1 Tools that manage md devices can be found at
2 http://www.<country>.kernel.org/pub/linux/utils/raid/....
5 Boot time assembly of RAID arrays
6 ---------------------------------
8 You can boot with your md device with the following kernel command
11 for old raid arrays without persistent superblocks:
12 md=<md device no.>,<raid level>,<chunk size factor>,<fault level>,dev0,dev1,...,devn
14 for raid arrays with persistent superblocks
15 md=<md device no.>,dev0,dev1,...,devn
16 or, to assemble a partitionable array:
17 md=d<md device no.>,dev0,dev1,...,devn
19 md device no. = the number of the md device ...
26 raid level = -1 linear mode
28 other modes are only supported with persistent super blocks
30 chunk size factor = (raid-0 and raid-1 only)
31 Set the chunk size as 4k << n.
33 fault level = totally ignored
35 dev0-devn: e.g. /dev/hda1,/dev/hdc1,/dev/sda1,/dev/sdb1
37 A possible loadlin line (Harald Hoyer <HarryH@Royal.Net>) looks like this:
39 e:\loadlin\loadlin e:\zimage root=/dev/md0 md=0,0,4,0,/dev/hdb2,/dev/hdc3 ro
42 Boot time autodetection of RAID arrays
43 --------------------------------------
45 When md is compiled into the kernel (not as module), partitions of
46 type 0xfd are scanned and automatically assembled into RAID arrays.
47 This autodetection may be suppressed with the kernel parameter
48 "raid=noautodetect". As of kernel 2.6.9, only drives with a type 0
49 superblock can be autodetected and run at boot time.
51 The kernel parameter "raid=partitionable" (or "raid=part") means
52 that all auto-detected arrays are assembled as partitionable.
54 Boot time assembly of degraded/dirty arrays
55 -------------------------------------------
57 If a raid5 or raid6 array is both dirty and degraded, it could have
58 undetectable data corruption. This is because the fact that it is
59 'dirty' means that the parity cannot be trusted, and the fact that it
60 is degraded means that some datablocks are missing and cannot reliably
61 be reconstructed (due to no parity).
63 For this reason, md will normally refuse to start such an array. This
64 requires the sysadmin to take action to explicitly start the array
65 desipite possible corruption. This is normally done with
66 mdadm --assemble --force ....
68 This option is not really available if the array has the root
69 filesystem on it. In order to support this booting from such an
70 array, md supports a module parameter "start_dirty_degraded" which,
71 when set to 1, bypassed the checks and will allows dirty degraded
74 So, to boot with a root filesystem of a dirty degraded raid[56], use
76 md-mod.start_dirty_degraded=1
82 The md driver can support a variety of different superblock formats.
83 Currently, it supports superblock formats "0.90.0" and the "md-1" format
84 introduced in the 2.5 development series.
86 The kernel will autodetect which format superblock is being used.
88 Superblock format '0' is treated differently to others for legacy
89 reasons - it is the original superblock format.
92 General Rules - apply for all superblock formats
93 ------------------------------------------------
95 An array is 'created' by writing appropriate superblocks to all
98 It is 'assembled' by associating each of these devices with an
99 particular md virtual device. Once it is completely assembled, it can
102 An array should be created by a user-space tool. This will write
103 superblocks to all devices. It will usually mark the array as
104 'unclean', or with some devices missing so that the kernel md driver
105 can create appropriate redundancy (copying in raid1, parity
106 calculation in raid4/5).
108 When an array is assembled, it is first initialized with the
109 SET_ARRAY_INFO ioctl. This contains, in particular, a major and minor
110 version number. The major version number selects which superblock
111 format is to be used. The minor number might be used to tune handling
112 of the format, such as suggesting where on each device to look for the
115 Then each device is added using the ADD_NEW_DISK ioctl. This
116 provides, in particular, a major and minor number identifying the
119 The array is started with the RUN_ARRAY ioctl.
121 Once started, new devices can be added. They should have an
122 appropriate superblock written to them, and then passed be in with
125 Devices that have failed or are not yet active can be detached from an
126 array using HOT_REMOVE_DISK.
129 Specific Rules that apply to format-0 super block arrays, and
130 arrays with no superblock (non-persistent).
131 -------------------------------------------------------------
133 An array can be 'created' by describing the array (level, chunksize
134 etc) in a SET_ARRAY_INFO ioctl. This must has major_version==0 and
137 Then uninitialized devices can be added with ADD_NEW_DISK. The
138 structure passed to ADD_NEW_DISK must specify the state of the device
139 and it's role in the array.
141 Once started with RUN_ARRAY, uninitialized spares can be added with
148 md devices appear in sysfs (/sys) as regular block devices,
152 Each 'md' device will contain a subdirectory called 'md' which
153 contains further md-specific information about the device.
155 All md devices contain:
157 a text file indicating the 'raid level'. This may be a standard
158 numerical level prefixed by "RAID-" - e.g. "RAID-5", or some
159 other name such as "linear" or "multipath".
160 If no raid level has been set yet (array is still being
161 assembled), this file will be empty.
164 a text file with a simple number indicating the number of devices
165 in a fully functional array. If this is not yet known, the file
166 will be empty. If an array is being resized (not currently
167 possible) this will contain the larger of the old and new sizes.
168 Some raid level (RAID1) allow this value to be set while the
169 array is active. This will reconfigure the array. Otherwise
170 it can only be set while assembling an array.
173 This is the size if bytes for 'chunks' and is only relevant to
174 raid levels that involve striping (1,4,5,6,10). The address space
175 of the array is conceptually divided into chunks and consecutive
176 chunks are striped onto neighbouring devices.
177 The size should be atleast PAGE_SIZE (4k) and should be a power
178 of 2. This can only be set while assembling an array
181 For arrays with data redundancy (i.e. not raid0, linear, faulty,
182 multipath), all components must be the same size - or at least
183 there must a size that they all provide space for. This is a key
184 part or the geometry of the array. It is measured in sectors
185 and can be read from here. Writing to this value may resize
186 the array if the personality supports it (raid1, raid5, raid6),
187 and if the component drives are large enough.
190 This indicates the format that is being used to record metadata
191 about the array. It can be 0.90 (traditional format), 1.0, 1.1,
192 1.2 (newer format in varying locations) or "none" indicating that
193 the kernel isn't managing metadata at all.
196 The raid 'level' for this array. The name will often (but not
197 always) be the same as the name of the module that implements the
198 level. To be auto-loaded the module must have an alias
199 md-$LEVEL e.g. md-raid5
200 This can be written only while the array is being assembled, not
204 The "layout" for the array for the particular level. This is
205 simply a number that is interpretted differently by different
206 levels. It can be written while assembling an array.
209 The point at which resync should start. If no resync is needed,
210 this will be a very large number. At array creation it will
211 default to 0, though starting the array as 'clean' will
215 This file can be written but not read. The value written should
216 be a block device number as major:minor. e.g. 8:0
217 This will cause that device to be attached to the array, if it is
218 available. It will then appear at md/dev-XXX (depending on the
219 name of the device) and further configuration is then possible.
222 When an md array has seen no write requests for a certain period
223 of time, it will be marked as 'clean'. When another write
224 request arrive, the array is marked as 'dirty' before the write
225 commenses. This is known as 'safe_mode'.
226 The 'certain period' is controlled by this file which stores the
227 period as a number of seconds. The default is 200msec (0.200).
228 Writing a value of 0 disables safemode.
231 This file contains a single word which describes the current
232 state of the array. In many cases, the state can be set by
233 writing the word for the desired state, however some states
234 cannot be explicitly set, and some transitions are not allowed.
237 No devices, no size, no level
238 Writing is equivalent to STOP_ARRAY ioctl
240 May have some settings, but array is not active
241 all IO results in error
242 When written, doesn't tear down array, but just stops it
243 suspended (not supported yet)
244 All IO requests will block. The array can be reconfigured.
245 Writing this, if accepted, will block until array is quiessent
247 no resync can happen. no superblocks get written.
250 like readonly, but behaves like 'clean' on a write request.
252 clean - no pending writes, but otherwise active.
253 When written to inactive array, starts without resync
254 If a write request arrives then
255 if metadata is known, mark 'dirty' and switch to 'active'.
256 if not known, block and switch to write-pending
257 If written to an active array that has pending writes, then fails.
259 fully active: IO and resync can be happening.
260 When written to inactive array, starts with resync
263 clean, but writes are blocked waiting for 'active' to be written.
266 like active, but no writes have been seen for a while (safe_mode_delay).
271 This are similar to /proc/sys/dev/raid/speed_limit_{min,max}
272 however they only apply to the particular array.
273 If no value has been written to these, of if the word 'system'
274 is written, then the system-wide value is used. If a value,
275 in kibibytes-per-second is written, then it is used.
276 When the files are read, they show the currently active value
277 followed by "(local)" or "(system)" depending on whether it is
278 a locally set or system-wide value.
281 This shows the number of sectors that have been completed of
282 whatever the current sync_action is, followed by the number of
283 sectors in total that could need to be processed. The two
284 numbers are separated by a '/' thus effectively showing one
285 value, a fraction of the process that is complete.
288 This shows the current actual speed, in K/sec, of the current
289 sync_action. It is averaged over the last 30 seconds.
292 As component devices are added to an md array, they appear in the 'md'
293 directory as new directories named
295 where XXX is a name that the kernel knows for the device, e.g. hdb1.
296 Each directory contains:
299 a symlink to the block device in /sys/block, e.g.
300 /sys/block/md0/md/dev-hdb1/block -> ../../../../block/hdb/hdb1
303 A file containing an image of the superblock read from, or
304 written to, that device.
307 A file recording the current state of the device in the array
308 which can be a comma separated list of
309 faulty - device has been kicked from active use due to
311 in_sync - device is a fully in-sync member of the array
312 spare - device is working, but not a full member.
313 This includes spares that are in the process
314 of being recoverred to
315 This list make grow in future.
316 This can be written to.
317 Writing "faulty" simulates a failure on the device.
318 Writing "remove" removes the device from the array.
321 An approximate count of read errors that have been detected on
322 this device but have not caused the device to be evicted from
323 the array (either because they were corrected or because they
324 happened while the array was read-only). When using version-1
325 metadata, this value persists across restarts of the array.
327 This value can be written while assembling an array thus
328 providing an ongoing count for arrays with metadata managed by
332 This gives the role that the device has in the array. It will
333 either be 'none' if the device is not active in the array
334 (i.e. is a spare or has failed) or an integer less than the
335 'raid_disks' number for the array indicating which possition
336 it currently fills. This can only be set while assembling an
337 array. A device for which this is set is assumed to be working.
340 This gives the location in the device (in sectors from the
341 start) where data from the array will be stored. Any part of
342 the device before this offset us not touched, unless it is
343 used for storing metadata (Formats 1.1 and 1.2).
346 The amount of the device, after the offset, that can be used
347 for storage of data. This will normally be the same as the
348 component_size. This can be written while assembling an
349 array. If a value less than the current component_size is
350 written, component_size will be reduced to this value.
353 An active md device will also contain and entry for each active device
354 in the array. These are named
358 where 'NN' is the possition in the array, starting from 0.
359 So for a 3 drive array there will be rd0, rd1, rd2.
360 These are symbolic links to the appropriate 'dev-XXX' entry.
362 cat /sys/block/md*/md/rd*/state
363 will show 'in_sync' on every line.
367 Active md devices for levels that support data redundancy (1,4,5,6)
371 a text file that can be used to monitor and control the rebuild
372 process. It contains one word which can be one of:
373 resync - redundancy is being recalculated after unclean
375 recover - a hot spare is being built to replace a
376 failed/missing device
377 idle - nothing is happening
378 check - A full check of redundancy was requested and is
379 happening. This reads all block and checks
380 them. A repair may also happen for some raid
382 repair - A full check and repair is happening. This is
383 similar to 'resync', but was requested by the
384 user, and the write-intent bitmap is NOT used to
385 optimise the process.
387 This file is writable, and each of the strings that could be
388 read are meaningful for writing.
390 'idle' will stop an active resync/recovery etc. There is no
391 guarantee that another resync/recovery may not be automatically
392 started again, though some event will be needed to trigger
394 'resync' or 'recovery' can be used to restart the
395 corresponding operation if it was stopped with 'idle'.
396 'check' and 'repair' will start the appropriate process
397 providing the current state is 'idle'.
400 When performing 'check' and 'repair', and possibly when
401 performing 'resync', md will count the number of errors that are
402 found. The count in 'mismatch_cnt' is the number of sectors
403 that were re-written, or (for 'check') would have been
404 re-written. As most raid levels work in units of pages rather
405 than sectors, this my be larger than the number of actual errors
406 by a factor of the number of sectors in a page.
408 Each active md device may also have attributes specific to the
409 personality module that manages it.
410 These are specific to the implementation of the module and could
411 change substantially if the implementation changes.
413 These currently include
415 stripe_cache_size (currently raid5 only)
416 number of entries in the stripe cache. This is writable, but
417 there are upper and lower limits (32768, 16). Default is 128.
418 strip_cache_active (currently raid5 only)
419 number of active entries in the stripe cache