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 despite 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'. e.g. raid0, raid1,
158 raid5, linear, multipath, faulty.
159 If no raid level has been set yet (array is still being
160 assembled), the value will reflect whatever has been written
161 to it, which may be a name like the above, or may be a number
162 such as '0', '5', etc.
165 a text file with a simple number indicating the number of devices
166 in a fully functional array. If this is not yet known, the file
167 will be empty. If an array is being resized (not currently
168 possible) this will contain the larger of the old and new sizes.
169 Some raid level (RAID1) allow this value to be set while the
170 array is active. This will reconfigure the array. Otherwise
171 it can only be set while assembling an array.
174 This is the size if bytes for 'chunks' and is only relevant to
175 raid levels that involve striping (1,4,5,6,10). The address space
176 of the array is conceptually divided into chunks and consecutive
177 chunks are striped onto neighbouring devices.
178 The size should be at least PAGE_SIZE (4k) and should be a power
179 of 2. This can only be set while assembling an array
182 The "layout" for the array for the particular level. This is
183 simply a number that is interpretted differently by different
184 levels. It can be written while assembling an array.
187 This is either "none" or a sector number within the devices of
188 the array where "reshape" is up to. If this is set, the three
189 attributes mentioned above (raid_disks, chunk_size, layout) can
190 potentially have 2 values, an old and a new value. If these
191 values differ, reading the attribute returns
193 and writing will effect the 'new' value, leaving the 'old'
197 For arrays with data redundancy (i.e. not raid0, linear, faulty,
198 multipath), all components must be the same size - or at least
199 there must a size that they all provide space for. This is a key
200 part or the geometry of the array. It is measured in sectors
201 and can be read from here. Writing to this value may resize
202 the array if the personality supports it (raid1, raid5, raid6),
203 and if the component drives are large enough.
206 This indicates the format that is being used to record metadata
207 about the array. It can be 0.90 (traditional format), 1.0, 1.1,
208 1.2 (newer format in varying locations) or "none" indicating that
209 the kernel isn't managing metadata at all.
212 The point at which resync should start. If no resync is needed,
213 this will be a very large number. At array creation it will
214 default to 0, though starting the array as 'clean' will
218 This file can be written but not read. The value written should
219 be a block device number as major:minor. e.g. 8:0
220 This will cause that device to be attached to the array, if it is
221 available. It will then appear at md/dev-XXX (depending on the
222 name of the device) and further configuration is then possible.
225 When an md array has seen no write requests for a certain period
226 of time, it will be marked as 'clean'. When another write
227 request arrives, the array is marked as 'dirty' before the write
228 commences. This is known as 'safe_mode'.
229 The 'certain period' is controlled by this file which stores the
230 period as a number of seconds. The default is 200msec (0.200).
231 Writing a value of 0 disables safemode.
234 This file contains a single word which describes the current
235 state of the array. In many cases, the state can be set by
236 writing the word for the desired state, however some states
237 cannot be explicitly set, and some transitions are not allowed.
239 Select/poll works on this file. All changes except between
240 active_idle and active (which can be frequent and are not
241 very interesting) are notified. active->active_idle is
242 reported if the metadata is externally managed.
245 No devices, no size, no level
246 Writing is equivalent to STOP_ARRAY ioctl
248 May have some settings, but array is not active
249 all IO results in error
250 When written, doesn't tear down array, but just stops it
251 suspended (not supported yet)
252 All IO requests will block. The array can be reconfigured.
253 Writing this, if accepted, will block until array is quiessent
255 no resync can happen. no superblocks get written.
258 like readonly, but behaves like 'clean' on a write request.
260 clean - no pending writes, but otherwise active.
261 When written to inactive array, starts without resync
262 If a write request arrives then
263 if metadata is known, mark 'dirty' and switch to 'active'.
264 if not known, block and switch to write-pending
265 If written to an active array that has pending writes, then fails.
267 fully active: IO and resync can be happening.
268 When written to inactive array, starts with resync
271 clean, but writes are blocked waiting for 'active' to be written.
274 like active, but no writes have been seen for a while (safe_mode_delay).
277 As component devices are added to an md array, they appear in the 'md'
278 directory as new directories named
280 where XXX is a name that the kernel knows for the device, e.g. hdb1.
281 Each directory contains:
284 a symlink to the block device in /sys/block, e.g.
285 /sys/block/md0/md/dev-hdb1/block -> ../../../../block/hdb/hdb1
288 A file containing an image of the superblock read from, or
289 written to, that device.
292 A file recording the current state of the device in the array
293 which can be a comma separated list of
294 faulty - device has been kicked from active use due to
296 in_sync - device is a fully in-sync member of the array
297 writemostly - device will only be subject to read
298 requests if there are no other options.
299 This applies only to raid1 arrays.
300 spare - device is working, but not a full member.
301 This includes spares that are in the process
302 of being recovered to
303 This list may grow in future.
304 This can be written to.
305 Writing "faulty" simulates a failure on the device.
306 Writing "remove" removes the device from the array.
307 Writing "writemostly" sets the writemostly flag.
308 Writing "-writemostly" clears the writemostly flag.
311 An approximate count of read errors that have been detected on
312 this device but have not caused the device to be evicted from
313 the array (either because they were corrected or because they
314 happened while the array was read-only). When using version-1
315 metadata, this value persists across restarts of the array.
317 This value can be written while assembling an array thus
318 providing an ongoing count for arrays with metadata managed by
322 This gives the role that the device has in the array. It will
323 either be 'none' if the device is not active in the array
324 (i.e. is a spare or has failed) or an integer less than the
325 'raid_disks' number for the array indicating which position
326 it currently fills. This can only be set while assembling an
327 array. A device for which this is set is assumed to be working.
330 This gives the location in the device (in sectors from the
331 start) where data from the array will be stored. Any part of
332 the device before this offset us not touched, unless it is
333 used for storing metadata (Formats 1.1 and 1.2).
336 The amount of the device, after the offset, that can be used
337 for storage of data. This will normally be the same as the
338 component_size. This can be written while assembling an
339 array. If a value less than the current component_size is
340 written, component_size will be reduced to this value.
343 An active md device will also contain and entry for each active device
344 in the array. These are named
348 where 'NN' is the position in the array, starting from 0.
349 So for a 3 drive array there will be rd0, rd1, rd2.
350 These are symbolic links to the appropriate 'dev-XXX' entry.
352 cat /sys/block/md*/md/rd*/state
353 will show 'in_sync' on every line.
357 Active md devices for levels that support data redundancy (1,4,5,6)
361 a text file that can be used to monitor and control the rebuild
362 process. It contains one word which can be one of:
363 resync - redundancy is being recalculated after unclean
365 recover - a hot spare is being built to replace a
366 failed/missing device
367 idle - nothing is happening
368 check - A full check of redundancy was requested and is
369 happening. This reads all block and checks
370 them. A repair may also happen for some raid
372 repair - A full check and repair is happening. This is
373 similar to 'resync', but was requested by the
374 user, and the write-intent bitmap is NOT used to
375 optimise the process.
377 This file is writable, and each of the strings that could be
378 read are meaningful for writing.
380 'idle' will stop an active resync/recovery etc. There is no
381 guarantee that another resync/recovery may not be automatically
382 started again, though some event will be needed to trigger
384 'resync' or 'recovery' can be used to restart the
385 corresponding operation if it was stopped with 'idle'.
386 'check' and 'repair' will start the appropriate process
387 providing the current state is 'idle'.
389 This file responds to select/poll. Any important change in the value
390 triggers a poll event. Sometimes the value will briefly be
391 "recover" if a recovery seems to be needed, but cannot be
392 achieved. In that case, the transition to "recover" isn't
393 notified, but the transition away is.
396 This contains a count of the number of devices by which the
397 arrays is degraded. So an optimal array with show '0'. A
398 single failed/missing drive will show '1', etc.
399 This file responds to select/poll, any increase or decrease
400 in the count of missing devices will trigger an event.
403 When performing 'check' and 'repair', and possibly when
404 performing 'resync', md will count the number of errors that are
405 found. The count in 'mismatch_cnt' is the number of sectors
406 that were re-written, or (for 'check') would have been
407 re-written. As most raid levels work in units of pages rather
408 than sectors, this my be larger than the number of actual errors
409 by a factor of the number of sectors in a page.
412 If the array has a write-intent bitmap, then writing to this
413 attribute can set bits in the bitmap, indicating that a resync
414 would need to check the corresponding blocks. Either individual
415 numbers or start-end pairs can be written. Multiple numbers
416 can be separated by a space.
417 Note that the numbers are 'bit' numbers, not 'block' numbers.
418 They should be scaled by the bitmap_chunksize.
422 This are similar to /proc/sys/dev/raid/speed_limit_{min,max}
423 however they only apply to the particular array.
424 If no value has been written to these, of if the word 'system'
425 is written, then the system-wide value is used. If a value,
426 in kibibytes-per-second is written, then it is used.
427 When the files are read, they show the currently active value
428 followed by "(local)" or "(system)" depending on whether it is
429 a locally set or system-wide value.
432 This shows the number of sectors that have been completed of
433 whatever the current sync_action is, followed by the number of
434 sectors in total that could need to be processed. The two
435 numbers are separated by a '/' thus effectively showing one
436 value, a fraction of the process that is complete.
437 A 'select' on this attribute will return when resync completes,
438 when it reaches the current sync_max (below) and possibly at
442 This is a number of sectors at which point a resync/recovery
443 process will pause. When a resync is active, the value can
444 only ever be increased, never decreased. The value of 'max'
445 effectively disables the limit.
449 This shows the current actual speed, in K/sec, of the current
450 sync_action. It is averaged over the last 30 seconds.
454 The two values, given as numbers of sectors, indicate a range
455 within the array where IO will be blocked. This is currently
456 only supported for raid4/5/6.
459 Each active md device may also have attributes specific to the
460 personality module that manages it.
461 These are specific to the implementation of the module and could
462 change substantially if the implementation changes.
464 These currently include
466 stripe_cache_size (currently raid5 only)
467 number of entries in the stripe cache. This is writable, but
468 there are upper and lower limits (32768, 16). Default is 128.
469 strip_cache_active (currently raid5 only)
470 number of active entries in the stripe cache
471 preread_bypass_threshold (currently raid5 only)
472 number of times a stripe requiring preread will be bypassed by
473 a stripe that does not require preread. For fairness defaults
474 to 1. Setting this to 0 disables bypass accounting and
475 requires preread stripes to wait until all full-width stripe-
476 writes are complete. Valid values are 0 to stripe_cache_size.