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.
259 Set first character of argv[0] to @ to indicate mdadm was launched
260 from initrd/initramfs and should not be shutdown by systemd as part of
261 the regular shutdown process. This option is normally only used by
262 the system's initscripts. Please see here for more details on how
263 systemd handled argv[0]:
265 .B http://www.freedesktop.org/wiki/Software/systemd/RootStorageDaemons
270 .BR \-f ", " \-\-force
271 Be more forceful about certain operations. See the various modes for
272 the exact meaning of this option in different contexts.
275 .BR \-c ", " \-\-config=
276 Specify the config file. Default is to use
277 .BR /etc/mdadm.conf ,
278 or if that is missing then
279 .BR /etc/mdadm/mdadm.conf .
280 If the config file given is
282 then nothing will be read, but
284 will act as though the config file contained exactly
285 .B "DEVICE partitions containers"
288 to find a list of devices to scan, and
290 to find a list of containers to examine.
293 is given for the config file, then
295 will act as though the config file were empty.
298 .BR \-s ", " \-\-scan
301 for missing information.
302 In general, this option gives
304 permission to get any missing information (like component devices,
305 array devices, array identities, and alert destination) from the
306 configuration file (see previous option);
307 one exception is MISC mode when using
313 says to get a list of array devices from
317 .BR \-e ", " \-\-metadata=
318 Declare the style of RAID metadata (superblock) to be used. The
319 default is {DEFAULT_METADATA} for
321 and to guess for other operations.
322 The default can be overridden by setting the
331 .ie '{DEFAULT_METADATA}'0.90'
332 .IP "0, 0.90, default"
336 Use the original 0.90 format superblock. This format limits arrays to
337 28 component devices and limits component devices of levels 1 and
338 greater to 2 terabytes. It is also possible for there to be confusion
339 about whether the superblock applies to a whole device or just the
340 last partition, if that partition starts on a 64K boundary.
341 .ie '{DEFAULT_METADATA}'0.90'
342 .IP "1, 1.0, 1.1, 1.2"
344 .IP "1, 1.0, 1.1, 1.2 default"
346 Use the new version-1 format superblock. This has fewer restrictions.
347 It can easily be moved between hosts with different endian-ness, and a
348 recovery operation can be checkpointed and restarted. The different
349 sub-versions store the superblock at different locations on the
350 device, either at the end (for 1.0), at the start (for 1.1) or 4K from
351 the start (for 1.2). "1" is equivalent to "1.2" (the commonly
352 preferred 1.x format).
353 'if '{DEFAULT_METADATA}'1.2' "default" is equivalent to "1.2".
355 Use the "Industry Standard" DDF (Disk Data Format) format defined by
357 When creating a DDF array a
359 will be created, and normal arrays can be created in that container.
361 Use the Intel(R) Matrix Storage Manager metadata format. This creates a
363 which is managed in a similar manner to DDF, and is supported by an
364 option-rom on some platforms:
366 .B http://www.intel.com/design/chipsets/matrixstorage_sb.htm
372 This will override any
374 setting in the config file and provides the identity of the host which
375 should be considered the home for any arrays.
377 When creating an array, the
379 will be recorded in the metadata. For version-1 superblocks, it will
380 be prefixed to the array name. For version-0.90 superblocks, part of
381 the SHA1 hash of the hostname will be stored in the later half of the
384 When reporting information about an array, any array which is tagged
385 for the given homehost will be reported as such.
387 When using Auto-Assemble, only arrays tagged for the given homehost
388 will be allowed to use 'local' names (i.e. not ending in '_' followed
389 by a digit string). See below under
390 .BR "Auto Assembly" .
392 .SH For create, build, or grow:
395 .BR \-n ", " \-\-raid\-devices=
396 Specify the number of active devices in the array. This, plus the
397 number of spare devices (see below) must equal the number of
399 (including "\fBmissing\fP" devices)
400 that are listed on the command line for
402 Setting a value of 1 is probably
403 a mistake and so requires that
405 be specified first. A value of 1 will then be allowed for linear,
406 multipath, RAID0 and RAID1. It is never allowed for RAID4, RAID5 or RAID6.
408 This number can only be changed using
410 for RAID1, RAID4, RAID5 and RAID6 arrays, and only on kernels which provide
411 the necessary support.
414 .BR \-x ", " \-\-spare\-devices=
415 Specify the number of spare (eXtra) devices in the initial array.
416 Spares can also be added
417 and removed later. The number of component devices listed
418 on the command line must equal the number of RAID devices plus the
419 number of spare devices.
422 .BR \-z ", " \-\-size=
423 Amount (in Kibibytes) of space to use from each drive in RAID levels 1/4/5/6.
424 This must be a multiple of the chunk size, and must leave about 128Kb
425 of space at the end of the drive for the RAID superblock.
426 If this is not specified
427 (as it normally is not) the smallest drive (or partition) sets the
428 size, though if there is a variance among the drives of greater than 1%, a warning is
431 A suffix of 'M' or 'G' can be given to indicate Megabytes or
432 Gigabytes respectively.
434 Sometimes a replacement drive can be a little smaller than the
435 original drives though this should be minimised by IDEMA standards.
436 Such a replacement drive will be rejected by
438 To guard against this it can be useful to set the initial size
439 slightly smaller than the smaller device with the aim that it will
440 still be larger than any replacement.
442 This value can be set with
444 for RAID level 1/4/5/6 though
446 based arrays such as those with IMSM metadata may not be able to
448 If the array was created with a size smaller than the currently
449 active drives, the extra space can be accessed using
451 The size can be given as
453 which means to choose the largest size that fits on all current drives.
455 Before reducing the size of the array (with
456 .BR "\-\-grow \-\-size=" )
457 you should make sure that space isn't needed. If the device holds a
458 filesystem, you would need to resize the filesystem to use less space.
460 After reducing the array size you should check that the data stored in
461 the device is still available. If the device holds a filesystem, then
462 an 'fsck' of the filesystem is a minimum requirement. If there are
463 problems the array can be made bigger again with no loss with another
464 .B "\-\-grow \-\-size="
467 This value cannot be used when creating a
469 such as with DDF and IMSM metadata, though it perfectly valid when
470 creating an array inside a container.
473 .BR \-Z ", " \-\-array\-size=
474 This is only meaningful with
476 and its effect is not persistent: when the array is stopped and
477 restarted the default array size will be restored.
479 Setting the array-size causes the array to appear smaller to programs
480 that access the data. This is particularly needed before reshaping an
481 array so that it will be smaller. As the reshape is not reversible,
482 but setting the size with
484 is, it is required that the array size is reduced as appropriate
485 before the number of devices in the array is reduced.
487 Before reducing the size of the array you should make sure that space
488 isn't needed. If the device holds a filesystem, you would need to
489 resize the filesystem to use less space.
491 After reducing the array size you should check that the data stored in
492 the device is still available. If the device holds a filesystem, then
493 an 'fsck' of the filesystem is a minimum requirement. If there are
494 problems the array can be made bigger again with no loss with another
495 .B "\-\-grow \-\-array\-size="
498 A suffix of 'M' or 'G' can be given to indicate Megabytes or
499 Gigabytes respectively.
502 restores the apparent size of the array to be whatever the real
503 amount of available space is.
506 .BR \-c ", " \-\-chunk=
507 Specify chunk size of kibibytes. The default when creating an
508 array is 512KB. To ensure compatibility with earlier versions, the
509 default when Building and array with no persistent metadata is 64KB.
510 This is only meaningful for RAID0, RAID4, RAID5, RAID6, and RAID10.
512 RAID4, RAID5, RAID6, and RAID10 require the chunk size to be a power
513 of 2. In any case it must be a multiple of 4KB.
515 A suffix of 'M' or 'G' can be given to indicate Megabytes or
516 Gigabytes respectively.
520 Specify rounding factor for a Linear array. The size of each
521 component will be rounded down to a multiple of this size.
522 This is a synonym for
524 but highlights the different meaning for Linear as compared to other
525 RAID levels. The default is 64K if a kernel earlier than 2.6.16 is in
526 use, and is 0K (i.e. no rounding) in later kernels.
529 .BR \-l ", " \-\-level=
530 Set RAID level. When used with
532 options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
533 raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty, container.
534 Obviously some of these are synonymous.
538 metadata type is requested, only the
540 level is permitted, and it does not need to be explicitly given.
544 only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid.
548 to change the RAID level in some cases. See LEVEL CHANGES below.
551 .BR \-p ", " \-\-layout=
552 This option configures the fine details of data layout for RAID5, RAID6,
553 and RAID10 arrays, and controls the failure modes for
556 The layout of the RAID5 parity block can be one of
557 .BR left\-asymmetric ,
558 .BR left\-symmetric ,
559 .BR right\-asymmetric ,
560 .BR right\-symmetric ,
561 .BR la ", " ra ", " ls ", " rs .
563 .BR left\-symmetric .
565 It is also possible to cause RAID5 to use a RAID4-like layout by
571 Finally for RAID5 there are DDF\-compatible layouts,
572 .BR ddf\-zero\-restart ,
573 .BR ddf\-N\-restart ,
575 .BR ddf\-N\-continue .
577 These same layouts are available for RAID6. There are also 4 layouts
578 that will provide an intermediate stage for converting between RAID5
579 and RAID6. These provide a layout which is identical to the
580 corresponding RAID5 layout on the first N\-1 devices, and has the 'Q'
581 syndrome (the second 'parity' block used by RAID6) on the last device.
583 .BR left\-symmetric\-6 ,
584 .BR right\-symmetric\-6 ,
585 .BR left\-asymmetric\-6 ,
586 .BR right\-asymmetric\-6 ,
588 .BR parity\-first\-6 .
590 When setting the failure mode for level
593 .BR write\-transient ", " wt ,
594 .BR read\-transient ", " rt ,
595 .BR write\-persistent ", " wp ,
596 .BR read\-persistent ", " rp ,
598 .BR read\-fixable ", " rf ,
599 .BR clear ", " flush ", " none .
601 Each failure mode can be followed by a number, which is used as a period
602 between fault generation. Without a number, the fault is generated
603 once on the first relevant request. With a number, the fault will be
604 generated after that many requests, and will continue to be generated
605 every time the period elapses.
607 Multiple failure modes can be current simultaneously by using the
609 option to set subsequent failure modes.
611 "clear" or "none" will remove any pending or periodic failure modes,
612 and "flush" will clear any persistent faults.
614 Finally, the layout options for RAID10 are one of 'n', 'o' or 'f' followed
615 by a small number. The default is 'n2'. The supported options are:
618 signals 'near' copies. Multiple copies of one data block are at
619 similar offsets in different devices.
622 signals 'offset' copies. Rather than the chunks being duplicated
623 within a stripe, whole stripes are duplicated but are rotated by one
624 device so duplicate blocks are on different devices. Thus subsequent
625 copies of a block are in the next drive, and are one chunk further
630 (multiple copies have very different offsets).
631 See md(4) for more detail about 'near', 'offset', and 'far'.
633 The number is the number of copies of each datablock. 2 is normal, 3
634 can be useful. This number can be at most equal to the number of
635 devices in the array. It does not need to divide evenly into that
636 number (e.g. it is perfectly legal to have an 'n2' layout for an array
637 with an odd number of devices).
639 When an array is converted between RAID5 and RAID6 an intermediate
640 RAID6 layout is used in which the second parity block (Q) is always on
641 the last device. To convert a RAID5 to RAID6 and leave it in this new
642 layout (which does not require re-striping) use
643 .BR \-\-layout=preserve .
644 This will try to avoid any restriping.
646 The converse of this is
647 .B \-\-layout=normalise
648 which will change a non-standard RAID6 layout into a more standard
655 (thus explaining the p of
659 .BR \-b ", " \-\-bitmap=
660 Specify a file to store a write-intent bitmap in. The file should not
663 is also given. The same file should be provided
664 when assembling the array. If the word
666 is given, then the bitmap is stored with the metadata on the array,
667 and so is replicated on all devices. If the word
671 mode, then any bitmap that is present is removed.
673 To help catch typing errors, the filename must contain at least one
674 slash ('/') if it is a real file (not 'internal' or 'none').
676 Note: external bitmaps are only known to work on ext2 and ext3.
677 Storing bitmap files on other filesystems may result in serious problems.
680 .BR \-\-bitmap\-chunk=
681 Set the chunksize of the bitmap. Each bit corresponds to that many
682 Kilobytes of storage.
683 When using a file based bitmap, the default is to use the smallest
684 size that is at-least 4 and requires no more than 2^21 chunks.
687 bitmap, the chunksize defaults to 64Meg, or larger if necessary to
688 fit the bitmap into the available space.
690 A suffix of 'M' or 'G' can be given to indicate Megabytes or
691 Gigabytes respectively.
694 .BR \-W ", " \-\-write\-mostly
695 subsequent devices listed in a
700 command will be flagged as 'write-mostly'. This is valid for RAID1
701 only and means that the 'md' driver will avoid reading from these
702 devices if at all possible. This can be useful if mirroring over a
706 .BR \-\-write\-behind=
707 Specify that write-behind mode should be enabled (valid for RAID1
708 only). If an argument is specified, it will set the maximum number
709 of outstanding writes allowed. The default value is 256.
710 A write-intent bitmap is required in order to use write-behind
711 mode, and write-behind is only attempted on drives marked as
715 .BR \-\-assume\-clean
718 that the array pre-existed and is known to be clean. It can be useful
719 when trying to recover from a major failure as you can be sure that no
720 data will be affected unless you actually write to the array. It can
721 also be used when creating a RAID1 or RAID10 if you want to avoid the
722 initial resync, however this practice \(em while normally safe \(em is not
723 recommended. Use this only if you really know what you are doing.
725 When the devices that will be part of a new array were filled
726 with zeros before creation the operator knows the array is
727 actually clean. If that is the case, such as after running
728 badblocks, this argument can be used to tell mdadm the
729 facts the operator knows.
731 When an array is resized to a larger size with
732 .B "\-\-grow \-\-size="
733 the new space is normally resynced in that same way that the whole
734 array is resynced at creation. From Linux version 3.0,
736 can be used with that command to avoid the automatic resync.
739 .BR \-\-backup\-file=
742 is used to increase the number of raid-devices in a RAID5 or RAID6 if
743 there are no spare devices available, or to shrink, change RAID level
744 or layout. See the GROW MODE section below on RAID\-DEVICES CHANGES.
745 The file must be stored on a separate device, not on the RAID array
750 This option is complementary to the
751 .B \-\-freeze-reshape
752 option for assembly. It is needed when
754 operation is interrupted and it is not restarted automatically due to
755 .B \-\-freeze-reshape
756 usage during array assembly. This option is used together with
760 ) command and device for a pending reshape to be continued.
761 All parameters required for reshape continuation will be read from array metadata.
765 .BR \-\-backup\-file=
766 option to be set, continuation option will require to have exactly the same
767 backup file given as well.
769 Any other parameter passed together with
771 option will be ignored.
774 .BR \-N ", " \-\-name=
777 for the array. This is currently only effective when creating an
778 array with a version-1 superblock, or an array in a DDF container.
779 The name is a simple textual string that can be used to identify array
780 components when assembling. If name is needed but not specified, it
781 is taken from the basename of the device that is being created.
793 run the array, even if some of the components
794 appear to be active in another array or filesystem. Normally
796 will ask for confirmation before including such components in an
797 array. This option causes that question to be suppressed.
800 .BR \-f ", " \-\-force
803 accept the geometry and layout specified without question. Normally
805 will not allow creation of an array with only one device, and will try
806 to create a RAID5 array with one missing drive (as this makes the
807 initial resync work faster). With
810 will not try to be so clever.
813 .BR \-a ", " "\-\-auto{=yes,md,mdp,part,p}{NN}"
814 Instruct mdadm how to create the device file if needed, possibly allocating
815 an unused minor number. "md" causes a non-partitionable array
816 to be used (though since Linux 2.6.28, these array devices are in fact
817 partitionable). "mdp", "part" or "p" causes a partitionable array (2.6 and
818 later) to be used. "yes" requires the named md device to have
819 a 'standard' format, and the type and minor number will be determined
820 from this. With mdadm 3.0, device creation is normally left up to
822 so this option is unlikely to be needed.
823 See DEVICE NAMES below.
825 The argument can also come immediately after
830 is not given on the command line or in the config file, then
836 is also given, then any
838 entries in the config file will override the
840 instruction given on the command line.
842 For partitionable arrays,
844 will create the device file for the whole array and for the first 4
845 partitions. A different number of partitions can be specified at the
846 end of this option (e.g.
848 If the device name ends with a digit, the partition names add a 'p',
850 .IR /dev/md/home1p3 .
851 If there is no trailing digit, then the partition names just have a
853 .IR /dev/md/scratch3 .
855 If the md device name is in a 'standard' format as described in DEVICE
856 NAMES, then it will be created, if necessary, with the appropriate
857 device number based on that name. If the device name is not in one of these
858 formats, then a unused device number will be allocated. The device
859 number will be considered unused if there is no active array for that
860 number, and there is no entry in /dev for that number and with a
861 non-standard name. Names that are not in 'standard' format are only
862 allowed in "/dev/md/".
864 This is meaningful with
871 .\".BR \-\-symlink = no
876 .\"to create devices in
878 .\"it will also create symlinks from
880 .\"with names starting with
886 .\"to suppress this, or
887 .\".B \-\-symlink=yes
888 .\"to enforce this even if it is suppressing
894 .BR \-a ", " "\-\-add"
895 This option can be used in Grow mode in two cases.
897 If the target array is a Linear array, then
899 can be used to add one or more devices to the array. They
900 are simply catenated on to the end of the array. Once added, the
901 devices cannot be removed.
905 option is being used to increase the number of devices in an array,
908 can be used to add some extra devices to be included in the array.
909 In most cases this is not needed as the extra devices can be added as
910 spares first, and then the number of raid-disks can be changed.
911 However for RAID0, it is not possible to add spares. So to increase
912 the number of devices in a RAID0, it is necessary to set the new
913 number of devices, and to add the new devices, in the same command.
918 .BR \-u ", " \-\-uuid=
919 uuid of array to assemble. Devices which don't have this uuid are
923 .BR \-m ", " \-\-super\-minor=
924 Minor number of device that array was created for. Devices which
925 don't have this minor number are excluded. If you create an array as
926 /dev/md1, then all superblocks will contain the minor number 1, even if
927 the array is later assembled as /dev/md2.
929 Giving the literal word "dev" for
933 to use the minor number of the md device that is being assembled.
936 .B \-\-super\-minor=dev
937 will look for super blocks with a minor number of 0.
940 is only relevant for v0.90 metadata, and should not normally be used.
946 .BR \-N ", " \-\-name=
947 Specify the name of the array to assemble. This must be the name
948 that was specified when creating the array. It must either match
949 the name stored in the superblock exactly, or it must match
952 prefixed to the start of the given name.
955 .BR \-f ", " \-\-force
956 Assemble the array even if the metadata on some devices appears to be
959 cannot find enough working devices to start the array, but can find
960 some devices that are recorded as having failed, then it will mark
961 those devices as working so that the array can be started.
962 An array which requires
964 to be started may contain data corruption. Use it carefully.
968 Attempt to start the array even if fewer drives were given than were
969 present last time the array was active. Normally if not all the
970 expected drives are found and
972 is not used, then the array will be assembled but not started.
975 an attempt will be made to start it anyway.
979 This is the reverse of
981 in that it inhibits the startup of array unless all expected drives
982 are present. This is only needed with
984 and can be used if the physical connections to devices are
985 not as reliable as you would like.
988 .BR \-a ", " "\-\-auto{=no,yes,md,mdp,part}"
989 See this option under Create and Build options.
992 .BR \-b ", " \-\-bitmap=
993 Specify the bitmap file that was given when the array was created. If
996 bitmap, there is no need to specify this when assembling the array.
999 .BR \-\-backup\-file=
1002 was used while reshaping an array (e.g. changing number of devices or
1003 chunk size) and the system crashed during the critical section, then the same
1005 must be presented to
1007 to allow possibly corrupted data to be restored, and the reshape
1011 .BR \-\-invalid\-backup
1012 If the file needed for the above option is not available for any
1013 reason an empty file can be given together with this option to
1014 indicate that the backup file is invalid. In this case the data that
1015 was being rearranged at the time of the crash could be irrecoverably
1016 lost, but the rest of the array may still be recoverable. This option
1017 should only be used as a last resort if there is no way to recover the
1022 .BR \-U ", " \-\-update=
1023 Update the superblock on each device while assembling the array. The
1024 argument given to this flag can be one of
1039 option will adjust the superblock of an array what was created on a Sparc
1040 machine running a patched 2.2 Linux kernel. This kernel got the
1041 alignment of part of the superblock wrong. You can use the
1042 .B "\-\-examine \-\-sparc2.2"
1045 to see what effect this would have.
1049 option will update the
1050 .B "preferred minor"
1051 field on each superblock to match the minor number of the array being
1053 This can be useful if
1055 reports a different "Preferred Minor" to
1057 In some cases this update will be performed automatically
1058 by the kernel driver. In particular the update happens automatically
1059 at the first write to an array with redundancy (RAID level 1 or
1060 greater) on a 2.6 (or later) kernel.
1064 option will change the uuid of the array. If a UUID is given with the
1066 option that UUID will be used as a new UUID and will
1068 be used to help identify the devices in the array.
1071 is given, a random UUID is chosen.
1075 option will change the
1077 of the array as stored in the superblock. This is only supported for
1078 version-1 superblocks.
1082 option will change the
1084 as recorded in the superblock. For version-0 superblocks, this is the
1085 same as updating the UUID.
1086 For version-1 superblocks, this involves updating the name.
1090 option will cause the array to be marked
1092 meaning that any redundancy in the array (e.g. parity for RAID5,
1093 copies for RAID1) may be incorrect. This will cause the RAID system
1094 to perform a "resync" pass to make sure that all redundant information
1099 option allows arrays to be moved between machines with different
1101 When assembling such an array for the first time after a move, giving
1102 .B "\-\-update=byteorder"
1105 to expect superblocks to have their byteorder reversed, and will
1106 correct that order before assembling the array. This is only valid
1107 with original (Version 0.90) superblocks.
1111 option will correct the summaries in the superblock. That is the
1112 counts of total, working, active, failed, and spare devices.
1116 option will rarely be of use. It applies to version 1.1 and 1.2 metadata
1117 only (where the metadata is at the start of the device) and is only
1118 useful when the component device has changed size (typically become
1119 larger). The version 1 metadata records the amount of the device that
1120 can be used to store data, so if a device in a version 1.1 or 1.2
1121 array becomes larger, the metadata will still be visible, but the
1122 extra space will not. In this case it might be useful to assemble the
1124 .BR \-\-update=devicesize .
1127 to determine the maximum usable amount of space on each device and
1128 update the relevant field in the metadata.
1132 option can be used when an array has an internal bitmap which is
1133 corrupt in some way so that assembling the array normally fails. It
1134 will cause any internal bitmap to be ignored.
1137 .BR \-\-freeze\-reshape
1138 Option is intended to be used in start-up scripts during initrd boot phase.
1139 When array under reshape is assembled during initrd phase, this option
1140 stops reshape after reshape critical section is being restored. This happens
1141 before file system pivot operation and avoids loss of file system context.
1142 Losing file system context would cause reshape to be broken.
1144 Reshape can be continued later using the
1146 option for the grow command.
1148 .SH For Manage mode:
1151 .BR \-t ", " \-\-test
1152 Unless a more serious error occurred,
1154 will exit with a status of 2 if no changes were made to the array and
1155 0 if at least one change was made.
1156 This can be useful when an indirect specifier such as
1161 is used in requesting an operation on the array.
1163 will report failure if these specifiers didn't find any match.
1166 .BR \-a ", " \-\-add
1167 hot-add listed devices.
1168 If a device appears to have recently been part of the array
1169 (possibly it failed or was removed) the device is re\-added as describe
1171 If that fails or the device was never part of the array, the device is
1172 added as a hot-spare.
1173 If the array is degraded, it will immediately start to rebuild data
1176 Note that this and the following options are only meaningful on array
1177 with redundancy. They don't apply to RAID0 or Linear.
1181 re\-add a device that was previous removed from an array.
1182 If the metadata on the device reports that it is a member of the
1183 array, and the slot that it used is still vacant, then the device will
1184 be added back to the array in the same position. This will normally
1185 cause the data for that device to be recovered. However based on the
1186 event count on the device, the recovery may only require sections that
1187 are flagged a write-intent bitmap to be recovered or may not require
1188 any recovery at all.
1190 When used on an array that has no metadata (i.e. it was built with
1192 it will be assumed that bitmap-based recovery is enough to make the
1193 device fully consistent with the array.
1197 can be accompanied by
1198 .BR \-\-update=devicesize .
1199 See the description of this option when used in Assemble mode for an
1200 explanation of its use.
1202 If the device name given is
1204 then mdadm will try to find any device that looks like it should be
1205 part of the array but isn't and will try to re\-add all such devices.
1208 .BR \-r ", " \-\-remove
1209 remove listed devices. They must not be active. i.e. they should
1210 be failed or spare devices. As well as the name of a device file
1219 The first causes all failed device to be removed. The second causes
1220 any device which is no longer connected to the system (i.e an 'open'
1223 to be removed. This will only succeed for devices that are spares or
1224 have already been marked as failed.
1227 .BR \-f ", " \-\-fail
1228 mark listed devices as faulty.
1229 As well as the name of a device file, the word
1231 can be given. This will cause any device that has been detached from
1232 the system to be marked as failed. It can then be removed.
1240 .BR \-\-write\-mostly
1241 Subsequent devices that are added or re\-added will have the 'write-mostly'
1242 flag set. This is only valid for RAID1 and means that the 'md' driver
1243 will avoid reading from these devices if possible.
1246 Subsequent devices that are added or re\-added will have the 'write-mostly'
1250 Each of these options requires that the first device listed is the array
1251 to be acted upon, and the remainder are component devices to be added,
1252 removed, marked as faulty, etc. Several different operations can be
1253 specified for different devices, e.g.
1255 mdadm /dev/md0 \-\-add /dev/sda1 \-\-fail /dev/sdb1 \-\-remove /dev/sdb1
1257 Each operation applies to all devices listed until the next
1260 If an array is using a write-intent bitmap, then devices which have
1261 been removed can be re\-added in a way that avoids a full
1262 reconstruction but instead just updates the blocks that have changed
1263 since the device was removed. For arrays with persistent metadata
1264 (superblocks) this is done automatically. For arrays created with
1266 mdadm needs to be told that this device we removed recently with
1269 Devices can only be removed from an array if they are not in active
1270 use, i.e. that must be spares or failed devices. To remove an active
1271 device, it must first be marked as
1277 .BR \-Q ", " \-\-query
1278 Examine a device to see
1279 (1) if it is an md device and (2) if it is a component of an md
1281 Information about what is discovered is presented.
1284 .BR \-D ", " \-\-detail
1285 Print details of one or more md devices.
1288 .BR \-\-detail\-platform
1289 Print details of the platform's RAID capabilities (firmware / hardware
1290 topology) for a given metadata format.
1293 .BR \-Y ", " \-\-export
1298 output will be formatted as
1300 pairs for easy import into the environment.
1303 .BR \-E ", " \-\-examine
1304 Print contents of the metadata stored on the named device(s).
1305 Note the contrast between
1310 applies to devices which are components of an array, while
1312 applies to a whole array which is currently active.
1315 If an array was created on a SPARC machine with a 2.2 Linux kernel
1316 patched with RAID support, the superblock will have been created
1317 incorrectly, or at least incompatibly with 2.4 and later kernels.
1322 will fix the superblock before displaying it. If this appears to do
1323 the right thing, then the array can be successfully assembled using
1324 .BR "\-\-assemble \-\-update=sparc2.2" .
1327 .BR \-X ", " \-\-examine\-bitmap
1328 Report information about a bitmap file.
1329 The argument is either an external bitmap file or an array component
1330 in case of an internal bitmap. Note that running this on an array
1333 does not report the bitmap for that array.
1336 .BR \-R ", " \-\-run
1337 start a partially assembled array. If
1339 did not find enough devices to fully start the array, it might leaving
1340 it partially assembled. If you wish, you can then use
1342 to start the array in degraded mode.
1345 .BR \-S ", " \-\-stop
1346 deactivate array, releasing all resources.
1349 .BR \-o ", " \-\-readonly
1350 mark array as readonly.
1353 .BR \-w ", " \-\-readwrite
1354 mark array as readwrite.
1357 .B \-\-zero\-superblock
1358 If the device contains a valid md superblock, the block is
1359 overwritten with zeros. With
1361 the block where the superblock would be is overwritten even if it
1362 doesn't appear to be valid.
1365 .B \-\-kill\-subarray=
1366 If the device is a container and the argument to \-\-kill\-subarray
1367 specifies an inactive subarray in the container, then the subarray is
1368 deleted. Deleting all subarrays will leave an 'empty-container' or
1369 spare superblock on the drives. See \-\-zero\-superblock for completely
1370 removing a superblock. Note that some formats depend on the subarray
1371 index for generating a UUID, this command will fail if it would change
1372 the UUID of an active subarray.
1375 .B \-\-update\-subarray=
1376 If the device is a container and the argument to \-\-update\-subarray
1377 specifies a subarray in the container, then attempt to update the given
1378 superblock field in the subarray. See below in
1383 .BR \-t ", " \-\-test
1388 is set to reflect the status of the device. See below in
1393 .BR \-W ", " \-\-wait
1394 For each md device given, wait for any resync, recovery, or reshape
1395 activity to finish before returning.
1397 will return with success if it actually waited for every device
1398 listed, otherwise it will return failure.
1402 For each md device given, or each device in /proc/mdstat if
1404 is given, arrange for the array to be marked clean as soon as possible.
1406 will return with success if the array uses external metadata and we
1407 successfully waited. For native arrays this returns immediately as the
1408 kernel handles dirty-clean transitions at shutdown. No action is taken
1409 if safe-mode handling is disabled.
1411 .SH For Incremental Assembly mode:
1413 .BR \-\-rebuild\-map ", " \-r
1414 Rebuild the map file
1415 .RB ( /var/run/mdadm/map )
1418 uses to help track which arrays are currently being assembled.
1421 .BR \-\-run ", " \-R
1422 Run any array assembled as soon as a minimal number of devices are
1423 available, rather than waiting until all expected devices are present.
1426 .BR \-\-scan ", " \-s
1427 Only meaningful with
1431 file for arrays that are being incrementally assembled and will try to
1432 start any that are not already started. If any such array is listed
1435 as requiring an external bitmap, that bitmap will be attached first.
1438 .BR \-\-fail ", " \-f
1439 This allows the hot-plug system to remove devices that have fully disappeared
1440 from the kernel. It will first fail and then remove the device from any
1441 array it belongs to.
1442 The device name given should be a kernel device name such as "sda",
1448 Only used with \-\-fail. The 'path' given will be recorded so that if
1449 a new device appears at the same location it can be automatically
1450 added to the same array. This allows the failed device to be
1451 automatically replaced by a new device without metadata if it appears
1452 at specified path. This option is normally only set by a
1456 .SH For Monitor mode:
1458 .BR \-m ", " \-\-mail
1459 Give a mail address to send alerts to.
1462 .BR \-p ", " \-\-program ", " \-\-alert
1463 Give a program to be run whenever an event is detected.
1466 .BR \-y ", " \-\-syslog
1467 Cause all events to be reported through 'syslog'. The messages have
1468 facility of 'daemon' and varying priorities.
1471 .BR \-d ", " \-\-delay
1472 Give a delay in seconds.
1474 polls the md arrays and then waits this many seconds before polling
1475 again. The default is 60 seconds. Since 2.6.16, there is no need to
1476 reduce this as the kernel alerts
1478 immediately when there is any change.
1481 .BR \-r ", " \-\-increment
1482 Give a percentage increment.
1484 will generate RebuildNN events with the given percentage increment.
1487 .BR \-f ", " \-\-daemonise
1490 to run as a background daemon if it decides to monitor anything. This
1491 causes it to fork and run in the child, and to disconnect from the
1492 terminal. The process id of the child is written to stdout.
1495 which will only continue monitoring if a mail address or alert program
1496 is found in the config file.
1499 .BR \-i ", " \-\-pid\-file
1502 is running in daemon mode, write the pid of the daemon process to
1503 the specified file, instead of printing it on standard output.
1506 .BR \-1 ", " \-\-oneshot
1507 Check arrays only once. This will generate
1509 events and more significantly
1515 .B " mdadm \-\-monitor \-\-scan \-1"
1517 from a cron script will ensure regular notification of any degraded arrays.
1520 .BR \-t ", " \-\-test
1523 alert for every array found at startup. This alert gets mailed and
1524 passed to the alert program. This can be used for testing that alert
1525 message do get through successfully.
1529 This inhibits the functionality for moving spares between arrays.
1530 Only one monitoring process started with
1532 but without this flag is allowed, otherwise the two could interfere
1539 .B mdadm \-\-assemble
1540 .I md-device options-and-component-devices...
1543 .B mdadm \-\-assemble \-\-scan
1544 .I md-devices-and-options...
1547 .B mdadm \-\-assemble \-\-scan
1551 This usage assembles one or more RAID arrays from pre-existing components.
1552 For each array, mdadm needs to know the md device, the identity of the
1553 array, and a number of component-devices. These can be found in a number of ways.
1555 In the first usage example (without the
1557 the first device given is the md device.
1558 In the second usage example, all devices listed are treated as md
1559 devices and assembly is attempted.
1560 In the third (where no devices are listed) all md devices that are
1561 listed in the configuration file are assembled. If no arrays are
1562 described by the configuration file, then any arrays that
1563 can be found on unused devices will be assembled.
1565 If precisely one device is listed, but
1571 was given and identity information is extracted from the configuration file.
1573 The identity can be given with the
1579 option, will be taken from the md-device record in the config file, or
1580 will be taken from the super block of the first component-device
1581 listed on the command line.
1583 Devices can be given on the
1585 command line or in the config file. Only devices which have an md
1586 superblock which contains the right identity will be considered for
1589 The config file is only used if explicitly named with
1591 or requested with (a possibly implicit)
1596 .B /etc/mdadm/mdadm.conf
1601 is not given, then the config file will only be used to find the
1602 identity of md arrays.
1604 Normally the array will be started after it is assembled. However if
1606 is not given and not all expected drives were listed, then the array
1607 is not started (to guard against usage errors). To insist that the
1608 array be started in this case (as may work for RAID1, 4, 5, 6, or 10),
1617 does not create any entries in
1621 It does record information in
1622 .B /var/run/mdadm/map
1625 to choose the correct name.
1629 detects that udev is not configured, it will create the devices in
1633 In Linux kernels prior to version 2.6.28 there were two distinctly
1634 different types of md devices that could be created: one that could be
1635 partitioned using standard partitioning tools and one that could not.
1636 Since 2.6.28 that distinction is no longer relevant as both type of
1637 devices can be partitioned.
1639 will normally create the type that originally could not be partitioned
1640 as it has a well defined major number (9).
1642 Prior to 2.6.28, it is important that mdadm chooses the correct type
1643 of array device to use. This can be controlled with the
1645 option. In particular, a value of "mdp" or "part" or "p" tells mdadm
1646 to use a partitionable device rather than the default.
1648 In the no-udev case, the value given to
1650 can be suffixed by a number. This tells
1652 to create that number of partition devices rather than the default of 4.
1656 can also be given in the configuration file as a word starting
1658 on the ARRAY line for the relevant array.
1665 and no devices are listed,
1667 will first attempt to assemble all the arrays listed in the config
1670 If no arrays are listed in the config (other than those marked
1672 it will look through the available devices for possible arrays and
1673 will try to assemble anything that it finds. Arrays which are tagged
1674 as belonging to the given homehost will be assembled and started
1675 normally. Arrays which do not obviously belong to this host are given
1676 names that are expected not to conflict with anything local, and are
1677 started "read-auto" so that nothing is written to any device until the
1678 array is written to. i.e. automatic resync etc is delayed.
1682 finds a consistent set of devices that look like they should comprise
1683 an array, and if the superblock is tagged as belonging to the given
1684 home host, it will automatically choose a device name and try to
1685 assemble the array. If the array uses version-0.90 metadata, then the
1687 number as recorded in the superblock is used to create a name in
1691 If the array uses version-1 metadata, then the
1693 from the superblock is used to similarly create a name in
1695 (the name will have any 'host' prefix stripped first).
1697 This behaviour can be modified by the
1701 configuration file. This line can indicate that specific metadata
1702 type should, or should not, be automatically assembled. If an array
1703 is found which is not listed in
1705 and has a metadata format that is denied by the
1707 line, then it will not be assembled.
1710 line can also request that all arrays identified as being for this
1711 homehost should be assembled regardless of their metadata type.
1714 for further details.
1716 Note: Auto assembly cannot be used for assembling and activating some
1717 arrays which are undergoing reshape. In particular as the
1719 cannot be given, any reshape which requires a backup-file to continue
1720 cannot be started by auto assembly. An array which is growing to more
1721 devices and has passed the critical section can be assembled using
1732 .BI \-\-raid\-devices= Z
1736 This usage is similar to
1738 The difference is that it creates an array without a superblock. With
1739 these arrays there is no difference between initially creating the array and
1740 subsequently assembling the array, except that hopefully there is useful
1741 data there in the second case.
1743 The level may raid0, linear, raid1, raid10, multipath, or faulty, or
1744 one of their synonyms. All devices must be listed and the array will
1745 be started once complete. It will often be appropriate to use
1746 .B \-\-assume\-clean
1747 with levels raid1 or raid10.
1758 .BI \-\-raid\-devices= Z
1762 This usage will initialise a new md array, associate some devices with
1763 it, and activate the array.
1765 The named device will normally not exist when
1766 .I "mdadm \-\-create"
1767 is run, but will be created by
1769 once the array becomes active.
1771 As devices are added, they are checked to see if they contain RAID
1772 superblocks or filesystems. They are also checked to see if the variance in
1773 device size exceeds 1%.
1775 If any discrepancy is found, the array will not automatically be run, though
1778 can override this caution.
1780 To create a "degraded" array in which some devices are missing, simply
1781 give the word "\fBmissing\fP"
1782 in place of a device name. This will cause
1784 to leave the corresponding slot in the array empty.
1785 For a RAID4 or RAID5 array at most one slot can be
1786 "\fBmissing\fP"; for a RAID6 array at most two slots.
1787 For a RAID1 array, only one real device needs to be given. All of the
1791 When creating a RAID5 array,
1793 will automatically create a degraded array with an extra spare drive.
1794 This is because building the spare into a degraded array is in general
1795 faster than resyncing the parity on a non-degraded, but not clean,
1796 array. This feature can be overridden with the
1800 When creating an array with version-1 metadata a name for the array is
1802 If this is not given with the
1806 will choose a name based on the last component of the name of the
1807 device being created. So if
1809 is being created, then the name
1814 is being created, then the name
1818 When creating a partition based array, using
1820 with version-1.x metadata, the partition type should be set to
1822 (non fs-data). This type selection allows for greater precision since
1823 using any other [RAID auto-detect (0xFD) or a GNU/Linux partition (0x83)],
1824 might create problems in the event of array recovery through a live cdrom.
1826 A new array will normally get a randomly assigned 128bit UUID which is
1827 very likely to be unique. If you have a specific need, you can choose
1828 a UUID for the array by giving the
1830 option. Be warned that creating two arrays with the same UUID is a
1831 recipe for disaster. Also, using
1833 when creating a v0.90 array will silently override any
1838 .\"option is given, it is not necessary to list any component-devices in this command.
1839 .\"They can be added later, before a
1843 .\"is given, the apparent size of the smallest drive given is used.
1845 When creating an array within a
1848 can be given either the list of devices to use, or simply the name of
1849 the container. The former case gives control over which devices in
1850 the container will be used for the array. The latter case allows
1852 to automatically choose which devices to use based on how much spare
1855 The General Management options that are valid with
1860 insist on running the array even if some devices look like they might
1865 start the array readonly \(em not supported yet.
1872 .I options... devices...
1875 This usage will allow individual devices in an array to be failed,
1876 removed or added. It is possible to perform multiple operations with
1877 on command. For example:
1879 .B " mdadm /dev/md0 \-f /dev/hda1 \-r /dev/hda1 \-a /dev/hda1"
1885 and will then remove it from the array and finally add it back
1886 in as a spare. However only one md array can be affected by a single
1889 When a device is added to an active array, mdadm checks to see if it
1890 has metadata on it which suggests that it was recently a member of the
1891 array. If it does, it tries to "re\-add" the device. If there have
1892 been no changes since the device was removed, or if the array has a
1893 write-intent bitmap which has recorded whatever changes there were,
1894 then the device will immediately become a full member of the array and
1895 those differences recorded in the bitmap will be resolved.
1905 MISC mode includes a number of distinct operations that
1906 operate on distinct devices. The operations are:
1909 The device is examined to see if it is
1910 (1) an active md array, or
1911 (2) a component of an md array.
1912 The information discovered is reported.
1916 The device should be an active md device.
1918 will display a detailed description of the array.
1922 will cause the output to be less detailed and the format to be
1923 suitable for inclusion in
1927 will normally be 0 unless
1929 failed to get useful information about the device(s); however, if the
1931 option is given, then the exit status will be:
1935 The array is functioning normally.
1938 The array has at least one failed device.
1941 The array has multiple failed devices such that it is unusable.
1944 There was an error while trying to get information about the device.
1948 .B \-\-detail\-platform
1949 Print detail of the platform's RAID capabilities (firmware / hardware
1950 topology). If the metadata is specified with
1954 then the return status will be:
1958 metadata successfully enumerated its platform components on this system
1961 metadata is platform independent
1964 metadata failed to find its platform components on this system
1968 .B \-\-update\-subarray=
1969 If the device is a container and the argument to \-\-update\-subarray
1970 specifies a subarray in the container, then attempt to update the given
1971 superblock field in the subarray. Similar to updating an array in
1972 "assemble" mode, the field to update is selected by
1976 option. Currently only
1982 option updates the subarray name in the metadata, it may not affect the
1983 device node name or the device node symlink until the subarray is
1984 re\-assembled. If updating
1986 would change the UUID of an active subarray this operation is blocked,
1987 and the command will end in an error.
1991 The device should be a component of an md array.
1993 will read the md superblock of the device and display the contents.
1998 is given, then multiple devices that are components of the one array
1999 are grouped together and reported in a single entry suitable
2005 without listing any devices will cause all devices listed in the
2006 config file to be examined.
2010 The devices should be active md arrays which will be deactivated, as
2011 long as they are not currently in use.
2015 This will fully activate a partially assembled md array.
2019 This will mark an active array as read-only, providing that it is
2020 not currently being used.
2026 array back to being read/write.
2030 For all operations except
2033 will cause the operation to be applied to all arrays listed in
2038 causes all devices listed in the config file to be examined.
2041 .BR \-b ", " \-\-brief
2042 Be less verbose. This is used with
2050 gives an intermediate level of verbosity.
2056 .B mdadm \-\-monitor
2057 .I options... devices...
2062 to periodically poll a number of md arrays and to report on any events
2065 will never exit once it decides that there are arrays to be checked,
2066 so it should normally be run in the background.
2068 As well as reporting events,
2070 may move a spare drive from one array to another if they are in the
2075 and if the destination array has a failed drive but no spares.
2077 If any devices are listed on the command line,
2079 will only monitor those devices. Otherwise all arrays listed in the
2080 configuration file will be monitored. Further, if
2082 is given, then any other md devices that appear in
2084 will also be monitored.
2086 The result of monitoring the arrays is the generation of events.
2087 These events are passed to a separate program (if specified) and may
2088 be mailed to a given E-mail address.
2090 When passing events to a program, the program is run once for each event,
2091 and is given 2 or 3 command-line arguments: the first is the
2092 name of the event (see below), the second is the name of the
2093 md device which is affected, and the third is the name of a related
2094 device if relevant (such as a component device that has failed).
2098 is given, then a program or an E-mail address must be specified on the
2099 command line or in the config file. If neither are available, then
2101 will not monitor anything.
2105 will continue monitoring as long as something was found to monitor. If
2106 no program or email is given, then each event is reported to
2109 The different events are:
2113 .B DeviceDisappeared
2114 An md array which previously was configured appears to no longer be
2115 configured. (syslog priority: Critical)
2119 was told to monitor an array which is RAID0 or Linear, then it will
2121 .B DeviceDisappeared
2122 with the extra information
2124 This is because RAID0 and Linear do not support the device-failed,
2125 hot-spare and resync operations which are monitored.
2129 An md array started reconstruction. (syslog priority: Warning)
2135 is a two-digit number (ie. 05, 48). This indicates that rebuild
2136 has passed that many percent of the total. The events are generated
2137 with fixed increment since 0. Increment size may be specified with
2138 a commandline option (default is 20). (syslog priority: Warning)
2142 An md array that was rebuilding, isn't any more, either because it
2143 finished normally or was aborted. (syslog priority: Warning)
2147 An active component device of an array has been marked as
2148 faulty. (syslog priority: Critical)
2152 A spare component device which was being rebuilt to replace a faulty
2153 device has failed. (syslog priority: Critical)
2157 A spare component device which was being rebuilt to replace a faulty
2158 device has been successfully rebuilt and has been made active.
2159 (syslog priority: Info)
2163 A new md array has been detected in the
2165 file. (syslog priority: Info)
2169 A newly noticed array appears to be degraded. This message is not
2172 notices a drive failure which causes degradation, but only when
2174 notices that an array is degraded when it first sees the array.
2175 (syslog priority: Critical)
2179 A spare drive has been moved from one array in a
2183 to another to allow a failed drive to be replaced.
2184 (syslog priority: Info)
2190 has been told, via the config file, that an array should have a certain
2191 number of spare devices, and
2193 detects that it has fewer than this number when it first sees the
2194 array, it will report a
2197 (syslog priority: Warning)
2201 An array was found at startup, and the
2204 (syslog priority: Info)
2214 cause Email to be sent. All events cause the program to be run.
2215 The program is run with two or three arguments: the event
2216 name, the array device and possibly a second device.
2218 Each event has an associated array device (e.g.
2220 and possibly a second device. For
2225 the second device is the relevant component device.
2228 the second device is the array that the spare was moved from.
2232 to move spares from one array to another, the different arrays need to
2233 be labeled with the same
2235 or the spares must be allowed to migrate through matching POLICY domains
2236 in the configuration file. The
2238 name can be any string; it is only necessary that different spare
2239 groups use different names.
2243 detects that an array in a spare group has fewer active
2244 devices than necessary for the complete array, and has no spare
2245 devices, it will look for another array in the same spare group that
2246 has a full complement of working drive and a spare. It will then
2247 attempt to remove the spare from the second drive and add it to the
2249 If the removal succeeds but the adding fails, then it is added back to
2252 If the spare group for a degraded array is not defined,
2254 will look at the rules of spare migration specified by POLICY lines in
2256 and then follow similar steps as above if a matching spare is found.
2259 The GROW mode is used for changing the size or shape of an active
2261 For this to work, the kernel must support the necessary change.
2262 Various types of growth are being added during 2.6 development.
2264 Currently the supported changes include
2266 change the "size" attribute for RAID1, RAID4, RAID5 and RAID6.
2268 increase or decrease the "raid\-devices" attribute of RAID0, RAID1, RAID4,
2271 change the chunk-size and layout of RAID0, RAID4, RAID5 and RAID6.
2273 convert between RAID1 and RAID5, between RAID5 and RAID6, between
2274 RAID0, RAID4, and RAID5, and between RAID0 and RAID10 (in the near-2 mode).
2276 add a write-intent bitmap to any array which supports these bitmaps, or
2277 remove a write-intent bitmap from such an array.
2280 Using GROW on containers is currently supported only for Intel's IMSM
2281 container format. The number of devices in a container can be
2282 increased - which affects all arrays in the container - or an array
2283 in a container can be converted between levels where those levels are
2284 supported by the container, and the conversion is on of those listed
2285 above. Resizing arrays in an IMSM container with
2287 is not yet supported.
2289 Grow functionality (e.g. expand a number of raid devices) for Intel's
2290 IMSM container format has an experimental status. It is guarded by the
2291 .B MDADM_EXPERIMENTAL
2292 environment variable which must be set to '1' for a GROW command to
2294 This is for the following reasons:
2297 Intel's native IMSM check-pointing is not fully tested yet.
2298 This can causes IMSM incompatibility during the grow process: an array
2299 which is growing cannot roam between Microsoft Windows(R) and Linux
2303 Interrupting a grow operation is not recommended, because it
2304 has not been fully tested for Intel's IMSM container format yet.
2307 Note: Intel's native checkpointing doesn't use
2309 option and it is transparent for assembly feature.
2312 Normally when an array is built the "size" is taken from the smallest
2313 of the drives. If all the small drives in an arrays are, one at a
2314 time, removed and replaced with larger drives, then you could have an
2315 array of large drives with only a small amount used. In this
2316 situation, changing the "size" with "GROW" mode will allow the extra
2317 space to start being used. If the size is increased in this way, a
2318 "resync" process will start to make sure the new parts of the array
2321 Note that when an array changes size, any filesystem that may be
2322 stored in the array will not automatically grow or shrink to use or
2323 vacate the space. The
2324 filesystem will need to be explicitly told to use the extra space
2325 after growing, or to reduce its size
2327 to shrinking the array.
2329 Also the size of an array cannot be changed while it has an active
2330 bitmap. If an array has a bitmap, it must be removed before the size
2331 can be changed. Once the change is complete a new bitmap can be created.
2333 .SS RAID\-DEVICES CHANGES
2335 A RAID1 array can work with any number of devices from 1 upwards
2336 (though 1 is not very useful). There may be times which you want to
2337 increase or decrease the number of active devices. Note that this is
2338 different to hot-add or hot-remove which changes the number of
2341 When reducing the number of devices in a RAID1 array, the slots which
2342 are to be removed from the array must already be vacant. That is, the
2343 devices which were in those slots must be failed and removed.
2345 When the number of devices is increased, any hot spares that are
2346 present will be activated immediately.
2348 Changing the number of active devices in a RAID5 or RAID6 is much more
2349 effort. Every block in the array will need to be read and written
2350 back to a new location. From 2.6.17, the Linux Kernel is able to
2351 increase the number of devices in a RAID5 safely, including restarting
2352 an interrupted "reshape". From 2.6.31, the Linux Kernel is able to
2353 increase or decrease the number of devices in a RAID5 or RAID6.
2355 From 2.6.35, the Linux Kernel is able to convert a RAID0 in to a RAID4
2358 uses this functionality and the ability to add
2359 devices to a RAID4 to allow devices to be added to a RAID0. When
2360 requested to do this,
2362 will convert the RAID0 to a RAID4, add the necessary disks and make
2363 the reshape happen, and then convert the RAID4 back to RAID0.
2365 When decreasing the number of devices, the size of the array will also
2366 decrease. If there was data in the array, it could get destroyed and
2367 this is not reversible, so you should firstly shrink the filesystem on
2368 the array to fit within the new size. To help prevent accidents,
2370 requires that the size of the array be decreased first with
2371 .BR "mdadm --grow --array-size" .
2372 This is a reversible change which simply makes the end of the array
2373 inaccessible. The integrity of any data can then be checked before
2374 the non-reversible reduction in the number of devices is request.
2376 When relocating the first few stripes on a RAID5 or RAID6, it is not
2377 possible to keep the data on disk completely consistent and
2378 crash-proof. To provide the required safety, mdadm disables writes to
2379 the array while this "critical section" is reshaped, and takes a
2380 backup of the data that is in that section. For grows, this backup may be
2381 stored in any spare devices that the array has, however it can also be
2382 stored in a separate file specified with the
2384 option, and is required to be specified for shrinks, RAID level
2385 changes and layout changes. If this option is used, and the system
2386 does crash during the critical period, the same file must be passed to
2388 to restore the backup and reassemble the array. When shrinking rather
2389 than growing the array, the reshape is done from the end towards the
2390 beginning, so the "critical section" is at the end of the reshape.
2394 Changing the RAID level of any array happens instantaneously. However
2395 in the RAID5 to RAID6 case this requires a non-standard layout of the
2396 RAID6 data, and in the RAID6 to RAID5 case that non-standard layout is
2397 required before the change can be accomplished. So while the level
2398 change is instant, the accompanying layout change can take quite a
2401 is required. If the array is not simultaneously being grown or
2402 shrunk, so that the array size will remain the same - for example,
2403 reshaping a 3-drive RAID5 into a 4-drive RAID6 - the backup file will
2404 be used not just for a "cricital section" but throughout the reshape
2405 operation, as described below under LAYOUT CHANGES.
2407 .SS CHUNK-SIZE AND LAYOUT CHANGES
2409 Changing the chunk-size of layout without also changing the number of
2410 devices as the same time will involve re-writing all blocks in-place.
2411 To ensure against data loss in the case of a crash, a
2413 must be provided for these changes. Small sections of the array will
2414 be copied to the backup file while they are being rearranged. This
2415 means that all the data is copied twice, once to the backup and once
2416 to the new layout on the array, so this type of reshape will go very
2419 If the reshape is interrupted for any reason, this backup file must be
2421 .B "mdadm --assemble"
2422 so the array can be reassembled. Consequently the file cannot be
2423 stored on the device being reshaped.
2428 A write-intent bitmap can be added to, or removed from, an active
2429 array. Either internal bitmaps, or bitmaps stored in a separate file,
2430 can be added. Note that if you add a bitmap stored in a file which is
2431 in a filesystem that is on the RAID array being affected, the system
2432 will deadlock. The bitmap must be on a separate filesystem.
2434 .SH INCREMENTAL MODE
2438 .B mdadm \-\-incremental
2444 .B mdadm \-\-incremental \-\-fail
2448 .B mdadm \-\-incremental \-\-rebuild\-map
2451 .B mdadm \-\-incremental \-\-run \-\-scan
2454 This mode is designed to be used in conjunction with a device
2455 discovery system. As devices are found in a system, they can be
2457 .B "mdadm \-\-incremental"
2458 to be conditionally added to an appropriate array.
2460 Conversely, it can also be used with the
2462 flag to do just the opposite and find whatever array a particular device
2463 is part of and remove the device from that array.
2465 If the device passed is a
2467 device created by a previous call to
2469 then rather than trying to add that device to an array, all the arrays
2470 described by the metadata of the container will be started.
2473 performs a number of tests to determine if the device is part of an
2474 array, and which array it should be part of. If an appropriate array
2475 is found, or can be created,
2477 adds the device to the array and conditionally starts the array.
2481 will normally only add devices to an array which were previously working
2482 (active or spare) parts of that array. The support for automatic
2483 inclusion of a new drive as a spare in some array requires
2484 a configuration through POLICY in config file.
2488 makes are as follow:
2490 Is the device permitted by
2492 That is, is it listed in a
2494 line in that file. If
2496 is absent then the default it to allow any device. Similar if
2498 contains the special word
2500 then any device is allowed. Otherwise the device name given to
2502 must match one of the names or patterns in a
2507 Does the device have a valid md superblock? If a specific metadata
2508 version is requested with
2512 then only that style of metadata is accepted, otherwise
2514 finds any known version of metadata. If no
2516 metadata is found, the device may be still added to an array
2517 as a spare if POLICY allows.
2521 Does the metadata match an expected array?
2522 The metadata can match in two ways. Either there is an array listed
2525 which identifies the array (either by UUID, by name, by device list,
2526 or by minor-number), or the array was created with a
2532 or on the command line.
2535 is not able to positively identify the array as belonging to the
2536 current host, the device will be rejected.
2541 keeps a list of arrays that it has partially assembled in
2542 .B /var/run/mdadm/map
2544 .B /var/run/mdadm.map
2545 if the directory doesn't exist. Or maybe even
2546 .BR /dev/.mdadm.map ).
2547 If no array exists which matches
2548 the metadata on the new device,
2550 must choose a device name and unit number. It does this based on any
2553 or any name information stored in the metadata. If this name
2554 suggests a unit number, that number will be used, otherwise a free
2555 unit number will be chosen. Normally
2557 will prefer to create a partitionable array, however if the
2561 suggests that a non-partitionable array is preferred, that will be
2564 If the array is not found in the config file and its metadata does not
2565 identify it as belonging to the "homehost", then
2567 will choose a name for the array which is certain not to conflict with
2568 any array which does belong to this host. It does this be adding an
2569 underscore and a small number to the name preferred by the metadata.
2571 Once an appropriate array is found or created and the device is added,
2573 must decide if the array is ready to be started. It will
2574 normally compare the number of available (non-spare) devices to the
2575 number of devices that the metadata suggests need to be active. If
2576 there are at least that many, the array will be started. This means
2577 that if any devices are missing the array will not be restarted.
2583 in which case the array will be run as soon as there are enough
2584 devices present for the data to be accessible. For a RAID1, that
2585 means one device will start the array. For a clean RAID5, the array
2586 will be started as soon as all but one drive is present.
2588 Note that neither of these approaches is really ideal. If it can
2589 be known that all device discovery has completed, then
2593 can be run which will try to start all arrays that are being
2594 incrementally assembled. They are started in "read-auto" mode in
2595 which they are read-only until the first write request. This means
2596 that no metadata updates are made and no attempt at resync or recovery
2597 happens. Further devices that are found before the first write can
2598 still be added safely.
2601 This section describes environment variables that affect how mdadm
2606 Setting this value to 1 will prevent mdadm from automatically launching
2607 mdmon. This variable is intended primarily for debugging mdadm/mdmon.
2613 does not create any device nodes in /dev, but leaves that task to
2617 appears not to be configured, or if this environment variable is set
2620 will create and devices that are needed.
2624 .B " mdadm \-\-query /dev/name-of-device"
2626 This will find out if a given device is a RAID array, or is part of
2627 one, and will provide brief information about the device.
2629 .B " mdadm \-\-assemble \-\-scan"
2631 This will assemble and start all arrays listed in the standard config
2632 file. This command will typically go in a system startup file.
2634 .B " mdadm \-\-stop \-\-scan"
2636 This will shut down all arrays that can be shut down (i.e. are not
2637 currently in use). This will typically go in a system shutdown script.
2639 .B " mdadm \-\-follow \-\-scan \-\-delay=120"
2641 If (and only if) there is an Email address or program given in the
2642 standard config file, then
2643 monitor the status of all arrays listed in that file by
2644 polling them ever 2 minutes.
2646 .B " mdadm \-\-create /dev/md0 \-\-level=1 \-\-raid\-devices=2 /dev/hd[ac]1"
2648 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
2651 .B " echo 'DEVICE /dev/hd*[0\-9] /dev/sd*[0\-9]' > mdadm.conf"
2653 .B " mdadm \-\-detail \-\-scan >> mdadm.conf"
2655 This will create a prototype config file that describes currently
2656 active arrays that are known to be made from partitions of IDE or SCSI drives.
2657 This file should be reviewed before being used as it may
2658 contain unwanted detail.
2660 .B " echo 'DEVICE /dev/hd[a\-z] /dev/sd*[a\-z]' > mdadm.conf"
2662 .B " mdadm \-\-examine \-\-scan \-\-config=mdadm.conf >> mdadm.conf"
2664 This will find arrays which could be assembled from existing IDE and
2665 SCSI whole drives (not partitions), and store the information in the
2666 format of a config file.
2667 This file is very likely to contain unwanted detail, particularly
2670 entries. It should be reviewed and edited before being used as an
2673 .B " mdadm \-\-examine \-\-brief \-\-scan \-\-config=partitions"
2675 .B " mdadm \-Ebsc partitions"
2677 Create a list of devices by reading
2678 .BR /proc/partitions ,
2679 scan these for RAID superblocks, and printout a brief listing of all
2682 .B " mdadm \-Ac partitions \-m 0 /dev/md0"
2684 Scan all partitions and devices listed in
2685 .BR /proc/partitions
2688 out of all such devices with a RAID superblock with a minor number of 0.
2690 .B " mdadm \-\-monitor \-\-scan \-\-daemonise > /var/run/mdadm"
2692 If config file contains a mail address or alert program, run mdadm in
2693 the background in monitor mode monitoring all md devices. Also write
2694 pid of mdadm daemon to
2695 .BR /var/run/mdadm .
2697 .B " mdadm \-Iq /dev/somedevice"
2699 Try to incorporate newly discovered device into some array as
2702 .B " mdadm \-\-incremental \-\-rebuild\-map \-\-run \-\-scan"
2704 Rebuild the array map from any current arrays, and then start any that
2707 .B " mdadm /dev/md4 --fail detached --remove detached"
2709 Any devices which are components of /dev/md4 will be marked as faulty
2710 and then remove from the array.
2712 .B " mdadm --grow /dev/md4 --level=6 --backup-file=/root/backup-md4"
2716 which is currently a RAID5 array will be converted to RAID6. There
2717 should normally already be a spare drive attached to the array as a
2718 RAID6 needs one more drive than a matching RAID5.
2720 .B " mdadm --create /dev/md/ddf --metadata=ddf --raid-disks 6 /dev/sd[a-f]"
2722 Create a DDF array over 6 devices.
2724 .B " mdadm --create /dev/md/home -n3 -l5 -z 30000000 /dev/md/ddf"
2726 Create a RAID5 array over any 3 devices in the given DDF set. Use
2727 only 30 gigabytes of each device.
2729 .B " mdadm -A /dev/md/ddf1 /dev/sd[a-f]"
2731 Assemble a pre-exist ddf array.
2733 .B " mdadm -I /dev/md/ddf1"
2735 Assemble all arrays contained in the ddf array, assigning names as
2738 .B " mdadm \-\-create \-\-help"
2740 Provide help about the Create mode.
2742 .B " mdadm \-\-config \-\-help"
2744 Provide help about the format of the config file.
2746 .B " mdadm \-\-help"
2748 Provide general help.
2758 lists all active md devices with information about them.
2760 uses this to find arrays when
2762 is given in Misc mode, and to monitor array reconstruction
2767 The config file lists which devices may be scanned to see if
2768 they contain MD super block, and gives identifying information
2769 (e.g. UUID) about known MD arrays. See
2773 .SS /var/run/mdadm/map
2776 mode is used, this file gets a list of arrays currently being created.
2779 does not exist as a directory, then
2780 .B /var/run/mdadm.map
2783 is not available (as may be the case during early boot),
2785 is used on the basis that
2787 is usually available very early in boot.
2792 understand two sorts of names for array devices.
2794 The first is the so-called 'standard' format name, which matches the
2795 names used by the kernel and which appear in
2798 The second sort can be freely chosen, but must reside in
2800 When giving a device name to
2802 to create or assemble an array, either full path name such as
2806 can be given, or just the suffix of the second sort of name, such as
2812 chooses device names during auto-assembly or incremental assembly, it
2813 will sometimes add a small sequence number to the end of the name to
2814 avoid conflicted between multiple arrays that have the same name. If
2816 can reasonably determine that the array really is meant for this host,
2817 either by a hostname in the metadata, or by the presence of the array
2820 then it will leave off the suffix if possible.
2821 Also if the homehost is specified as
2824 will only use a suffix if a different array of the same name already
2825 exists or is listed in the config file.
2827 The standard names for non-partitioned arrays (the only sort of md
2828 array available in 2.4 and earlier) are of the form
2832 where NN is a number.
2833 The standard names for partitionable arrays (as available from 2.6
2834 onwards) are of the form
2838 Partition numbers should be indicated by added "pMM" to these, thus "/dev/md/d1p2".
2840 From kernel version, 2.6.28 the "non-partitioned array" can actually
2841 be partitioned. So the "md_dNN" names are no longer needed, and
2842 partitions such as "/dev/mdNNpXX" are possible.
2846 was previously known as
2850 is completely separate from the
2852 package, and does not use the
2854 configuration file at all.
2857 For further information on mdadm usage, MD and the various levels of
2860 .B http://raid.wiki.kernel.org/
2862 (based upon Jakob \(/Ostergaard's Software\-RAID.HOWTO)
2864 .\"for new releases of the RAID driver check out:
2867 .\".UR ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2868 .\"ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
2873 .\".UR http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2874 .\"http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
2877 The latest version of
2879 should always be available from
2881 .B http://www.kernel.org/pub/linux/utils/raid/mdadm/