to be specified for shrinks, RAID level
2316 changes and layout changes. If this option is used, and the system
2317 does crash during the critical period, the same file must be passed to
2319 to restore the backup and reassemble the array. When shrinking rather
2320 than growing the array, the reshape is done from the end towards the
2321 beginning, so the "critical section" is at the end of the reshape.
2325 Changing the RAID level of any array happens instantaneously. However
2326 in the RAID5 to RAID6 case this requires a non-standard layout of the
2327 RAID6 data, and in the RAID6 to RAID5 case that non-standard layout is
2328 required before the change can be accomplished. So while the level
2329 change is instant, the accompanying layout change can take quite a
2332 is required. If the array is not simultaneously being grown or
2333 shrunk, so that the array size will remain the same - for example,
2334 reshaping a 3-drive RAID5 into a 4-drive RAID6 - the backup file will
2335 be used not just for a "cricital section" but throughout the reshape
2336 operation, as described below under LAYOUT CHANGES.
2338 .SS CHUNK-SIZE AND LAYOUT CHANGES
2340 Changing the chunk-size of layout without also changing the number of
2341 devices as the same time will involve re-writing all blocks in-place.
2342 To ensure against data loss in the case of a crash, a
2344 must be provided for these changes. Small sections of the array will
2345 be copied to the backup file while they are being rearranged. This
2346 means that all the data is copied twice, once to the backup and once
2347 to the new layout on the array, so this type of reshape will go very
2350 If the reshape is interrupted for any reason, this backup file must be
2352 .B "mdadm --assemble"
2353 so the array can be reassembled. Consequently the file cannot be
2354 stored on the device being reshaped.
2359 A write-intent bitmap can be added to, or removed from, an active
2360 array. Either internal bitmaps, or bitmaps stored in a separate file,
2361 can be added. Note that if you add a bitmap stored in a file which is
2362 in a filesystem that is on the RAID array being affected, the system
2363 will deadlock. The bitmap must be on a separate filesystem.
2365 .SH INCREMENTAL MODE
2369 .B mdadm \-\-incremental
2375 .B mdadm \-\-incremental \-\-fail
2379 .B mdadm \-\-incremental \-\-rebuild\-map
2382 .B mdadm \-\-incremental \-\-run \-\-scan
2385 This mode is designed to be used in conjunction with a device
2386 discovery system. As devices are found in a system, they can be
2388 .B "mdadm \-\-incremental"
2389 to be conditionally added to an appropriate array.
2391 Conversely, it can also be used with the
2393 flag to do just the opposite and find whatever array a particular device
2394 is part of and remove the device from that array.
2396 If the device passed is a
2398 device created by a previous call to
2400 then rather than trying to add that device to an array, all the arrays
2401 described by the metadata of the container will be started.
2404 performs a number of tests to determine if the device is part of an
2405 array, and which array it should be part of. If an appropriate array
2406 is found, or can be created,
2408 adds the device to the array and conditionally starts the array.
2412 will normally only add devices to an array which were previously working
2413 (active or spare) parts of that array. The support for automatic
2414 inclusion of a new drive as a spare in some array requires
2415 a configuration through POLICY in config file.
2419 makes are as follow:
2421 Is the device permitted by
2423 That is, is it listed in a
2425 line in that file. If
2427 is absent then the default it to allow any device. Similar if
2429 contains the special word
2431 then any device is allowed. Otherwise the device name given to
2433 must match one of the names or patterns in a
2438 Does the device have a valid md superblock. If a specific metadata
2439 version is request with
2443 then only that style of metadata is accepted, otherwise
2445 finds any known version of metadata. If no
2447 metadata is found, the device may be still added to an array
2448 as a spare if POLICY allows.
2452 Does the metadata match an expected array?
2453 The metadata can match in two ways. Either there is an array listed
2456 which identifies the array (either by UUID, by name, by device list,
2457 or by minor-number), or the array was created with a
2463 or on the command line.
2466 is not able to positively identify the array as belonging to the
2467 current host, the device will be rejected.
2471 keeps a list of arrays that it has partially assembled in
2472 .B /var/run/mdadm/map
2474 .B /var/run/mdadm.map
2475 if the directory doesn't exist. Or maybe even
2476 .BR /dev/.mdadm.map ).
2477 If no array exists which matches
2478 the metadata on the new device,
2480 must choose a device name and unit number. It does this based on any
2483 or any name information stored in the metadata. If this name
2484 suggests a unit number, that number will be used, otherwise a free
2485 unit number will be chosen. Normally
2487 will prefer to create a partitionable array, however if the
2491 suggests that a non-partitionable array is preferred, that will be
2494 If the array is not found in the config file and its metadata does not
2495 identify it as belonging to the "homehost", then
2497 will choose a name for the array which is certain not to conflict with
2498 any array which does belong to this host. It does this be adding an
2499 underscore and a small number to the name preferred by the metadata.
2501 Once an appropriate array is found or created and the device is added,
2503 must decide if the array is ready to be started. It will
2504 normally compare the number of available (non-spare) devices to the
2505 number of devices that the metadata suggests need to be active. If
2506 there are at least that many, the array will be started. This means
2507 that if any devices are missing the array will not be restarted.
2513 in which case the array will be run as soon as there are enough
2514 devices present for the data to be accessible. For a RAID1, that
2515 means one device will start the array. For a clean RAID5, the array
2516 will be started as soon as all but one drive is present.
2518 Note that neither of these approaches is really ideal. If it can
2519 be known that all device discovery has completed, then
2523 can be run which will try to start all arrays that are being
2524 incrementally assembled. They are started in "read-auto" mode in
2525 which they are read-only until the first write request. This means
2526 that no metadata updates are made and no attempt at resync or recovery
2527 happens. Further devices that are found before the first write can
2528 still be added safely.
2531 This section describes environment variables that affect how mdadm
2536 Setting this value to 1 will prevent mdadm from automatically launching
2537 mdmon. This variable is intended primarily for debugging mdadm/mdmon.
2543 does not create any device nodes in /dev, but leaves that task to
2547 appears not to be configured, or if this environment variable is set
2550 will create and devices that are needed.
2554 .B " mdadm \-\-query /dev/name-of-device"
2556 This will find out if a given device is a RAID array, or is part of
2557 one, and will provide brief information about the device.
2559 .B " mdadm \-\-assemble \-\-scan"
2561 This will assemble and start all arrays listed in the standard config
2562 file. This command will typically go in a system startup file.
2564 .B " mdadm \-\-stop \-\-scan"
2566 This will shut down all arrays that can be shut down (i.e. are not
2567 currently in use). This will typically go in a system shutdown script.
2569 .B " mdadm \-\-follow \-\-scan \-\-delay=120"
2571 If (and only if) there is an Email address or program given in the
2572 standard config file, then
2573 monitor the status of all arrays listed in that file by
2574 polling them ever 2 minutes.
2576 .B " mdadm \-\-create /dev/md0 \-\-level=1 \-\-raid\-devices=2 /dev/hd[ac]1"
2578 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
2581 .B " echo 'DEVICE /dev/hd*[0\-9] /dev/sd*[0\-9]' > mdadm.conf"
2583 .B " mdadm \-\-detail \-\-scan >> mdadm.conf"
2585 This will create a prototype config file that describes currently
2586 active arrays that are known to be made from partitions of IDE or SCSI drives.
2587 This file should be reviewed before being used as it may
2588 contain unwanted detail.
2590 .B " echo 'DEVICE /dev/hd[a\-z] /dev/sd*[a\-z]' > mdadm.conf"
2592 .B " mdadm \-\-examine \-\-scan \-\-config=mdadm.conf >> mdadm.conf"
2594 This will find arrays which could be assembled from existing IDE and
2595 SCSI whole drives (not partitions), and store the information in the
2596 format of a config file.
2597 This file is very likely to contain unwanted detail, particularly
2600 entries. It should be reviewed and edited before being used as an
2603 .B " mdadm \-\-examine \-\-brief \-\-scan \-\-config=partitions"
2605 .B " mdadm \-Ebsc partitions"
2607 Create a list of devices by reading
2608 .BR /proc/partitions ,
2609 scan these for RAID superblocks, and printout a brief listing of all
2612 .B " mdadm \-Ac partitions \-m 0 /dev/md0"
2614 Scan all partitions and devices listed in
2615 .BR /proc/partitions
2618 out of all such devices with a RAID superblock with a minor number of 0.
2620 .B " mdadm \-\-monitor \-\-scan \-\-daemonise > /var/run/mdadm"
2622 If config file contains a mail address or alert program, run mdadm in
2623 the background in monitor mode monitoring all md devices. Also write
2624 pid of mdadm daemon to
2625 .BR /var/run/mdadm .
2627 .B " mdadm \-Iq /dev/somedevice"
2629 Try to incorporate newly discovered device into some array as
2632 .B " mdadm \-\-incremental \-\-rebuild\-map \-\-run \-\-scan"
2634 Rebuild the array map from any current arrays, and then start any that
2637 .B " mdadm /dev/md4 --fail detached --remove detached"
2639 Any devices which are components of /dev/md4 will be marked as faulty
2640 and then remove from the array.
2642 .B " mdadm --grow /dev/md4 --level=6 --backup-file=/root/backup-md4
2646 which is currently a RAID5 array will be converted to RAID6. There
2647 should normally already be a spare drive attached to the array as a
2648 RAID6 needs one more drive than a matching RAID5.
2650 .B " mdadm --create /dev/md/ddf --metadata=ddf --raid-disks 6 /dev/sd[a-f]"
2652 Create a DDF array over 6 devices.
2654 .B " mdadm --create /dev/md/home -n3 -l5 -z 30000000 /dev/md/ddf"
2656 Create a RAID5 array over any 3 devices in the given DDF set. Use
2657 only 30 gigabytes of each device.
2659 .B " mdadm -A /dev/md/ddf1 /dev/sd[a-f]"
2661 Assemble a pre-exist ddf array.
2663 .B " mdadm -I /dev/md/ddf1"
2665 Assemble all arrays contained in the ddf array, assigning names as
2668 .B " mdadm \-\-create \-\-help"
2670 Provide help about the Create mode.
2672 .B " mdadm \-\-config \-\-help"
2674 Provide help about the format of the config file.
2676 .B " mdadm \-\-help"
2678 Provide general help.
2688 lists all active md devices with information about them.
2690 uses this to find arrays when
2692 is given in Misc mode, and to monitor array reconstruction
2697 The config file lists which devices may be scanned to see if
2698 they contain MD super block, and gives identifying information
2699 (e.g. UUID) about known MD arrays. See
2703 .SS /var/run/mdadm/map
2706 mode is used, this file gets a list of arrays currently being created.
2709 does not exist as a directory, then
2710 .B /var/run/mdadm.map
2713 is not available (as may be the case during early boot),
2715 is used on the basis that
2717 is usually available very early in boot.
2722 understand two sorts of names for array devices.
2724 The first is the so-called 'standard' format name, which matches the
2725 names used by the kernel and which appear in
2728 The second sort can be freely chosen, but must reside in
2730 When giving a device name to
2732 to create or assemble an array, either full path name such as
2736 can be given, or just the suffix of the second sort of name, such as
2742 chooses device names during auto-assembly or incremental assembly, it
2743 will sometimes add a small sequence number to the end of the name to
2744 avoid conflicted between multiple arrays that have the same name. If
2746 can reasonably determine that the array really is meant for this host,
2747 either by a hostname in the metadata, or by the presence of the array
2750 then it will leave off the suffix if possible.
2751 Also if the homehost is specified as
2754 will only use a suffix if a different array of the same name already
2755 exists or is listed in the config file.
2757 The standard names for non-partitioned arrays (the only sort of md
2758 array available in 2.4 and earlier) are of the form
2762 where NN is a number.
2763 The standard names for partitionable arrays (as available from 2.6
2764 onwards) are of the form
2768 Partition numbers should be indicated by added "pMM" to these, thus "/dev/md/d1p2".
2770 From kernel version, 2.6.28 the "non-partitioned array" can actually
2771 be partitioned. So the "md_dNN" names are no longer needed, and
2772 partitions such as "/dev/mdNNpXX" are possible.
2776 was previously known as
2780 is completely separate from the
2782 package, and does not use the
2784 configuration file at all.
2787 For further information on mdadm usage, MD and the various levels of
2790 .B http://linux\-raid.osdl.org/
2792 (based upon Jakob \(/Ostergaard's Software\-RAID.HOWTO)
2794 .\"for new releases of the RAID driver check out:
2797 .\".UR ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2798 .\"ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2803 .\".UR http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2804 .\"http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2807 The latest version of
2809 should always be available from
2811 .B http://www.kernel.org/pub/linux/utils/raid/mdadm/