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 there-of) 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
45 is not a Software RAID mechanism, but does involve
48 each device is a path to one common physical storage device.
51 is also not true RAID, and it only involves one device. It
52 provides a layer over a true device that can be used to inject faults.
55 '''is a program that can be used to create, manage, and monitor
57 '''such it provides a similar set of functionality to the
60 '''The key differences between
67 '''is a single program and not a collection of programs.
70 '''can perform (almost) all of its functions without having a
71 '''configuration file and does not use one by default. Also
73 '''helps with management of the configuration
77 '''can provide information about your arrays (through Query, Detail, and Examine)
87 '''configuration file, at all. It has a different configuration file
88 '''with a different format and a different purpose.
91 mdadm has 7 major modes of operation:
94 Assemble the parts of a previously created
95 array into an active array. Components can be explicitly given
96 or can be searched for.
98 checks that the components
99 do form a bona fide array, and can, on request, fiddle superblock
100 information so as to assemble a faulty array.
104 Build an array that doesn't have per-device superblocks. For these
107 cannot differentiate between initial creation and subsequent assembly
108 of an array. It also cannot perform any checks that appropriate
109 devices have been requested. Because of this, the
111 mode should only be used together with a complete understanding of
116 Create a new array with per-device superblocks.
118 '''in several step create-add-add-run or it can all happen with one command.
121 .B "Follow or Monitor"
122 Monitor one or more md devices and act on any state changes. This is
123 only meaningful for raid1, 4, 5, 6, 10 or multipath arrays as
124 only these have interesting state. raid0 or linear never have
125 missing, spare, or failed drives, so there is nothing to monitor.
129 Grow (or shrink) an array, or otherwise reshape it in some way.
130 Currently supported growth options including changing the active size
131 of component devices in RAID level 1/4/5/6 and changing the number of
132 active devices in RAID1.
136 This is for doing things to specific components of an array such as
137 adding new spares and removing faulty devices.
141 This is an 'everything else' mode that supports operations on active
142 arrays, operations on component devices such as erasing old superblocks, and
143 information gathering operations.
144 '''This mode allows operations on independent devices such as examine MD
145 '''superblocks, erasing old superblocks and stopping active arrays.
149 .SH Options for selecting a mode are:
152 .BR -A ", " --assemble
153 Assemble a pre-existing array.
157 Build a legacy array without superblocks.
164 .BR -F ", " --follow ", " --monitor
171 Change the size or shape of an active array.
173 If a device is given before any options, or if the first option is
178 then the MANAGE mode is assume.
179 Anything other than these will cause the
183 .SH Options that are not mode-specific are:
187 Display general help message or, after one of the above options, a
188 mode specific help message.
192 Display more detailed help about command line parsing and some commonly
196 .BR -V ", " --version
197 Print version information for mdadm.
200 .BR -v ", " --verbose
201 Be more verbose about what is happening. This can be used twice to be
203 The extra verbosity currently only affects
206 .BR "--examine --scan" .
210 Avoid printing purely informative messages. With this,
212 will be silent unless there is something really important to report.
216 Be less verbose. This is used with
224 gives an intermediate level of verbosity.
228 Be more forceful about certain operations. See the various modes of
229 the exact meaning of this option in different contexts.
232 .BR -c ", " --config=
233 Specify the config file. Default is to use
234 .BR /etc/mdadm.conf ,
235 or if that is missing, then
236 .BR /etc/mdadm/mdadm.conf .
237 If the config file given is
239 then nothing will be read, but
241 will act as though the config file contained exactly
242 .B "DEVICE partitions"
245 to find a list of devices to scan.
248 is given for the config file, then
250 will act as though the config file were empty.
256 for missing information.
257 In general, this option gives
259 permission to get any missing information, like component devices,
260 array devices, array identities, and alert destination from the
262 .BR /etc/mdadm.conf .
263 One exception is MISC mode when using
269 says to get a list of array devices from
273 .B -e ", " --metadata=
274 Declare the style of superblock (raid metadata) to be used. The
275 default is 0.90 for --create, and to guess for other operations.
276 The default can be overridden by setting the
285 .IP "0, 0.90, default"
286 Use the original 0.90 format superblock. This format limits arrays to
287 28 componenet devices and limits component devices of levels 1 and
288 greater to 2 terabytes.
289 .IP "1, 1.0, 1.1, 1.2"
290 Use the new version-1 format superblock. This has few restrictions.
291 The different subversion store the superblock at different locations
292 on the device, either at the end (for 1.0), at the start (for 1.1) or
293 4K from the start (for 1.2).
298 This will over-ride any
300 setting in the config file and provides the identify of the host which
301 should be considered the home for any arrays.
303 When creating an array, the
305 will be recorded in the superblock. For version-1 superblocks, it will
306 be prefixed to the array name. For version-0.90 superblocks part of
307 the SHA1 hash of the hostname will be stored in the later half of the
310 When reporting information about an array, any array which is tagged
311 for the given homehost will be reported as such.
313 When using Auto-Assemble, only arrays tagged for the given homehost
316 .SH For create, build, or grow:
319 .BR -n ", " --raid-devices=
320 Specify the number of active devices in the array. This, plus the
321 number of spare devices (see below) must equal the number of
323 (including "\fBmissing\fP" devices)
324 that are listed on the command line for
326 Setting a value of 1 is probably
327 a mistake and so requires that
329 be specified first. A value of 1 will then be allowed for linear,
330 multipath, raid0 and raid1. It is never allowed for raid4 or raid5.
332 This number can only be changed using
334 for RAID1 arrays, and only on kernels which provide necessary support.
337 .BR -x ", " --spare-devices=
338 Specify the number of spare (eXtra) devices in the initial array.
339 Spares can also be added
340 and removed later. The number of component devices listed
341 on the command line must equal the number of raid devices plus the
342 number of spare devices.
347 Amount (in Kibibytes) of space to use from each drive in RAID1/4/5/6.
348 This must be a multiple of the chunk size, and must leave about 128Kb
349 of space at the end of the drive for the RAID superblock.
350 If this is not specified
351 (as it normally is not) the smallest drive (or partition) sets the
352 size, though if there is a variance among the drives of greater than 1%, a warning is
355 This value can be set with
357 for RAID level 1/4/5/6. If the array was created with a size smaller
358 than the currently active drives, the extra space can be accessed
361 The size can be given as
363 which means to choose the largest size that fits on all current drives.
367 Specify chunk size of kibibytes. The default is 64.
371 Specify rounding factor for linear array (==chunk size)
375 Set raid level. When used with
377 options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4,
378 raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty. Obviously some of these are synonymous.
382 only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid.
384 Not yet supported with
388 .BR -p ", " --layout=
389 This option configures the fine details of data layout for raid5,
390 and raid10 arrays, and controls the failure modes for
393 The layout of the raid5 parity block can be one of
398 la, ra, ls, rs. The default is left-symmetric.
400 When setting the failure mode for
418 Each mode can be followed by a number which is used as a period
419 between fault generation. Without a number, the fault is generated
420 once on the first relevant request. With a number, the fault will be
421 generated after that many request, and will continue to be generated
422 every time the period elapses.
424 Multiple failure modes can be current simultaneously by using the
425 "--grow" option to set subsequent failure modes.
427 "clear" or "none" will remove any pending or periodic failure modes,
428 and "flush" will clear any persistent faults.
430 To set the parity with "--grow", the level of the array ("faulty")
431 must be specified before the fault mode is specified.
433 Finally, the layout options for RAID10 are one of 'n', 'o' or 'p' followed
434 by a small number. The default is 'n2'.
437 signals 'near' copies. Multiple copies of one data block are at
438 similar offsets in different devices.
441 signals 'offset' copies. Rather than the chunks being duplicated
442 within a stripe, whole stripes are duplicated but are rotated by one
443 device so duplicate blocks are on different devices. Thus subsequent
444 copies of a block are in the next drive, and are one chunk further
449 (multiple copies have very different offsets). See md(4) for more
450 detail about 'near' and 'far'.
452 The number is the number of copies of each datablock. 2 is normal, 3
453 can be useful. This number can be at most equal to the number of
454 devices in the array. It does not need to divide evenly into that
455 number (e.g. it is perfectly legal to have an 'n2' layout for an array
456 with an odd number of devices).
460 same as --layout (thus explaining the p of
464 .BR -b ", " --bitmap=
465 Specify a file to store a write-intent bitmap in. The file should not
466 exist unless --force is also given. The same file should be provided
467 when assembling the array. If the word
469 is given, then the bitmap is stored with the metadata on the array,
470 and so is replicated on all devices. If the word
474 mode, then any bitmap that is present is removed.
476 To help catch typing errors, the filename must contain at least one
477 slash ('/') if it is a real file (not 'internal' or 'none').
479 Note: external bitmaps are only known to work on ext2 and ext3.
480 Storing bitmap files on other filesystems may result in serious problems.
484 Set the chunksize of the bitmap. Each bit corresponds to that many
485 Kilobytes of storage.
486 When using a file based bitmap, the default is to use the smallest
487 size that is atleast 4 and requires no more than 2^21 chunks.
490 bitmap, the chunksize is automatically determined to make best use of
495 .BR -W ", " --write-mostly
496 subsequent devices lists in a
501 command will be flagged as 'write-mostly'. This is valid for RAID1
502 only and means that the 'md' driver will avoid reading from these
503 devices if at all possible. This can be useful if mirroring over a
508 Specify that write-behind mode should be enabled (valid for RAID1
509 only). If an argument is specified, it will set the maximum number
510 of outstanding writes allowed. The default value is 256.
511 A write-intent bitmap is required in order to use write-behind
512 mode, and write-behind is only attempted on drives marked as
519 that the array pre-existed and is known to be clean. It can be useful
520 when trying to recover from a major failure as you can be sure that no
521 data will be affected unless you actually write to the array. It can
522 also be used when creating a RAID1 or RAID10 if you want to avoid the
523 initial resync, however this practice - while normally safe - is not
524 recommended. Use this ony if you really know what you are doing.
528 This is needed when --grow is used to increase the number of
529 raid-devices in a RAID5 if there are no spare devices available.
530 See the section below on RAID_DEVICE CHANGES. The file should be
531 stored on a separate device, not on the raid array being reshaped.
537 for the array. This is currently only effective when creating an
538 array with a version-1 superblock. The name is a simple textual
539 string that can be used to identify array components when assembling.
545 run the array, even if some of the components
546 appear to be active in another array or filesystem. Normally
548 will ask for confirmation before including such components in an
549 array. This option causes that question to be suppressed.
555 accept the geometry and layout specified without question. Normally
557 will not allow creation of an array with only one device, and will try
558 to create a raid5 array with one missing drive (as this makes the
559 initial resync work faster). With
562 will not try to be so clever.
565 .BR -a ", " "--auto{=no,yes,md,mdp,part,p}{NN}"
566 Instruct mdadm to create the device file if needed, possibly allocating
567 an unused minor number. "md" causes a non-partitionable array
568 to be used. "mdp", "part" or "p" causes a partitionable array (2.6 and
569 later) to be used. "yes" requires the named md device to have
570 a 'standard' format, and the type and minor number will be determined
571 from this. See DEVICE NAMES below.
573 The argument can also come immediately after
576 If --auto is not given on the command line or in the config file, then
582 is also given, then any
584 entries in the config file will over-ride the
586 instruction given on the command line.
588 For partitionable arrays,
590 will create the device file for the whole array and for the first 4
591 partitions. A different number of partitions can be specified at the
592 end of this option (e.g.
594 If the device name ends with a digit, the partition names add a 'p',
595 and a number, e.g. "/dev/home1p3". If there is no
596 trailing digit, then the partition names just have a number added,
597 e.g. "/dev/scratch3".
599 If the md device name is in a 'standard' format as described in DEVICE
600 NAMES, then it will be created, if necessary, with the appropriate
601 number based on that name. If the device name is not in one of these
602 formats, then a unused minor number will be allocated. The minor
603 number will be considered unused if there is no active array for that
604 number, and there is no entry in /dev for that number and with a
615 it will also create symlinks from
617 with names starting with
625 to enforce this even if it is suppressing
633 uuid of array to assemble. Devices which don't have this uuid are
637 .BR -m ", " --super-minor=
638 Minor number of device that array was created for. Devices which
639 don't have this minor number are excluded. If you create an array as
640 /dev/md1, then all superblocks will contain the minor number 1, even if
641 the array is later assembled as /dev/md2.
643 Giving the literal word "dev" for
647 to use the minor number of the md device that is being assembled.
651 will look for super blocks with a minor number of 0.
655 Specify the name of the array to assemble. This must be the name
656 that was specified when creating the array. It must either match
657 then name stored in the superblock exactly, or it must match
660 is added to the start of the given name.
664 Assemble the array even if some superblocks appear out-of-date
668 Attempt to start the array even if fewer drives were given than were
669 present last time the array was active. Normally if not all the
670 expected drives are found and
672 is not used, then the array will be assembled but not started.
675 an attempt will be made to start it anyway.
679 This is the reverse of
681 in that it inhibits the started if array unless all expected drives
682 are present. This is only needed with
684 and can be used if you physical connections to devices are
685 not as reliable as you would like.
688 .BR -a ", " "--auto{=no,yes,md,mdp,part}"
689 See this option under Create and Build options.
692 .BR -b ", " --bitmap=
693 Specify the bitmap file that was given when the array was created. If
696 bitmap, there is no need to specify this when assembling the array.
702 was used to grow the number of raid-devices in a RAID5, and the system
703 crashed during the critical section, then the same
705 must be presented to --assemble to allow possibly corrupted data to be
709 .BR -U ", " --update=
710 Update the superblock on each device while assembling the array. The
711 argument given to this flag can be one of
725 option will adjust the superblock of an array what was created on a Sparc
726 machine running a patched 2.2 Linux kernel. This kernel got the
727 alignment of part of the superblock wrong. You can use the
728 .B "--examine --sparc2.2"
731 to see what effect this would have.
735 option will update the
737 field on each superblock to match the minor number of the array being
739 This can be useful if
741 reports a different "Preferred Minor" to
743 In some cases this update will be performed automatically
744 by the kernel driver. In particular the update happens automatically
745 at the first write to an array with redundancy (RAID level 1 or
746 greater) on a 2.6 (or later) kernel.
750 option will change the uuid of the array. If a UUID is given with the
751 "--uuid" option that UUID will be used as a new UUID and will
753 be used to help identify the devices in the array.
754 If no "--uuid" is given, a random uuid is chosen.
758 option will change the
760 of the array as stored in the superblock. This is only supported for
761 version-1 superblocks.
765 option will change the
767 as recorded in the superblock. For version-0 superblocks, this is the
768 same as updating the UUID.
769 For version-1 superblocks, this involves updating the name.
773 option will cause the array to be marked
775 meaning that any redundancy in the array (e.g. parity for raid5,
776 copies for raid1) may be incorrect. This will cause the raid system
777 to perform a "resync" pass to make sure that all redundant information
782 option allows arrays to be moved between machines with different
784 When assembling such an array for the first time after a move, giving
785 .B "--update=byteorder"
788 to expect superblocks to have their byteorder reversed, and will
789 correct that order before assembling the array. This is only valid
790 with original (Version 0.90) superblocks.
794 option will correct the summaries in the superblock. That is the
795 counts of total, working, active, failed, and spare devices.
799 will rarely be of use. It applies to version 1.1 and 1.2 metadata
800 only (where the metadata is at the start of the device) and is only
801 useful when the component device has changed size (typically become
802 larger). The version 1 metadata records the amount of the device that
803 can be used to store data, so if a device in a version 1.1 or 1.2
804 array becomes larger, the metadata will still be visible, but the
805 extra space will not. In this case it might be useful to assemble the
807 .BR --update=devicesize .
810 to determine the maximum usable amount of space on each device and
811 update the relevant field in the metadata.
814 .B --auto-update-homehost
815 This flag is only meaning with auto-assembly (see discussion below).
816 In that situation, if no suitable arrays are found for this homehost,
818 will recan for any arrays at all and will assemble them and update the
819 homehost to match the current host.
825 hot-add listed devices.
829 re-add a device that was recently removed from an array.
833 remove listed devices. They must not be active. i.e. they should
834 be failed or spare devices.
838 mark listed devices as faulty.
845 Each of these options require that the first device list is the array
846 to be acted upon and the remainder are component devices to be added,
847 removed, or marked as fault. Several different operations can be
848 specified for different devices, e.g.
850 mdadm /dev/md0 --add /dev/sda1 --fail /dev/sdb1 --remove /dev/sdb1
852 Each operation applies to all devices listed until the next
855 If an array is using a write-intent bitmap, then devices which have
856 been removed can be re-added in a way that avoids a full
857 reconstruction but instead just updated the blocks that have changed
858 since the device was removed. For arrays with persistent metadata
859 (superblocks) this is done automatically. For arrays created with
861 mdadm needs to be told that this device we removed recently with
864 Devices can only be removed from an array if they are not in active
865 use. i.e. that must be spares or failed devices. To remove an active
866 device, it must be marked as
874 Examine a device to see
875 (1) if it is an md device and (2) if it is a component of an md
877 Information about what is discovered is presented.
881 Print detail of one or more md devices.
884 .BR -E ", " --examine
885 Print content of md superblock on device(s).
888 If an array was created on a 2.2 Linux kernel patched with RAID
889 support, the superblock will have been created incorrectly, or at
890 least incompatibly with 2.4 and later kernels. Using the
894 will fix the superblock before displaying it. If this appears to do
895 the right thing, then the array can be successfully assembled using
896 .BR "--assemble --update=sparc2.2" .
899 .BR -X ", " --examine-bitmap
900 Report information about a bitmap file.
904 start a partially built array.
908 deactivate array, releasing all resources.
911 .BR -o ", " --readonly
912 mark array as readonly.
915 .BR -w ", " --readwrite
916 mark array as readwrite.
920 If the device contains a valid md superblock, the block is
921 over-written with zeros. With
923 the block where the superblock would be is over-written even if it
924 doesn't appear to be valid.
932 is set to reflect the status of the device.
936 For each md device given, wait for any resync, recovery, or reshape
937 activity to finish before returning.
939 will return with success if it actually waited for every device
940 listed, otherwise it will return failure.
942 .SH For Monitor mode:
945 Give a mail address to send alerts to.
948 .BR -p ", " --program ", " --alert
949 Give a program to be run whenever an event is detected.
953 Cause all events to be reported through 'syslog'. The messages have
954 facility of 'daemon' and varying priorities.
958 Give a delay in seconds.
960 polls the md arrays and then waits this many seconds before polling
961 again. The default is 60 seconds.
964 .BR -f ", " --daemonise
967 to run as a background daemon if it decides to monitor anything. This
968 causes it to fork and run in the child, and to disconnect form the
969 terminal. The process id of the child is written to stdout.
972 which will only continue monitoring if a mail address or alert program
973 is found in the config file.
976 .BR -i ", " --pid-file
979 is running in daemon mode, write the pid of the daemon process to
980 the specified file, instead of printing it on standard output.
983 .BR -1 ", " --oneshot
984 Check arrays only once. This will generate
986 events and more significantly
992 .B " mdadm --monitor --scan -1"
994 from a cron script will ensure regular notification of any degraded arrays.
1000 alert for every array found at startup. This alert gets mailed and
1001 passed to the alert program. This can be used for testing that alert
1002 message do get through successfully.
1009 .I md-device options-and-component-devices...
1012 .B mdadm --assemble --scan
1013 .I md-devices-and-options...
1016 .B mdadm --assemble --scan
1020 This usage assembles one or more raid arrays from pre-existing components.
1021 For each array, mdadm needs to know the md device, the identity of the
1022 array, and a number of component-devices. These can be found in a number of ways.
1024 In the first usage example (without the
1026 the first device given is the md device.
1027 In the second usage example, all devices listed are treated as md
1028 devices and assembly is attempted.
1029 In the third (where no devices are listed) all md devices that are
1030 listed in the configuration file are assembled.
1032 If precisely one device is listed, but
1038 was given and identify information is extracted from the configuration file.
1040 The identity can be given with the
1044 option, can be found in the config file, or will be taken from the
1045 super block on the first component-device listed on the command line.
1047 Devices can be given on the
1049 command line or in the config file. Only devices which have an md
1050 superblock which contains the right identity will be considered for
1053 The config file is only used if explicitly named with
1055 or requested with (a possibly implicit)
1063 is not given, then the config file will only be used to find the
1064 identity of md arrays.
1066 Normally the array will be started after it is assembled. However if
1068 is not given and insufficient drives were listed to start a complete
1069 (non-degraded) array, then the array is not started (to guard against
1070 usage errors). To insist that the array be started in this case (as
1071 may work for RAID1, 4, 5, 6, or 10), give the
1075 If the md device does not exist, then it will be created providing the
1076 intent is clear. i.e. the name must be in a standard form, or the
1078 option must be given to clarify how and whether the device should be
1081 This can be useful for handling partitioned devices (which don't have
1082 a stable device number - it can change after a reboot) and when using
1083 "udev" to manage your
1085 tree (udev cannot handle md devices because of the unusual device
1086 initialisation conventions).
1088 If the option to "auto" is "mdp" or "part" or (on the command line
1089 only) "p", then mdadm will create a partitionable array, using the
1090 first free one that is not in use, and does not already have an entry
1091 in /dev (apart from numeric /dev/md* entries).
1093 If the option to "auto" is "yes" or "md" or (on the command line)
1094 nothing, then mdadm will create a traditional, non-partitionable md
1097 It is expected that the "auto" functionality will be used to create
1098 device entries with meaningful names such as "/dev/md/home" or
1099 "/dev/md/root", rather than names based on the numerical array number.
1101 When using this option to create a partitionable array, the device
1102 files for the first 4 partitions are also created. If a different
1103 number is required it can be simply appended to the auto option.
1104 e.g. "auto=part8". Partition names are created by appending a digit
1105 string to the device name, with an intervening "p" if the device name
1110 option is also available in Build and Create modes. As those modes do
1111 not use a config file, the "auto=" config option does not apply to
1119 and no devices are listed,
1121 will first attempt to assemble all the arrays listed in the config
1126 has been specified (either in the config file or on the command line),
1128 will look further for possible arrays and will try to assemble
1129 anything that it finds which is tagged as belonging to the given
1130 homehost. This is the only situation where
1132 will assemble arrays without being given specific device name or
1133 identify information for the array.
1137 finds a consistent set of devices that look like they should comprise
1138 an array, and if the superblock is tagged as belonging to the given
1139 home host, it will automatically choose a device name and try to
1140 assemble the array. If the array uses version-0.90 metadata, then the
1142 number as recorded in the superblock is used to create a name in
1146 If the array uses version-1 metadata, then the
1148 from the superblock is used to similarly create a name in
1150 The name will have any 'host' prefix stripped first.
1154 cannot find any array for the given host at all, and if
1155 .B --auto-update-homehost
1158 will search again for any array (not just an array created for this
1159 host) and will assemble each assuming
1160 .IR --update=homehost .
1161 This will change the host tag in the superblock so that on the next run,
1162 these arrays will be found without the second pass. The intention of
1163 this feature is to support transitioning a set of md arrays to using
1166 The reason for requiring arrays to be tagged with the homehost for
1167 auto assembly is to guard against problems that can arise when moving
1168 devices from one host to another.
1178 .BI --raid-devices= Z
1182 This usage is similar to
1184 The difference is that it creates an array without a superblock. With
1185 these arrays there is no difference between initially creating the array and
1186 subsequently assembling the array, except that hopefully there is useful
1187 data there in the second case.
1189 The level may raid0, linear, multipath, or faulty, or one of their
1190 synonyms. All devices must be listed and the array will be started
1202 .BI --raid-devices= Z
1206 This usage will initialise a new md array, associate some devices with
1207 it, and activate the array.
1211 option is given (as described in more detail in the section on
1212 Assemble mode), then the md device will be created with a suitable
1213 device number if necessary.
1215 As devices are added, they are checked to see if they contain raid
1216 superblocks or filesystems. They are also checked to see if the variance in
1217 device size exceeds 1%.
1219 If any discrepancy is found, the array will not automatically be run, though
1222 can override this caution.
1224 To create a "degraded" array in which some devices are missing, simply
1225 give the word "\fBmissing\fP"
1226 in place of a device name. This will cause
1228 to leave the corresponding slot in the array empty.
1229 For a RAID4 or RAID5 array at most one slot can be
1230 "\fBmissing\fP"; for a RAID6 array at most two slots.
1231 For a RAID1 array, only one real device needs to be given. All of the
1235 When creating a RAID5 array,
1237 will automatically create a degraded array with an extra spare drive.
1238 This is because building the spare into a degraded array is in general faster than resyncing
1239 the parity on a non-degraded, but not clean, array. This feature can
1240 be over-ridden with the
1244 When creating an array with version-1 metadata a name for the host is
1246 If this is not given with the
1250 will chose a name based on the last component of the name of the
1251 device being created. So if
1253 is being created, then the name
1258 is being created, then the name
1262 A new array will normally get a randomly assigned 128bit UUID which is
1263 very likely to be unique. If you have a specific need, you can choose
1264 a UUID for the array by giving the
1266 option. Be warned that creating two arrays with the same UUID is a
1267 recipe for disaster. Also, using
1269 when creating a v0.90 array will silently override any
1274 '''option is given, it is not necessary to list any component-devices in this command.
1275 '''They can be added later, before a
1279 '''is given, the apparent size of the smallest drive given is used.
1281 The General Management options that are valid with --create are:
1284 insist on running the array even if some devices look like they might
1289 start the array readonly - not supported yet.
1297 .I options... devices...
1300 This usage will allow individual devices in an array to be failed,
1301 removed or added. It is possible to perform multiple operations with
1302 on command. For example:
1304 .B " mdadm /dev/md0 -f /dev/hda1 -r /dev/hda1 -a /dev/hda1"
1310 and will then remove it from the array and finally add it back
1311 in as a spare. However only one md array can be affected by a single
1322 MISC mode includes a number of distinct operations that
1323 operate on distinct devices. The operations are:
1326 The device is examined to see if it is
1327 (1) an active md array, or
1328 (2) a component of an md array.
1329 The information discovered is reported.
1333 The device should be an active md device.
1335 will display a detailed description of the array.
1339 will cause the output to be less detailed and the format to be
1340 suitable for inclusion in
1341 .BR /etc/mdadm.conf .
1344 will normally be 0 unless
1346 failed to get useful information about the device(s). However if the
1348 option is given, then the exit status will be:
1352 The array is functioning normally.
1355 The array has at least one failed device.
1358 The array has multiple failed devices and hence is unusable (raid4 or
1362 There was an error while trying to get information about the device.
1367 The device should be a component of an md array.
1369 will read the md superblock of the device and display the contents.
1374 then multiple devices that are components of the one array
1375 are grouped together and reported in a single entry suitable
1377 .BR /etc/mdadm.conf .
1381 without listing any devices will cause all devices listed in the
1382 config file to be examined.
1386 The devices should be active md arrays which will be deactivated, as
1387 long as they are not currently in use.
1391 This will fully activate a partially assembled md array.
1395 This will mark an active array as read-only, providing that it is
1396 not currently being used.
1402 array back to being read/write.
1406 For all operations except
1409 will cause the operation to be applied to all arrays listed in
1414 causes all devices listed in the config file to be examined.
1422 .I options... devices...
1427 to periodically poll a number of md arrays and to report on any events
1430 will never exit once it decides that there are arrays to be checked,
1431 so it should normally be run in the background.
1433 As well as reporting events,
1435 may move a spare drive from one array to another if they are in the
1438 and if the destination array has a failed drive but no spares.
1440 If any devices are listed on the command line,
1442 will only monitor those devices. Otherwise all arrays listed in the
1443 configuration file will be monitored. Further, if
1445 is given, then any other md devices that appear in
1447 will also be monitored.
1449 The result of monitoring the arrays is the generation of events.
1450 These events are passed to a separate program (if specified) and may
1451 be mailed to a given E-mail address.
1453 When passing event to program, the program is run once for each event
1454 and is given 2 or 3 command-line arguments. The first is the
1455 name of the event (see below). The second is the name of the
1456 md device which is affected, and the third is the name of a related
1457 device if relevant, such as a component device that has failed.
1461 is given, then a program or an E-mail address must be specified on the
1462 command line or in the config file. If neither are available, then
1464 will not monitor anything.
1468 will continue monitoring as long as something was found to monitor. If
1469 no program or email is given, then each event is reported to
1472 The different events are:
1476 .B DeviceDisappeared
1477 An md array which previously was configured appears to no longer be
1478 configured. (syslog priority: Critical)
1482 was told to monitor an array which is RAID0 or Linear, then it will
1484 .B DeviceDisappeared
1485 with the extra information
1487 This is because RAID0 and Linear do not support the device-failed,
1488 hot-spare and resync operations which are monitored.
1492 An md array started reconstruction. (syslog priority: Warning)
1498 is 20, 40, 60, or 80, this indicates that rebuild has passed that many
1499 percentage of the total. (syslog priority: Warning)
1503 An md array that was rebuilding, isn't any more, either because it
1504 finished normally or was aborted. (syslog priority: Warning)
1508 An active component device of an array has been marked as
1509 faulty. (syslog priority: Critical)
1513 A spare component device which was being rebuilt to replace a faulty
1514 device has failed. (syslog priority: Critial)
1518 A spare component device which was being rebuilt to replace a faulty
1519 device has been successfully rebuilt and has been made active.
1520 (syslog priority: Info)
1524 A new md array has been detected in the
1526 file. (syslog priority: Info)
1530 A newly noticed array appears to be degraded. This message is not
1533 notices a drive failure which causes degradation, but only when
1535 notices that an array is degraded when it first sees the array.
1536 (syslog priority: Critial)
1540 A spare drive has been moved from one array in a
1542 to another to allow a failed drive to be replaced.
1543 (syslog priority: Info)
1549 has been told, via the config file, that an array should have a certain
1550 number of spare devices, and
1552 detects that it has fewer that this number when it first sees the
1553 array, it will report a
1556 (syslog priority: Warning)
1560 An array was found at startup, and the
1563 (syslog priority: Info)
1573 cause Email to be sent. All events cause the program to be run.
1574 The program is run with two or three arguments, they being the event
1575 name, the array device and possibly a second device.
1577 Each event has an associated array device (e.g.
1579 and possibly a second device. For
1584 the second device is the relevant component device.
1587 the second device is the array that the spare was moved from.
1591 to move spares from one array to another, the different arrays need to
1592 be labelled with the same
1594 in the configuration file. The
1596 name can be any string. It is only necessary that different spare
1597 groups use different names.
1601 detects that an array which is in a spare group has fewer active
1602 devices than necessary for the complete array, and has no spare
1603 devices, it will look for another array in the same spare group that
1604 has a full complement of working drive and a spare. It will then
1605 attempt to remove the spare from the second drive and add it to the
1607 If the removal succeeds but the adding fails, then it is added back to
1611 The GROW mode is used for changing the size or shape of an active
1613 For this to work, the kernel must support the necessary change.
1614 Various types of growth are being added during 2.6 development,
1615 including restructuring a raid5 array to have more active devices.
1617 Currently the only support available is to
1619 change the "size" attribute
1620 for RAID1, RAID5 and RAID6.
1622 increase the "raid-disks" attribute of RAID1 and RAID5.
1624 add a write-intent bitmap to any array which support these bitmaps, or
1625 remove a write-intent bitmap from such an array.
1629 Normally when an array is built the "size" it taken from the smallest
1630 of the drives. If all the small drives in an arrays are, one at a
1631 time, removed and replaced with larger drives, then you could have an
1632 array of large drives with only a small amount used. In this
1633 situation, changing the "size" with "GROW" mode will allow the extra
1634 space to start being used. If the size is increased in this way, a
1635 "resync" process will start to make sure the new parts of the array
1638 Note that when an array changes size, any filesystem that may be
1639 stored in the array will not automatically grow to use the space. The
1640 filesystem will need to be explicitly told to use the extra space.
1642 .SS RAID-DEVICES CHANGES
1644 A RAID1 array can work with any number of devices from 1 upwards
1645 (though 1 is not very useful). There may be times which you want to
1646 increase or decrease the number of active devices. Note that this is
1647 different to hot-add or hot-remove which changes the number of
1650 When reducing the number of devices in a RAID1 array, the slots which
1651 are to be removed from the array must already be vacant. That is, the
1652 devices that which were in those slots must be failed and removed.
1654 When the number of devices is increased, any hot spares that are
1655 present will be activated immediately.
1657 Increasing the number of active devices in a RAID5 is much more
1658 effort. Every block in the array will need to be read and written
1659 back to a new location. From 2.6.17, the Linux Kernel is able to do
1660 this safely, including restart and interrupted "reshape".
1662 When relocating the first few stripes on a raid5, it is not possible
1663 to keep the data on disk completely consistent and crash-proof. To
1664 provide the required safety, mdadm disables writes to the array while
1665 this "critical section" is reshaped, and takes a backup of the data
1666 that is in that section. This backup is normally stored in any spare
1667 devices that the array has, however it can also be stored in a
1668 separate file specified with the
1670 option. If this option is used, and the system does crash during the
1671 critical period, the same file must be passed to
1673 to restore the backup and reassemble the array.
1677 A write-intent bitmap can be added to, or removed from, an active
1678 array. Either internal bitmaps, or bitmaps stored in a separate file
1679 can be added. Note that if you add a bitmap stored in a file which is
1680 in a filesystem that is on the raid array being affected, the system
1681 will deadlock. The bitmap must be on a separate filesystem.
1685 .B " mdadm --query /dev/name-of-device"
1687 This will find out if a given device is a raid array, or is part of
1688 one, and will provide brief information about the device.
1690 .B " mdadm --assemble --scan"
1692 This will assemble and start all arrays listed in the standard config file
1693 file. This command will typically go in a system startup file.
1695 .B " mdadm --stop --scan"
1697 This will shut down all array that can be shut down (i.e. are not
1698 currently in use). This will typically go in a system shutdown script.
1700 .B " mdadm --follow --scan --delay=120"
1702 If (and only if) there is an Email address or program given in the
1703 standard config file, then
1704 monitor the status of all arrays listed in that file by
1705 polling them ever 2 minutes.
1707 .B " mdadm --create /dev/md0 --level=1 --raid-devices=2 /dev/hd[ac]1"
1709 Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1.
1712 .B " echo 'DEVICE /dev/hd*[0-9] /dev/sd*[0-9]' > mdadm.conf"
1714 .B " mdadm --detail --scan >> mdadm.conf"
1716 This will create a prototype config file that describes currently
1717 active arrays that are known to be made from partitions of IDE or SCSI drives.
1718 This file should be reviewed before being used as it may
1719 contain unwanted detail.
1721 .B " echo 'DEVICE /dev/hd[a-z] /dev/sd*[a-z]' > mdadm.conf"
1723 .B " mdadm --examine --scan --config=mdadm.conf >> mdadm.conf"
1725 This will find what arrays could be assembled from existing IDE and
1726 SCSI whole drives (not partitions) and store the information is the
1727 format of a config file.
1728 This file is very likely to contain unwanted detail, particularly
1731 entries. It should be reviewed and edited before being used as an
1734 .B " mdadm --examine --brief --scan --config=partitions"
1736 .B " mdadm -Ebsc partitions"
1738 Create a list of devices by reading
1739 .BR /proc/partitions ,
1740 scan these for RAID superblocks, and printout a brief listing of all
1743 .B " mdadm -Ac partitions -m 0 /dev/md0"
1745 Scan all partitions and devices listed in
1746 .BR /proc/partitions
1749 out of all such devices with a RAID superblock with a minor number of 0.
1751 .B " mdadm --monitor --scan --daemonise > /var/run/mdadm"
1753 If config file contains a mail address or alert program, run mdadm in
1754 the background in monitor mode monitoring all md devices. Also write
1755 pid of mdadm daemon to
1756 .BR /var/run/mdadm .
1758 .B " mdadm --create --help"
1760 Provide help about the Create mode.
1762 .B " mdadm --config --help"
1764 Provide help about the format of the config file.
1768 Provide general help.
1779 lists all active md devices with information about them.
1781 uses this to find arrays when
1783 is given in Misc mode, and to monitor array reconstruction
1789 The config file lists which devices may be scanned to see if
1790 they contain MD super block, and gives identifying information
1791 (e.g. UUID) about known MD arrays. See
1797 While entries in the /dev directory can have any format you like,
1799 has an understanding of 'standard' formats which it uses to guide its
1800 behaviour when creating device files via the
1804 The standard names for non-partitioned arrays (the only sort of md
1805 array available in 2.4 and earlier) either of
1811 where NN is a number.
1812 The standard names for partitionable arrays (as available from 2.6
1819 Partition numbers should be indicated by added "pMM" to these, thus "/dev/md/d1p2".
1823 was previously known as
1827 is completely separate from the
1829 package, and does not use the
1831 configuration file at all.
1834 For information on the various levels of
1838 .UR http://ostenfeld.dk/~jakob/Software-RAID.HOWTO/
1839 http://ostenfeld.dk/~jakob/Software-RAID.HOWTO/
1842 '''for new releases of the RAID driver check out:
1845 '''.UR ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
1846 '''ftp://ftp.kernel.org/pub/linux/kernel/people/mingo/raid-patches
1851 '''.UR http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
1852 '''http://www.cse.unsw.edu.au/~neilb/patches/linux-stable/
1855 The latest version of
1857 should always be available from
1859 .UR http://www.kernel.org/pub/linux/utils/raid/mdadm/
1860 http://www.kernel.org/pub/linux/utils/raid/mdadm/