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
37 .B MULTIPATH is not a Software RAID mechanism, but does involve
40 each device is a path to one common physical storage device.
42 .B FAULTY is also no true RAID, and it only involves one device. It
43 provides a layer over a true device that can be used to inject faults.
46 is a program that can be used to create, manage, and monitor
48 such it provides a similar set of functionality to the
51 The key differences between
58 is a single program and not a collection of programs.
61 can perform (almost) all of its functions without having a
62 configuration file and does not use one by default. Also
64 helps with management of the configuration
68 can provide information about your arrays (through Query, Detail, and Examine)
78 configuration file, at all. It has a different configuration file
79 with a different format and an different purpose.
82 mdadm has 7 major modes of operation:
85 Assemble the parts of a previously created
86 array into an active array. Components can be explicitly given
87 or can be searched for.
89 checks that the components
90 do form a bona fide array, and can, on request, fiddle superblock
91 information so as to assemble a faulty array.
95 Build a legacy array without per-device superblocks.
99 Create a new array with per-device superblocks.
101 '''in several step create-add-add-run or it can all happen with one command.
105 This is for doing things to specific components of an array such as
106 adding new spares and removing faulty devices.
110 This mode allows operations on independent devices such as examine MD
111 superblocks, erasing old superblocks and stopping active arrays.
114 .B "Follow or Monitor"
115 Monitor one or more md devices and act on any state changes. This is
116 only meaningful for raid1, 4, 5, 6 or multipath arrays as
117 only these have interesting state. raid0 or linear never have
118 missing, spare, or failed drives, so there is nothing to monitor.
122 Grow (or shrink) an array, or otherwise reshape it in some way.
123 Currently supported growth options including changing the active size
124 of componenet devices in RAID level 1/4/5/6 and changing the number of
125 active devices in RAID1.
129 Available options are:
132 .BR -A ", " --assemble
133 Assemble a pre-existing array.
137 Build a legacy array without superblocks.
145 Examine a device to see
146 (1) if it is an md device and (2) if it is a component of an md
148 Information about what is discovered is presented.
152 Print detail of one or more md devices.
155 .BR -E ", " --examine
156 Print content of md superblock on device(s).
159 .BR -F ", " --follow ", " --monitor
166 Change the size or shape of an active array.
170 Display help message or, after above option, mode specific help
175 Display more detailed help about command line parsing and some commonly
179 .BR -V ", " --version
180 Print version information for mdadm.
183 .BR -v ", " --verbose
184 Be more verbose about what is happening.
188 Be less verbose. This is used with
195 Be more forceful about certain operations. See the various modes of
196 the exact meaning of this option in different contexts.
199 .BR -c ", " --config=
200 Specify the config file. Default is
201 .BR /etc/mdadm.conf .
202 If the config file given is
204 then nothing will be read, but
206 will act as though the config file contained exactly
207 .B "DEVICE partitions"
210 to find a list of devices to scan.
213 is given for the config file, then
215 will act as though the config file were empty.
221 for missing information.
222 In general, this option gives
224 permission to get any missing information, like component devices,
225 array devices, array identities, and alert destination from the
227 .BR /etc/mdadm.conf .
228 One exception is MISC mode when using
234 says to get a list of array devices from
237 .SH For create or build:
241 Specify chunk size of kibibytes. The default is 64.
245 Specify rounding factor for linear array (==chunk size)
249 Set raid level. When used with
251 options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
252 raid5, 5, raid6, 6, multipath, mp, fautly. Obviously some of these are synonymous.
256 only linear, raid0, 0, stripe are valid.
259 .BR -p ", " --parity=
260 Set raid5 parity algorithm. Options are:
265 la, ra, ls, rs. The default is left-symmetric.
267 This option is also used to set the failure mode for
285 Each mode can be followed by a number which is used as a period
286 between fault generation. Without a number, the fault is generated
287 once on the first relevant request. With a number, the fault will be
288 generated after that many request, and will continue to be generated
289 every time the period elapses.
291 Multiple failure modes can be current simultaneously by using the
292 "--grow" option to set subsequent failure modes.
294 "clear" or "none" will remove any pending or periodic failure modes,
295 and "flush" will clear any persistant faults.
297 To set the parity with "--grow", the level of the array ("faulty")
298 must be specified before the fault mode is specified.
305 .BR -n ", " --raid-devices=
306 Specify the number of active devices in the array. This, plus the
307 number of spare devices (see below) must equal the number of
309 (including "\fBmissing\fP" devices)
310 that are listed on the command line for
312 Setting a value of 1 is probably
313 a mistake and so requires that
315 be specified first. A value of 1 will then be allowed for linear,
316 multipath, raid0 and raid1. It is never allowed for raid4 or raid5.
318 This number can only be changed using
320 for RAID1 arrays, and only on kernels which provide necessary support.
323 .BR -x ", " --spare-devices=
324 Specify the number of spare (eXtra) devices in the initial array.
325 Spares can also be added
326 and removed later. The number of component devices listed
327 on the command line must equal the number of raid devices plus the
328 number of spare devices.
333 Amount (in Kibibytes) of space to use from each drive in RAID1/4/5/6.
334 This must be a multiple of the chunk size, and must leave about 128Kb
335 of space at the end of the drive for the RAID superblock.
336 If this is not specified
337 (as it normally is not) the smallest drive (or partition) sets the
338 size, though if there is a variance among the drives of greater than 1%, a warning is
341 This value can be set with
343 for RAID level 1/4/5/6. If the array was created with a size smaller
344 than the currently active drives, the extra space can be accessed
352 that the array pre-existed and is known to be clean. This is only
353 really useful for Building RAID1 array. Only use this if you really
354 know what you are doing. This is currently only supported for --build.
360 run the array, even if some of the components
361 appear to be active in another array or filesystem. Normally
363 will ask for confirmation before including such components in an
364 array. This option causes that question to be suppressed.
370 accept the geometry and layout specified without question. Normally
372 will not allow creation of an array with only one device, and will try
373 to create a raid5 array with one missing drive (as this makes the
374 initial resync work faster). With
377 will not try to be so clever.
380 .BR -a ", " "--auto{=no,yes,md,mdp,part,p}{NN}"
381 Instruct mdadm to create the device file if needed, and to allocate
382 an unused minor number. "yes" or "md" causes a non-partitionable array
383 to be used. "mdp", "part" or "p" causes a partitionable array (2.6 and
384 later) to be used. The argumentment can also come immediately after
387 For partitionable arrays,
389 will create the device file for the whole array and for the first 4
390 partitions. A different number of partitions can be specified at the
391 end of this option (e.g.
393 If the device name ends with a digit, the partition names add an
394 underscore, a 'p', and a number, e.g. "/dev/home1_p3". If there is no
395 trailing digit, then the partition names just have a number added,
396 e.g. "/dev/scratch3".
402 uuid of array to assemble. Devices which don't have this uuid are
406 .BR -m ", " --super-minor=
407 Minor number of device that array was created for. Devices which
408 don't have this minor number are excluded. If you create an array as
409 /dev/md1, then all superblocks will contain the minor number 1, even if
410 the array is later assembled as /dev/md2.
412 Giving the literal word "dev" for
416 to use the minor number of the md device that is being assembled.
420 will look for super blocks with a minor number of 0.
424 Assemble the array even if some superblocks appear out-of-date
428 Attempt to start the array even if fewer drives were given than are
429 needed for a full array. Normally if not all drives are found and
431 is not used, then the array will be assembled but not started.
434 an attempt will be made to start it anyway.
437 .BR -a ", " "--auto{=no,yes,md,mdp,part}"
438 See this option under Create and Build options.
441 .BR -U ", " --update=
442 Update the superblock on each device while assembling the array. The
443 argument given to this flag can be one of
452 option will adjust the superblock of an array what was created on a Sparc
453 machine running a patched 2.2 Linux kernel. This kernel got the
454 alignment of part of the superblock wrong. You can use the
455 .B "--examine --sparc2.2"
458 to see what effect this would have.
462 option will update the
464 field on each superblock to match the minor number of the array being
465 assembled. This is not needed on 2.6 and later kernels as they make
466 this adjustment automatically.
470 option will cause the array to be marked
472 meaning that any redundancy in the array (e.g. parity for raid5,
473 copies for raid1) may be incorrect. This will cause the raid system
474 to perform a "resync" pass to make sure that all redundant information
479 option will correct the summaries in the superblock. That is the
480 counts of total, working, active, failed, and spare devices.
487 hotadd listed devices.
491 remove listed devices. They must not be active. i.e. they should
492 be failed or spare devices.
496 mark listed devices as faulty.
502 .SH For Examine mode:
506 In an array was created on a 2.2 Linux kernel patched with RAID
507 support, the superblock will have been created incorrectly, or at
508 least incompatibly with 2.4 and later kernels. Using the
512 will fix the superblock before displaying it. If this appears to do
513 the right thing, then the array can be successfully assembled using
514 .BR "--assemble --update=sparc2.2" .
520 start a partially built array.
524 deactivate array, releasing all resources.
527 .BR -o ", " --readonly
528 mark array as readonly.
531 .BR -w ", " --readwrite
532 mark array as readwrite.
536 If the device contains a valid md superblock, the block is
537 over-written with zeros. With
539 the block where the superblock would be is over-written even if it
540 doesn't appear to be valid.
548 is set to reflect the status of the device.
550 .SH For Monitor mode:
553 Give a mail address to send alerts to.
556 .BR -p ", " --program ", " --alert
557 Give a program to be run whenever an event is detected.
561 Give a delay in seconds.
563 polls the md arrays and then waits this many seconds before polling
564 again. The default is 60 seconds.
567 .BR -f ", " --daemonise
570 to run as a background daemon if it decides to monitor anything. This
571 causes it to fork and run in the child, and to disconnect form the
572 terminal. The process id of the child is written to stdout.
575 which will only continue monitoring if a mail address or alert program
576 is found in the config file.
579 .BR -i ", " --pid-file
582 is running in daemon mode, write the pid of the daemon process to
583 the specified file, instead of printing it on standard output.
586 .BR -1 ", " --oneshot
587 Check arrays only once. This will generate
589 events and more significantly
593 .B " mdadm --monitor --scan -1"
595 from a cron script will ensure regular notification of any degraded arrays.
601 alert for every array found at startup. This alert gets mailed and
602 passed to the alert program. This can be used for testing that alert
603 message to get through successfully.
610 .I md-device options-and-component-devices...
613 .B mdadm --assemble --scan
614 .I md-devices-and-options...
617 .B mdadm --assemble --scan
621 This usage assembles one or more raid arrays from pre-existing components.
622 For each array, mdadm needs to know the md device, the identity of the
623 array, and a number of component-devices. These can be found in a number of ways.
625 In the first usage example (without the
627 the first device given is the md device.
628 In the second usage example, all devices listed are treated as md
629 devices and assembly is attempted.
630 In the third (where no devices are listed) all md devices that are
631 listed in the configuration file are assembled.
633 If precisely one device is listed, but
639 was given and identify information is extracted from the configuration file.
641 The identity can be given with the
645 option, can be found in the config file, or will be taken from the
646 super block on the first component-device listed on the command line.
648 Devices can be given on the
650 command line or in the config file. Only devices which have an md
651 superblock which contains the right identity will be considered for
654 The config file is only used if explicitly named with
656 or requested with (a possibly implicit)
664 is not given, then the config file will only be used to find the
665 identity of md arrays.
667 Normally the array will be started after it is assembled. However if
669 is not given and insufficient drives were listed to start a complete
670 (non-degraded) array, then the array is not started (to guard against
671 usage errors). To insist that the array be started in this case (as
672 may work for RAID1, 4, 5 or 6), give the
678 option is given, either on the command line (--auto) or in the
679 configuration file (e.g. auto=part), then
681 will create the md device if necessary or will re-create it if it
682 doesn't look usable as it is.
684 This can be useful for handling partitioned devices (which don't have
685 a stable device number - it can change after a reboot) and when using
686 "udev" to manage your
688 tree (udev cannot handle md devices because of the unusual device
689 initialisation conventions).
691 If the option to "auto" is "mdp" or "part" or (on the command line
692 only) "p", then mdadm will create a partitionable array, using the
693 first free one that is not inuse, and does not already have an entry
694 in /dev (apart from numeric /dev/md* entries).
696 If the option to "auto" is "yes" or "md" or (on the command line)
697 nothing, then mdadm will create a traditional, non-partitionable md
700 It is expected that the "auto" functionality will be used to create
701 device entries with meaningful names such as "/dev/md/home" or
702 "/dev/md/root", rather than names based on the numerical array number.
704 When using this option to create a partitionable array, the device
705 files for the first 4 partitions are also created. If a different
706 number is required it can be simply appended to the auto option.
707 e.g. "auto=part8". Partition names are created by appending a digit
708 string to the device name, with an intervening "_p" if the device name
713 option is also available in Build and Create modes. As those modes do
714 not use a config file, the "auto=" config option does not apply to
725 .BI --raid-devices= Z
729 This usage is similar to
731 The difference is that it creates a legacy array without a superblock. With
732 these arrays there is no difference between initially creating the array and
733 subsequently assembling the array, except that hopefully there is useful
734 data there in the second case.
736 The level may only be 0, raid0, or linear. All devices must be listed
737 and the array will be started once complete.
748 .BI --raid-devices= Z
752 This usage will initialise a new md array, associate some devices with
753 it, and activate the array.
757 option is given (as described in more detail in the section on
758 Assemble mode), then the md device will be created with a suitable
759 device number if necessary.
761 As devices are added, they are checked to see if they contain raid
762 superblocks or filesystems. They are also checked to see if the variance in
763 device size exceeds 1%.
765 If any discrepancy is found, the array will not automatically be run, though
768 can override this caution.
770 To create a "degraded" array in which some devices are missing, simply
771 give the word "\fBmissing\fP"
772 in place of a device name. This will cause
774 to leave the corresponding slot in the array empty.
775 For a RAID4 or RAID5 array at most one slot can be
776 "\fBmissing\fP"; for a RAID6 array at most two slots.
777 For a RAID1 array, only one real device needs to be given. All of the
781 When creating a RAID5 array,
783 will automatically create a degraded array with an extra spare drive.
784 This is because building the spare into a degraded array is in general faster than resyncing
785 the parity on a non-degraded, but not clean, array. This feature can
786 be over-ridden with the
792 '''option is given, it is not necessary to list any component-devices in this command.
793 '''They can be added later, before a
797 '''is given, the apparent size of the smallest drive given is used.
799 The General Management options that are valid with --create are:
802 insist on running the array even if some devices look like they might
807 start the array readonly - not supported yet.
814 .I options... devices...
817 This usage will allow individual devices in an array to be failed,
818 removed or added. It is possible to perform multiple operations with
819 on command. For example:
821 .B " mdadm /dev/md0 -f /dev/hda1 -r /dev/hda1 -a /dev/hda1"
827 and will then remove it from the array and finally add it back
828 in as a spare. However only one md array can be affected by a single
839 MISC mode includes a number of distinct operations that
840 operate on distinct devices. The operations are:
843 The device is examined to see if it is
844 (1) an active md array, or
845 (2) a component of an md array.
846 The information discovered is reported.
850 The device should be an active md device.
852 will display a detailed description of the array.
856 will cause the output to be less detailed and the format to be
857 suitable for inclusion in
858 .BR /etc/mdadm.conf .
861 will normally be 0 unless
863 failed to get useful information about the device(s). However if the
865 option is given, then the exit status will be:
869 The array is functioning normally.
872 The array has at least one failed device.
875 The array has multiple failed devices and hence is unusable (raid4 or
879 There was an error while trying to get information about the device.
884 The device should be a component of an md array.
886 will read the md superblock of the device and display the contents.
891 then multiple devices that are components of the one array
892 are grouped together and reported in a single entry suitable
894 .BR /etc/mdadm.conf .
898 without listing any devices will cause all devices listed in the
899 config file to be examined.
903 The devices should be active md arrays which will be deactivated, as
904 long as they are not currently in use.
908 This will fully activate a partially assembled md array.
912 This will mark an active array as read-only, providing that it is
913 not currently being used.
919 array back to being read/write.
923 For all operations except
926 will cause the operation to be applied to all arrays listed in
931 causes all devices listed in the config file to be examined.
939 .I options... devices...
944 to periodically poll a number of md arrays and to report on any events
947 will never exit once it decides that there are arrays to be checked,
948 so it should normally be run in the background.
950 As well as reporting events,
952 may move a spare drive from one array to another if they are in the
955 and if the destination array has a failed drive but not spares.
957 If any devices are listed on the command line,
959 will only monitor those devices. Otherwise all arrays listed in the
960 configuration file will be monitored. Further, if
962 is given, then any other md devices that appear in
964 will also be monitored.
966 The result of monitoring the arrays is the generation of events.
967 These events are passed to a separate program (if specified) and may
968 be mailed to a given E-mail address.
970 When passing event to program, the program is run once for each event
971 and is given 2 or 3 command-line arguements. The first is the
972 name of the event (see below). The second is the name of the
973 md device which is affected, and the third is the name of a related
974 device if relevant, such as a component device that has failed.
978 is given, then a program or an E-mail address must be specified on the
979 command line or in the config file. If neither are available, then
981 will not monitor anything.
985 will continue monitoring as long as something was found to monitor. If
986 no program or email is given, then each event is reported to
989 The different events are:
994 An md array which previously was configured appears to no longer be
999 An md array started reconstruction.
1005 is 20, 40, 60, or 80, this indicates that rebuild has passed that many
1006 percentage of the total.
1010 An md array that was rebuilding, isn't any more, either because it
1011 finished normally or was aborted.
1015 An active component device of an array has been marked as faulty.
1019 A spare component device which was being rebuilt to replace a faulty
1024 A spare component device which was being rebuilt to replace a faulty
1025 device as been successfully rebuild and has been made active.
1029 A new md array has been detected in the
1035 A newly noticed array appears to be degraded. This message is not
1038 notices a drive failure which causes degradation, but only when
1040 notices that an array is degraded when it first sees the array.
1044 A spare drive has been moved from one array in a
1046 to another to allow a failed drive to be replaced.
1050 An array was found at startup, and the
1061 cause Email to be sent. All events cause the program to be run.
1062 The program is run with two or three arguments, they being the event
1063 name, the array device and possibly a second device.
1065 Each event has an associated array device (e.g.
1067 and possibly a second device. For
1072 the second device is the relevant component device.
1075 the second device is the array that the spare was moved from.
1079 to move spares from one array to another, the different arrays need to
1080 be labelled with the same
1082 in the configuration file. The
1084 name can be any string. It is only necessary that different spare
1085 groups use different names.
1089 detects that an array which is in a spare group has fewer active
1090 devices than necessary for the complete array, and has no spare
1091 devices, it will look for another array in the same spare group that
1092 has a full complement of working drive and a spare. It will then
1093 attempt to remove the spare from the second drive and add it to the
1095 If the removal succeeds but the adding fails, then it is added back to
1099 The GROW mode is used for changing the size or shape of an active
1101 For this to work, the kernel must support the necessary change.
1102 Various types of growth may be added during 2.6 development, possibly
1103 including restructuring a raid5 array to have more active devices.
1105 Currently the only support available is to change the "size" attribute
1106 for arrays with redundancy, and the raid-disks attribute of RAID1
1109 Normally when an array is build the "size" it taken from the smallest
1110 of the drives. If all the small drives in an arrays are, one at a
1111 time, removed and replaced with larger drives, then you could have an
1112 array of large drives with only a small amount used. In this
1113 situation, changing the "size" with "GROW" mode will allow the extra
1114 space to start being used. If the size is increased in this way, a
1115 "resync" process will start to make sure the new parts of the array
1118 Note that when an array changes size, any filesystem that may be
1119 stored in the array will not automatically grow to use the space. The
1120 filesystem will need to be explicitly told to use the extra space.
1122 A RAID1 array can work with any number of devices from 1 upwards
1123 (though 1 is not very useful). There may be times which you want to
1124 increase or decrease the number of active devices. Note that this is
1125 different to hot-add or hot-remove which changes the number of
1128 When reducing the number of devices in a RAID1 array, the slots which
1129 are to be removed from the array must already be vacant. That is, the
1130 devices that which were in those slots must be failed and removed.
1132 When the number of devices is increased, any hot spares that are
1133 present may be activated immediately.
1137 .B " mdadm --query /dev/name-of-device"
1139 This will find out if a given device is a raid array, or is part of
1140 one, and will provide brief information about the device.
1142 .B " mdadm --assemble --scan"
1144 This will assemble and start all arrays listed in the standard confile
1145 file. This command will typically go in a system startup file.
1147 .B " mdadm --stop --scan"
1149 This will shut down all array that can be shut down (i.e. are not
1150 currently in use). This will typically go in a system shutdown script.
1152 .B " mdadm --follow --scan --delay=120"
1154 If (and only if) there is an Email address or program given in the
1155 standard config file, then
1156 monitor the status of all arrays listed in that file by
1157 polling them ever 2 minutes.
1159 .B " mdadm --create /dev/md0 --level=1 --raid-devices=2 /dev/hd[ac]1"
1161 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
1164 .B " echo 'DEVICE /dev/hd*[0-9] /dev/sd*[0-9]' > mdadm.conf"
1166 .B " mdadm --detail --scan >> mdadm.conf"
1168 This will create a prototype config file that describes currently
1169 active arrays that are known to be made from partitions of IDE or SCSI drives.
1170 This file should be reviewed before being used as it may
1171 contain unwanted detail.
1173 .B " echo 'DEVICE /dev/hd[a-z] /dev/sd*[a-z]' > mdadm.conf"
1175 .B " mdadm --examine --scan --config=mdadm.conf >> mdadm.conf"
1177 This will find what arrays could be assembled from existign IDE and
1178 SCSI whole drives (not partitions) and store the information is the
1179 format of a config file.
1180 This file is very likely to contain unwanted detail, particularly
1183 entries. It should be reviewed and edited before being used as an
1186 .B " mdadm --examine --brief --scan --config=partitions"
1188 .B " mdadm -Ebsc partitions"
1190 Create a list of devices by reading
1191 .BR /proc/partitions ,
1192 scan these for RAID superblocks, and printout a brief listing of all
1195 .B " mdadm -Ac partitions -m 0 /dev/md0"
1197 Scan all partitions and devices listed in
1198 .BR /proc/partitions
1201 out of all such devices with a RAID superblock with a minor number of 0.
1203 .B " mdadm --monitor --scan --daemonise > /var/run/mdadm"
1205 If config file contains a mail address or alert program, run mdadm in
1206 the background in monitor mode monitoring all md devices. Also write
1207 pid of mdadm daemon to
1208 .BR /var/run/mdadm .
1210 .B " mdadm --create --help"
1212 Providew help about the Create mode.
1214 .B " mdadm --config --help"
1216 Provide help about the format of the config file.
1220 Provide general help.
1231 lists all active md devices with information about them.
1233 uses this to find arrays when
1235 is given in Misc mode, and to monitor array reconstruction
1241 The config file lists which devices may be scanned to see if
1242 they contain MD super block, and gives identifying information
1243 (e.g. UUID) about known MD arrays. See
1250 was previously known as
1254 For information on the various levels of
1258 .UR http://ostenfeld.dk/~jakob/Software-RAID.HOWTO/
1259 http://ostenfeld.dk/~jakob/Software-RAID.HOWTO/
1262 for new releases of the RAID driver check out:
1265 .UR ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
1266 ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
1271 .UR http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
1272 http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/