4 mdadm \- manage MD devices
10 .BI mdadm " [mode] <raiddevice> [options] <component-devices>"
13 RAID devices are virtual devices created from two or more
14 real block devices. This allows multiple devices (typically disk
15 drives or partitions there-of) to be combined into a single device to
16 hold (for example) a single filesystem.
17 Some RAID levels include redundancy and so can survive some degree of
20 Linux Software RAID devices are implemented through the md (Multiple
21 Devices) device driver.
23 Currently, Linux supports
39 is not a Software RAID mechanism, but does involve
42 each device is a path to one common physical storage device.
45 is also not true RAID, and it only involves one device. It
46 provides a layer over a true device that can be used to inject faults.
49 '''is a program that can be used to create, manage, and monitor
51 '''such it provides a similar set of functionality to the
54 '''The key differences between
61 '''is a single program and not a collection of programs.
64 '''can perform (almost) all of its functions without having a
65 '''configuration file and does not use one by default. Also
67 '''helps with management of the configuration
71 '''can provide information about your arrays (through Query, Detail, and Examine)
81 '''configuration file, at all. It has a different configuration file
82 '''with a different format and a different purpose.
85 mdadm has 7 major modes of operation:
88 Assemble the parts of a previously created
89 array into an active array. Components can be explicitly given
90 or can be searched for.
92 checks that the components
93 do form a bona fide array, and can, on request, fiddle superblock
94 information so as to assemble a faulty array.
98 Build an array that doesn't have per-device superblocks. For these
101 cannot differentiate between initial creation and subsequent assembly
102 of an array. It also cannot perform any checks that appropriate
103 devices have been requested. Because of this, the
105 mode should only be used together with a complete understanding of
110 Create a new array with per-device superblocks.
112 '''in several step create-add-add-run or it can all happen with one command.
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 in RAID level 1/4/5/6 and changing the number of
126 active devices in RAID1.
130 This is for doing things to specific components of an array such as
131 adding new spares and removing faulty devices.
135 This is an 'everything else' mode that supports operations on active
136 arrays, operations on component devices such as erasing old superblocks, and
137 information gathering operations.
138 '''This mode allows operations on independent devices such as examine MD
139 '''superblocks, erasing old superblocks and stopping active arrays.
143 .SH Options for selecting a mode are:
146 .BR -A ", " --assemble
147 Assemble a pre-existing array.
151 Build a legacy array without superblocks.
158 .BR -F ", " --follow ", " --monitor
165 Change the size or shape of an active array.
167 If a device is given before any options, or if the first option is
172 then the MANAGE mode is assume.
173 Anything other than these will cause the
177 .SH Options that are not mode-specific are:
181 Display general help message or, after one of the above options, a
182 mode specific help message.
186 Display more detailed help about command line parsing and some commonly
190 .BR -V ", " --version
191 Print version information for mdadm.
194 .BR -v ", " --verbose
195 Be more verbose about what is happening. This can be used twice to be
197 The extra verbosity currently only affects
200 .BR "--examine --scan" .
204 Avoid printing purely informative messages. With this,
206 will be silent unless there is something really important to report.
210 Be less verbose. This is used with
218 gives an intermediate level of verbosity.
222 Be more forceful about certain operations. See the various modes of
223 the exact meaning of this option in different contexts.
226 .BR -c ", " --config=
227 Specify the config file. Default is to use
228 .BR /etc/mdadm.conf ,
229 or if that is missing, then
230 .BR /etc/mdadm/mdadm.conf .
231 If the config file given is
233 then nothing will be read, but
235 will act as though the config file contained exactly
236 .B "DEVICE partitions"
239 to find a list of devices to scan.
242 is given for the config file, then
244 will act as though the config file were empty.
250 for missing information.
251 In general, this option gives
253 permission to get any missing information, like component devices,
254 array devices, array identities, and alert destination from the
256 .BR /etc/mdadm.conf .
257 One exception is MISC mode when using
263 says to get a list of array devices from
267 .B -e ", " --metadata=
268 Declare the style of superblock (raid metadata) to be used. The
269 default is 0.90 for --create, and to guess for other operations.
273 .IP "0, 0.90, default"
274 Use the original 0.90 format superblock. This format limits arrays to
275 28 componenet devices and limits component devices of levels 1 and
276 greater to 2 terabytes.
277 .IP "1, 1.0, 1.1, 1.2"
278 Use the new version-1 format superblock. This has few restrictions.
279 The different subversion store the superblock at different locations
280 on the device, either at the end (for 1.0), at the start (for 1.1) or
281 4K from the start (for 1.2).
286 This will over-ride any
288 setting in the config file and provides the identify of the host which
289 should be considered the home for any arrays.
291 When creating an array, the
293 will be recorded in the superblock. For version-1 superblocks, it will
294 be prefixed to the array name. For version-0.90 superblocks part of
295 the SHA1 hash of the hostname will be stored in the later half of the
298 When reporting information about an array, any array which is tagged
299 for the given homehost will be reported as such.
301 When using Auto-Assemble, only arrays tagged for the given homehost
304 .SH For create, build, or grow:
307 .BR -n ", " --raid-devices=
308 Specify the number of active devices in the array. This, plus the
309 number of spare devices (see below) must equal the number of
311 (including "\fBmissing\fP" devices)
312 that are listed on the command line for
314 Setting a value of 1 is probably
315 a mistake and so requires that
317 be specified first. A value of 1 will then be allowed for linear,
318 multipath, raid0 and raid1. It is never allowed for raid4 or raid5.
320 This number can only be changed using
322 for RAID1 arrays, and only on kernels which provide necessary support.
325 .BR -x ", " --spare-devices=
326 Specify the number of spare (eXtra) devices in the initial array.
327 Spares can also be added
328 and removed later. The number of component devices listed
329 on the command line must equal the number of raid devices plus the
330 number of spare devices.
335 Amount (in Kibibytes) of space to use from each drive in RAID1/4/5/6.
336 This must be a multiple of the chunk size, and must leave about 128Kb
337 of space at the end of the drive for the RAID superblock.
338 If this is not specified
339 (as it normally is not) the smallest drive (or partition) sets the
340 size, though if there is a variance among the drives of greater than 1%, a warning is
343 This value can be set with
345 for RAID level 1/4/5/6. If the array was created with a size smaller
346 than the currently active drives, the extra space can be accessed
349 The size can be given as
351 which means to choose the largest size that fits on all current drives.
355 Specify chunk size of kibibytes. The default is 64.
359 Specify rounding factor for linear array (==chunk size)
363 Set raid level. When used with
365 options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
366 raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty. Obviously some of these are synonymous.
370 only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid.
372 Not yet supported with
376 .BR -p ", " --layout=
377 This option configures the fine details of data layout for raid5,
378 and raid10 arrays, and controls the failure modes for
381 The layout of the raid5 parity block can be one of
386 la, ra, ls, rs. The default is left-symmetric.
388 When setting the failure mode for
406 Each mode can be followed by a number which is used as a period
407 between fault generation. Without a number, the fault is generated
408 once on the first relevant request. With a number, the fault will be
409 generated after that many request, and will continue to be generated
410 every time the period elapses.
412 Multiple failure modes can be current simultaneously by using the
413 "--grow" option to set subsequent failure modes.
415 "clear" or "none" will remove any pending or periodic failure modes,
416 and "flush" will clear any persistent faults.
418 To set the parity with "--grow", the level of the array ("faulty")
419 must be specified before the fault mode is specified.
421 Finally, the layout options for RAID10 are one of 'n', 'o' or 'p' followed
422 by a small number. The default is 'n2'.
425 signals 'near' copies. Multiple copies of one data block are at
426 similar offsets in different devices.
429 signals 'offset' copies. Rather than the chunks being duplicated
430 within a stripe, whole stripes are duplicated but are rotated by one
431 device so duplicate blocks are on different devices. Thus subsequent
432 copies of a block are in the next drive, and are one chunk further
437 (multiple copies have very different offsets). See md(4) for more
438 detail about 'near' and 'far'.
440 The number is the number of copies of each datablock. 2 is normal, 3
441 can be useful. This number can be at most equal to the number of
442 devices in the array. It does not need to divide evenly into that
443 number (e.g. it is perfectly legal to have an 'n2' layout for an array
444 with an odd number of devices).
448 same as --layout (thus explaining the p of
452 .BR -b ", " --bitmap=
453 Specify a file to store a write-intent bitmap in. The file should not
454 exist unless --force is also given. The same file should be provided
455 when assembling the array. If the word
457 is given, then the bitmap is stored with the metadata on the array,
458 and so is replicated on all devices. If the word
462 mode, then any bitmap that is present is removed.
464 To help catch typing errors, the filename must contain at least one
465 slash ('/') if it is a real file (not 'internal' or 'none').
467 Note: external bitmaps are only known to work on ext2 and ext3.
468 Storing bitmap files on other filesystems may result in serious problems.
472 Set the chunksize of the bitmap. Each bit corresponds to that many
473 Kilobytes of storage.
474 When using a file based bitmap, the default is to use the smallest
475 size that is atleast 4 and requires no more than 2^21 chunks.
478 bitmap, the chunksize is automatically determined to make best use of
483 .BR -W ", " --write-mostly
484 subsequent devices lists in a
489 command will be flagged as 'write-mostly'. This is valid for RAID1
490 only and means that the 'md' driver will avoid reading from these
491 devices if at all possible. This can be useful if mirroring over a
496 Specify that write-behind mode should be enabled (valid for RAID1
497 only). If an argument is specified, it will set the maximum number
498 of outstanding writes allowed. The default value is 256.
499 A write-intent bitmap is required in order to use write-behind
500 mode, and write-behind is only attempted on drives marked as
507 that the array pre-existed and is known to be clean. It can be useful
508 when trying to recover from a major failure as you can be sure that no
509 data will be affected unless you actually write to the array. It can
510 also be used when creating a RAID1 or RAID10 if you want to avoid the
511 initial resync, however this practice - while normally safe - is not
512 recommended. Use this ony if you really know what you are doing.
516 This is needed when --grow is used to increase the number of
517 raid-devices in a RAID5 if there are no spare devices available.
518 See the section below on RAID_DEVICE CHANGES. The file should be
519 stored on a separate device, not on the raid array being reshaped.
525 for the array. This is currently only effective when creating an
526 array with a version-1 superblock. The name is a simple textual
527 string that can be used to identify array components when assembling.
533 run the array, even if some of the components
534 appear to be active in another array or filesystem. Normally
536 will ask for confirmation before including such components in an
537 array. This option causes that question to be suppressed.
543 accept the geometry and layout specified without question. Normally
545 will not allow creation of an array with only one device, and will try
546 to create a raid5 array with one missing drive (as this makes the
547 initial resync work faster). With
550 will not try to be so clever.
553 .BR -a ", " "--auto{=no,yes,md,mdp,part,p}{NN}"
554 Instruct mdadm to create the device file if needed, possibly allocating
555 an unused minor number. "md" causes a non-partitionable array
556 to be used. "mdp", "part" or "p" causes a partitionable array (2.6 and
557 later) to be used. "yes" requires the named md device to have
558 a 'standard' format, and the type and minor number will be determined
559 from this. See DEVICE NAMES below.
561 The argument can also come immediately after
566 is also given, then any
568 entries in the config file will over-ride the
570 instruction given on the command line.
572 For partitionable arrays,
574 will create the device file for the whole array and for the first 4
575 partitions. A different number of partitions can be specified at the
576 end of this option (e.g.
578 If the device name ends with a digit, the partition names add a 'p',
579 and a number, e.g. "/dev/home1p3". If there is no
580 trailing digit, then the partition names just have a number added,
581 e.g. "/dev/scratch3".
583 If the md device name is in a 'standard' format as described in DEVICE
584 NAMES, then it will be created, if necessary, with the appropriate
585 number based on that name. If the device name is not in one of these
586 formats, then a unused minor number will be allocated. The minor
587 number will be considered unused if there is no active array for that
588 number, and there is no entry in /dev for that number and with a
595 uuid of array to assemble. Devices which don't have this uuid are
599 .BR -m ", " --super-minor=
600 Minor number of device that array was created for. Devices which
601 don't have this minor number are excluded. If you create an array as
602 /dev/md1, then all superblocks will contain the minor number 1, even if
603 the array is later assembled as /dev/md2.
605 Giving the literal word "dev" for
609 to use the minor number of the md device that is being assembled.
613 will look for super blocks with a minor number of 0.
617 Specify the name of the array to assemble. This must be the name
618 that was specified when creating the array. It must either match
619 then name stored in the superblock exactly, or it must match
622 is added to the start of the given name.
626 Assemble the array even if some superblocks appear out-of-date
630 Attempt to start the array even if fewer drives were given than were
631 present last time the array was active. Normally if not all the
632 expected drives are found and
634 is not used, then the array will be assembled but not started.
637 an attempt will be made to start it anyway.
641 This is the reverse of
643 in that it inhibits the started if array unless all expected drives
644 are present. This is only needed with
646 and can be used if you physical connections to devices are
647 not as reliable as you would like.
650 .BR -a ", " "--auto{=no,yes,md,mdp,part}"
651 See this option under Create and Build options.
654 .BR -b ", " --bitmap=
655 Specify the bitmap file that was given when the array was created. If
658 bitmap, there is no need to specify this when assembling the array.
664 was used to grow the number of raid-devices in a RAID5, and the system
665 crashed during the critical section, then the same
667 must be presented to --assemble to allow possibly corrupted data to be
671 .BR -U ", " --update=
672 Update the superblock on each device while assembling the array. The
673 argument given to this flag can be one of
686 option will adjust the superblock of an array what was created on a Sparc
687 machine running a patched 2.2 Linux kernel. This kernel got the
688 alignment of part of the superblock wrong. You can use the
689 .B "--examine --sparc2.2"
692 to see what effect this would have.
696 option will update the
698 field on each superblock to match the minor number of the array being
699 assembled. This is not needed on 2.6 and later kernels as they make
700 this adjustment automatically.
704 option will change the uuid of the array. If a UUID is given with the
705 "--uuid" option that UUID will be used as a new UUID and with
707 be used to help identify the devices in the array.
708 If no "--uuid" is given, a random uuid is chosen.
712 option will change the
714 of the array as stored in the superblock. This is only supported for
715 version-1 superblocks.
719 option will change the
721 as recorded in the superblock. For version-0 superblocks, this is the
722 same as updating the UUID.
723 For version-1 superblocks, this involves updating the name.
727 option will cause the array to be marked
729 meaning that any redundancy in the array (e.g. parity for raid5,
730 copies for raid1) may be incorrect. This will cause the raid system
731 to perform a "resync" pass to make sure that all redundant information
736 option allows arrays to be moved between machines with different
738 When assembling such an array for the first time after a move, giving
739 .B "--update=byteorder"
742 to expect superblocks to have their byteorder reversed, and will
743 correct that order before assembling the array. This is only valid
744 with original (Version 0.90) superblocks.
748 option will correct the summaries in the superblock. That is the
749 counts of total, working, active, failed, and spare devices.
752 .B --auto-update-homehost
753 This flag is only meaning with auto-assembly (see discussion below).
754 In that situation, if no suitable arrays are found for this homehost,
756 will recan for any arrays at all and will assemble them and update the
757 homehost to match the current host.
763 hot-add listed devices.
767 re-add a device that was recently removed from an array.
771 remove listed devices. They must not be active. i.e. they should
772 be failed or spare devices.
776 mark listed devices as faulty.
783 Each of these options require that the first device list is the array
784 to be acted upon and the remainder are component devices to be added,
785 removed, or marked as fault. Several different operations can be
786 specified for different devices, e.g.
788 mdadm /dev/md0 --add /dev/sda1 --fail /dev/sdb1 --remove /dev/sdb1
790 Each operation applies to all devices listed until the next
793 If an array is using a write-intent bitmap, then devices which have
794 been removed can be re-added in a way that avoids a full
795 reconstruction but instead just updated the blocks that have changed
796 since the device was removed. For arrays with persistent metadata
797 (superblocks) this is done automatically. For arrays created with
799 mdadm needs to be told that this device we removed recently with
802 Devices can only be removed from an array if they are not in active
803 use. i.e. that must be spares or failed devices. To remove an active
804 device, it must be marked as
812 Examine a device to see
813 (1) if it is an md device and (2) if it is a component of an md
815 Information about what is discovered is presented.
819 Print detail of one or more md devices.
822 .BR -E ", " --examine
823 Print content of md superblock on device(s).
826 If an array was created on a 2.2 Linux kernel patched with RAID
827 support, the superblock will have been created incorrectly, or at
828 least incompatibly with 2.4 and later kernels. Using the
832 will fix the superblock before displaying it. If this appears to do
833 the right thing, then the array can be successfully assembled using
834 .BR "--assemble --update=sparc2.2" .
837 .BR -X ", " --examine-bitmap
838 Report information about a bitmap file.
842 start a partially built array.
846 deactivate array, releasing all resources.
849 .BR -o ", " --readonly
850 mark array as readonly.
853 .BR -w ", " --readwrite
854 mark array as readwrite.
858 If the device contains a valid md superblock, the block is
859 over-written with zeros. With
861 the block where the superblock would be is over-written even if it
862 doesn't appear to be valid.
870 is set to reflect the status of the device.
872 .SH For Monitor mode:
875 Give a mail address to send alerts to.
878 .BR -p ", " --program ", " --alert
879 Give a program to be run whenever an event is detected.
883 Cause all events to be reported through 'syslog'. The messages have
884 facility of 'daemon' and varying priorities.
888 Give a delay in seconds.
890 polls the md arrays and then waits this many seconds before polling
891 again. The default is 60 seconds.
894 .BR -f ", " --daemonise
897 to run as a background daemon if it decides to monitor anything. This
898 causes it to fork and run in the child, and to disconnect form the
899 terminal. The process id of the child is written to stdout.
902 which will only continue monitoring if a mail address or alert program
903 is found in the config file.
906 .BR -i ", " --pid-file
909 is running in daemon mode, write the pid of the daemon process to
910 the specified file, instead of printing it on standard output.
913 .BR -1 ", " --oneshot
914 Check arrays only once. This will generate
916 events and more significantly
922 .B " mdadm --monitor --scan -1"
924 from a cron script will ensure regular notification of any degraded arrays.
930 alert for every array found at startup. This alert gets mailed and
931 passed to the alert program. This can be used for testing that alert
932 message do get through successfully.
939 .I md-device options-and-component-devices...
942 .B mdadm --assemble --scan
943 .I md-devices-and-options...
946 .B mdadm --assemble --scan
950 This usage assembles one or more raid arrays from pre-existing components.
951 For each array, mdadm needs to know the md device, the identity of the
952 array, and a number of component-devices. These can be found in a number of ways.
954 In the first usage example (without the
956 the first device given is the md device.
957 In the second usage example, all devices listed are treated as md
958 devices and assembly is attempted.
959 In the third (where no devices are listed) all md devices that are
960 listed in the configuration file are assembled.
962 If precisely one device is listed, but
968 was given and identify information is extracted from the configuration file.
970 The identity can be given with the
974 option, can be found in the config file, or will be taken from the
975 super block on the first component-device listed on the command line.
977 Devices can be given on the
979 command line or in the config file. Only devices which have an md
980 superblock which contains the right identity will be considered for
983 The config file is only used if explicitly named with
985 or requested with (a possibly implicit)
993 is not given, then the config file will only be used to find the
994 identity of md arrays.
996 Normally the array will be started after it is assembled. However if
998 is not given and insufficient drives were listed to start a complete
999 (non-degraded) array, then the array is not started (to guard against
1000 usage errors). To insist that the array be started in this case (as
1001 may work for RAID1, 4, 5, 6, or 10), give the
1007 option is given, either on the command line (--auto) or in the
1008 configuration file (e.g. auto=part), then
1010 will create the md device if necessary or will re-create it if it
1011 doesn't look usable as it is.
1013 This can be useful for handling partitioned devices (which don't have
1014 a stable device number - it can change after a reboot) and when using
1015 "udev" to manage your
1017 tree (udev cannot handle md devices because of the unusual device
1018 initialisation conventions).
1020 If the option to "auto" is "mdp" or "part" or (on the command line
1021 only) "p", then mdadm will create a partitionable array, using the
1022 first free one that is not in use, and does not already have an entry
1023 in /dev (apart from numeric /dev/md* entries).
1025 If the option to "auto" is "yes" or "md" or (on the command line)
1026 nothing, then mdadm will create a traditional, non-partitionable md
1029 It is expected that the "auto" functionality will be used to create
1030 device entries with meaningful names such as "/dev/md/home" or
1031 "/dev/md/root", rather than names based on the numerical array number.
1033 When using this option to create a partitionable array, the device
1034 files for the first 4 partitions are also created. If a different
1035 number is required it can be simply appended to the auto option.
1036 e.g. "auto=part8". Partition names are created by appending a digit
1037 string to the device name, with an intervening "p" if the device name
1042 option is also available in Build and Create modes. As those modes do
1043 not use a config file, the "auto=" config option does not apply to
1051 and no devices are listed,
1053 will first attempt to assemble all the arrays listed in the config
1058 has been specified (either in the config file or on the command line),
1060 will look further for possible arrays and will try to assemble
1061 anything that it finds which is tagged as belonging to the given
1062 homehost. This is the only situation where
1064 will assemble arrays without being given specific device name or
1065 identify information for the array.
1069 finds a consistent set of devices that look like they should comprise
1070 an array, and if the superblock is tagged as belonging to the given
1071 home host, it will automatically choose a device name and try to
1072 assemble the array. If the array uses version-0.90 metadata, then the
1074 number as recorded in the superblock is used to create a name in
1078 If the array uses version-1 metadata, then the
1080 from the superblock is used to similarly create a name in
1082 The name will have any 'host' prefix stripped first.
1086 cannot find any array for the given host at all, and if
1087 .B --auto-update-homehost
1090 will search again for any array (not just an array created for this
1091 host) and will assemble each assuming
1092 .IR --update=homehost .
1093 This will change the host tag in the superblock so that on the next run,
1094 these arrays will be found without the second pass. The intention of
1095 this feature is to support transitioning a set of md arrays to using
1098 The reason for requiring arrays to be tagged with the homehost for
1099 auto assembly is to guard against problems that can arise when moving
1100 devices from one host to another.
1110 .BI --raid-devices= Z
1114 This usage is similar to
1116 The difference is that it creates an array without a superblock. With
1117 these arrays there is no difference between initially creating the array and
1118 subsequently assembling the array, except that hopefully there is useful
1119 data there in the second case.
1121 The level may raid0, linear, multipath, or faulty, or one of their
1122 synonyms. All devices must be listed and the array will be started
1134 .BI --raid-devices= Z
1138 This usage will initialise a new md array, associate some devices with
1139 it, and activate the array.
1143 option is given (as described in more detail in the section on
1144 Assemble mode), then the md device will be created with a suitable
1145 device number if necessary.
1147 As devices are added, they are checked to see if they contain raid
1148 superblocks or filesystems. They are also checked to see if the variance in
1149 device size exceeds 1%.
1151 If any discrepancy is found, the array will not automatically be run, though
1154 can override this caution.
1156 To create a "degraded" array in which some devices are missing, simply
1157 give the word "\fBmissing\fP"
1158 in place of a device name. This will cause
1160 to leave the corresponding slot in the array empty.
1161 For a RAID4 or RAID5 array at most one slot can be
1162 "\fBmissing\fP"; for a RAID6 array at most two slots.
1163 For a RAID1 array, only one real device needs to be given. All of the
1167 When creating a RAID5 array,
1169 will automatically create a degraded array with an extra spare drive.
1170 This is because building the spare into a degraded array is in general faster than resyncing
1171 the parity on a non-degraded, but not clean, array. This feature can
1172 be over-ridden with the
1176 When creating an array with version-1 metadata a name for the host is
1178 If this is not given with the
1182 will chose a name based on the last component of the name of the
1183 device being created. So if
1185 is being created, then the name
1190 is being created, then the name
1196 '''option is given, it is not necessary to list any component-devices in this command.
1197 '''They can be added later, before a
1201 '''is given, the apparent size of the smallest drive given is used.
1203 The General Management options that are valid with --create are:
1206 insist on running the array even if some devices look like they might
1211 start the array readonly - not supported yet.
1219 .I options... devices...
1222 This usage will allow individual devices in an array to be failed,
1223 removed or added. It is possible to perform multiple operations with
1224 on command. For example:
1226 .B " mdadm /dev/md0 -f /dev/hda1 -r /dev/hda1 -a /dev/hda1"
1232 and will then remove it from the array and finally add it back
1233 in as a spare. However only one md array can be affected by a single
1244 MISC mode includes a number of distinct operations that
1245 operate on distinct devices. The operations are:
1248 The device is examined to see if it is
1249 (1) an active md array, or
1250 (2) a component of an md array.
1251 The information discovered is reported.
1255 The device should be an active md device.
1257 will display a detailed description of the array.
1261 will cause the output to be less detailed and the format to be
1262 suitable for inclusion in
1263 .BR /etc/mdadm.conf .
1266 will normally be 0 unless
1268 failed to get useful information about the device(s). However if the
1270 option is given, then the exit status will be:
1274 The array is functioning normally.
1277 The array has at least one failed device.
1280 The array has multiple failed devices and hence is unusable (raid4 or
1284 There was an error while trying to get information about the device.
1289 The device should be a component of an md array.
1291 will read the md superblock of the device and display the contents.
1296 then multiple devices that are components of the one array
1297 are grouped together and reported in a single entry suitable
1299 .BR /etc/mdadm.conf .
1303 without listing any devices will cause all devices listed in the
1304 config file to be examined.
1308 The devices should be active md arrays which will be deactivated, as
1309 long as they are not currently in use.
1313 This will fully activate a partially assembled md array.
1317 This will mark an active array as read-only, providing that it is
1318 not currently being used.
1324 array back to being read/write.
1328 For all operations except
1331 will cause the operation to be applied to all arrays listed in
1336 causes all devices listed in the config file to be examined.
1344 .I options... devices...
1349 to periodically poll a number of md arrays and to report on any events
1352 will never exit once it decides that there are arrays to be checked,
1353 so it should normally be run in the background.
1355 As well as reporting events,
1357 may move a spare drive from one array to another if they are in the
1360 and if the destination array has a failed drive but no spares.
1362 If any devices are listed on the command line,
1364 will only monitor those devices. Otherwise all arrays listed in the
1365 configuration file will be monitored. Further, if
1367 is given, then any other md devices that appear in
1369 will also be monitored.
1371 The result of monitoring the arrays is the generation of events.
1372 These events are passed to a separate program (if specified) and may
1373 be mailed to a given E-mail address.
1375 When passing event to program, the program is run once for each event
1376 and is given 2 or 3 command-line arguments. The first is the
1377 name of the event (see below). The second is the name of the
1378 md device which is affected, and the third is the name of a related
1379 device if relevant, such as a component device that has failed.
1383 is given, then a program or an E-mail address must be specified on the
1384 command line or in the config file. If neither are available, then
1386 will not monitor anything.
1390 will continue monitoring as long as something was found to monitor. If
1391 no program or email is given, then each event is reported to
1394 The different events are:
1398 .B DeviceDisappeared
1399 An md array which previously was configured appears to no longer be
1400 configured. (syslog priority: Critical)
1404 was told to monitor an array which is RAID0 or Linear, then it will
1406 .B DeviceDisappeared
1407 with the extra information
1409 This is because RAID0 and Linear do not support the device-failed,
1410 hot-spare and resync operations which are monitored.
1414 An md array started reconstruction. (syslog priority: Warning)
1420 is 20, 40, 60, or 80, this indicates that rebuild has passed that many
1421 percentage of the total. (syslog priority: Warning)
1425 An md array that was rebuilding, isn't any more, either because it
1426 finished normally or was aborted. (syslog priority: Warning)
1430 An active component device of an array has been marked as
1431 faulty. (syslog priority: Critical)
1435 A spare component device which was being rebuilt to replace a faulty
1436 device has failed. (syslog priority: Critial)
1440 A spare component device which was being rebuilt to replace a faulty
1441 device has been successfully rebuilt and has been made active.
1442 (syslog priority: Info)
1446 A new md array has been detected in the
1448 file. (syslog priority: Info)
1452 A newly noticed array appears to be degraded. This message is not
1455 notices a drive failure which causes degradation, but only when
1457 notices that an array is degraded when it first sees the array.
1458 (syslog priority: Critial)
1462 A spare drive has been moved from one array in a
1464 to another to allow a failed drive to be replaced.
1465 (syslog priority: Info)
1471 has been told, via the config file, that an array should have a certain
1472 number of spare devices, and
1474 detects that it has fewer that this number when it first sees the
1475 array, it will report a
1478 (syslog priority: Warning)
1482 An array was found at startup, and the
1485 (syslog priority: Info)
1495 cause Email to be sent. All events cause the program to be run.
1496 The program is run with two or three arguments, they being the event
1497 name, the array device and possibly a second device.
1499 Each event has an associated array device (e.g.
1501 and possibly a second device. For
1506 the second device is the relevant component device.
1509 the second device is the array that the spare was moved from.
1513 to move spares from one array to another, the different arrays need to
1514 be labelled with the same
1516 in the configuration file. The
1518 name can be any string. It is only necessary that different spare
1519 groups use different names.
1523 detects that an array which is in a spare group has fewer active
1524 devices than necessary for the complete array, and has no spare
1525 devices, it will look for another array in the same spare group that
1526 has a full complement of working drive and a spare. It will then
1527 attempt to remove the spare from the second drive and add it to the
1529 If the removal succeeds but the adding fails, then it is added back to
1533 The GROW mode is used for changing the size or shape of an active
1535 For this to work, the kernel must support the necessary change.
1536 Various types of growth are being added during 2.6 development,
1537 including restructuring a raid5 array to have more active devices.
1539 Currently the only support available is to
1541 change the "size" attribute
1542 for RAID1, RAID5 and RAID6.
1544 increase the "raid-disks" attribute of RAID1 and RAID5.
1546 add a write-intent bitmap to any array which support these bitmaps, or
1547 remove a write-intent bitmap from such an array.
1551 Normally when an array is built the "size" it taken from the smallest
1552 of the drives. If all the small drives in an arrays are, one at a
1553 time, removed and replaced with larger drives, then you could have an
1554 array of large drives with only a small amount used. In this
1555 situation, changing the "size" with "GROW" mode will allow the extra
1556 space to start being used. If the size is increased in this way, a
1557 "resync" process will start to make sure the new parts of the array
1560 Note that when an array changes size, any filesystem that may be
1561 stored in the array will not automatically grow to use the space. The
1562 filesystem will need to be explicitly told to use the extra space.
1564 .SS RAID-DEVICES CHANGES
1566 A RAID1 array can work with any number of devices from 1 upwards
1567 (though 1 is not very useful). There may be times which you want to
1568 increase or decrease the number of active devices. Note that this is
1569 different to hot-add or hot-remove which changes the number of
1572 When reducing the number of devices in a RAID1 array, the slots which
1573 are to be removed from the array must already be vacant. That is, the
1574 devices that which were in those slots must be failed and removed.
1576 When the number of devices is increased, any hot spares that are
1577 present will be activated immediately.
1579 Increasing the number of active devices in a RAID5 is much more
1580 effort. Every block in the array will need to be read and written
1581 back to a new location. From 2.6.17, the Linux Kernel is able to do
1582 this safely, including restart and interrupted "reshape".
1584 When relocating the first few stripes on a raid5, it is not possible
1585 to keep the data on disk completely consistent and crash-proof. To
1586 provide the required safety, mdadm disables writes to the array while
1587 this "critical section" is reshaped, and takes a backup of the data
1588 that is in that section. This backup is normally stored in any spare
1589 devices that the array has, however it can also be stored in a
1590 separate file specified with the
1592 option. If this option is used, and the system does crash during the
1593 critical period, the same file must be passed to
1595 to restore the backup and reassemble the array.
1599 A write-intent bitmap can be added to, or removed from, an active
1600 array. Either internal bitmaps, or bitmaps stored in a separate file
1601 can be added. Note that if you add a bitmap stored in a file which is
1602 in a filesystem that is on the raid array being affected, the system
1603 will deadlock. The bitmap must be on a separate filesystem.
1607 .B " mdadm --query /dev/name-of-device"
1609 This will find out if a given device is a raid array, or is part of
1610 one, and will provide brief information about the device.
1612 .B " mdadm --assemble --scan"
1614 This will assemble and start all arrays listed in the standard config file
1615 file. This command will typically go in a system startup file.
1617 .B " mdadm --stop --scan"
1619 This will shut down all array that can be shut down (i.e. are not
1620 currently in use). This will typically go in a system shutdown script.
1622 .B " mdadm --follow --scan --delay=120"
1624 If (and only if) there is an Email address or program given in the
1625 standard config file, then
1626 monitor the status of all arrays listed in that file by
1627 polling them ever 2 minutes.
1629 .B " mdadm --create /dev/md0 --level=1 --raid-devices=2 /dev/hd[ac]1"
1631 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
1634 .B " echo 'DEVICE /dev/hd*[0-9] /dev/sd*[0-9]' > mdadm.conf"
1636 .B " mdadm --detail --scan >> mdadm.conf"
1638 This will create a prototype config file that describes currently
1639 active arrays that are known to be made from partitions of IDE or SCSI drives.
1640 This file should be reviewed before being used as it may
1641 contain unwanted detail.
1643 .B " echo 'DEVICE /dev/hd[a-z] /dev/sd*[a-z]' > mdadm.conf"
1645 .B " mdadm --examine --scan --config=mdadm.conf >> mdadm.conf"
1647 This will find what arrays could be assembled from existing IDE and
1648 SCSI whole drives (not partitions) and store the information is the
1649 format of a config file.
1650 This file is very likely to contain unwanted detail, particularly
1653 entries. It should be reviewed and edited before being used as an
1656 .B " mdadm --examine --brief --scan --config=partitions"
1658 .B " mdadm -Ebsc partitions"
1660 Create a list of devices by reading
1661 .BR /proc/partitions ,
1662 scan these for RAID superblocks, and printout a brief listing of all
1665 .B " mdadm -Ac partitions -m 0 /dev/md0"
1667 Scan all partitions and devices listed in
1668 .BR /proc/partitions
1671 out of all such devices with a RAID superblock with a minor number of 0.
1673 .B " mdadm --monitor --scan --daemonise > /var/run/mdadm"
1675 If config file contains a mail address or alert program, run mdadm in
1676 the background in monitor mode monitoring all md devices. Also write
1677 pid of mdadm daemon to
1678 .BR /var/run/mdadm .
1680 .B " mdadm --create --help"
1682 Provide help about the Create mode.
1684 .B " mdadm --config --help"
1686 Provide help about the format of the config file.
1690 Provide general help.
1701 lists all active md devices with information about them.
1703 uses this to find arrays when
1705 is given in Misc mode, and to monitor array reconstruction
1711 The config file lists which devices may be scanned to see if
1712 they contain MD super block, and gives identifying information
1713 (e.g. UUID) about known MD arrays. See
1719 While entries in the /dev directory can have any format you like,
1721 has an understanding of 'standard' formats which it uses to guide its
1722 behaviour when creating device files via the
1726 The standard names for non-partitioned arrays (the only sort of md
1727 array available in 2.4 and earlier) either of
1733 where NN is a number.
1734 The standard names for partitionable arrays (as available from 2.6
1741 Partition numbers should be indicated by added "pMM" to these, thus "/dev/md/d1p2".
1745 was previously known as
1749 is completely separate from the
1751 package, and does not use the
1753 configuration file at all.
1756 For information on the various levels of
1760 .UR http://ostenfeld.dk/~jakob/Software-RAID.HOWTO/
1761 http://ostenfeld.dk/~jakob/Software-RAID.HOWTO/
1764 '''for new releases of the RAID driver check out:
1767 '''.UR ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
1768 '''ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
1773 '''.UR http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
1774 '''http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
1777 The latest version of
1779 should always be available from
1781 .UR http://www.kernel.org/pub/linux/utils/raid/mdadm/
1782 http://www.kernel.org/pub/linux/utils/raid/mdadm/