]>
Commit | Line | Data |
---|---|---|
1 | .\" -*- nroff -*- | |
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. | |
8 | .TH MDADM 8 "" v4.1-rc2 | |
9 | .SH NAME | |
10 | mdadm \- manage MD devices | |
11 | .I aka | |
12 | Linux Software RAID | |
13 | ||
14 | .SH SYNOPSIS | |
15 | ||
16 | .BI mdadm " [mode] <raiddevice> [options] <component-devices>" | |
17 | ||
18 | .SH DESCRIPTION | |
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 thereof) 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 | |
24 | device failure. | |
25 | ||
26 | Linux Software RAID devices are implemented through the md (Multiple | |
27 | Devices) device driver. | |
28 | ||
29 | Currently, Linux supports | |
30 | .B LINEAR | |
31 | md devices, | |
32 | .B RAID0 | |
33 | (striping), | |
34 | .B RAID1 | |
35 | (mirroring), | |
36 | .BR RAID4 , | |
37 | .BR RAID5 , | |
38 | .BR RAID6 , | |
39 | .BR RAID10 , | |
40 | .BR MULTIPATH , | |
41 | .BR FAULTY , | |
42 | and | |
43 | .BR CONTAINER . | |
44 | ||
45 | .B MULTIPATH | |
46 | is not a Software RAID mechanism, but does involve | |
47 | multiple devices: | |
48 | each device is a path to one common physical storage device. | |
49 | New installations should not use md/multipath as it is not well | |
50 | supported and has no ongoing development. Use the Device Mapper based | |
51 | multipath-tools instead. | |
52 | ||
53 | .B FAULTY | |
54 | is also not true RAID, and it only involves one device. It | |
55 | provides a layer over a true device that can be used to inject faults. | |
56 | ||
57 | .B CONTAINER | |
58 | is different again. A | |
59 | .B CONTAINER | |
60 | is a collection of devices that are | |
61 | managed as a set. This is similar to the set of devices connected to | |
62 | a hardware RAID controller. The set of devices may contain a number | |
63 | of different RAID arrays each utilising some (or all) of the blocks from a | |
64 | number of the devices in the set. For example, two devices in a 5-device set | |
65 | might form a RAID1 using the whole devices. The remaining three might | |
66 | have a RAID5 over the first half of each device, and a RAID0 over the | |
67 | second half. | |
68 | ||
69 | With a | |
70 | .BR CONTAINER , | |
71 | there is one set of metadata that describes all of | |
72 | the arrays in the container. So when | |
73 | .I mdadm | |
74 | creates a | |
75 | .B CONTAINER | |
76 | device, the device just represents the metadata. Other normal arrays (RAID1 | |
77 | etc) can be created inside the container. | |
78 | ||
79 | .SH MODES | |
80 | mdadm has several major modes of operation: | |
81 | .TP | |
82 | .B Assemble | |
83 | Assemble the components of a previously created | |
84 | array into an active array. Components can be explicitly given | |
85 | or can be searched for. | |
86 | .I mdadm | |
87 | checks that the components | |
88 | do form a bona fide array, and can, on request, fiddle superblock | |
89 | information so as to assemble a faulty array. | |
90 | ||
91 | .TP | |
92 | .B Build | |
93 | Build an array that doesn't have per-device metadata (superblocks). For these | |
94 | sorts of arrays, | |
95 | .I mdadm | |
96 | cannot differentiate between initial creation and subsequent assembly | |
97 | of an array. It also cannot perform any checks that appropriate | |
98 | components have been requested. Because of this, the | |
99 | .B Build | |
100 | mode should only be used together with a complete understanding of | |
101 | what you are doing. | |
102 | ||
103 | .TP | |
104 | .B Create | |
105 | Create a new array with per-device metadata (superblocks). | |
106 | Appropriate metadata is written to each device, and then the array | |
107 | comprising those devices is activated. A 'resync' process is started | |
108 | to make sure that the array is consistent (e.g. both sides of a mirror | |
109 | contain the same data) but the content of the device is left otherwise | |
110 | untouched. | |
111 | The array can be used as soon as it has been created. There is no | |
112 | need to wait for the initial resync to finish. | |
113 | ||
114 | .TP | |
115 | .B "Follow or Monitor" | |
116 | Monitor one or more md devices and act on any state changes. This is | |
117 | only meaningful for RAID1, 4, 5, 6, 10 or multipath arrays, as | |
118 | only these have interesting state. RAID0 or Linear never have | |
119 | missing, spare, or failed drives, so there is nothing to monitor. | |
120 | ||
121 | .TP | |
122 | .B "Grow" | |
123 | Grow (or shrink) an array, or otherwise reshape it in some way. | |
124 | Currently supported growth options including changing the active size | |
125 | of component devices and changing the number of active devices in | |
126 | Linear and RAID levels 0/1/4/5/6, | |
127 | changing the RAID level between 0, 1, 5, and 6, and between 0 and 10, | |
128 | changing the chunk size and layout for RAID 0,4,5,6,10 as well as adding or | |
129 | removing a write-intent bitmap and changing the array's consistency policy. | |
130 | ||
131 | .TP | |
132 | .B "Incremental Assembly" | |
133 | Add a single device to an appropriate array. If the addition of the | |
134 | device makes the array runnable, the array will be started. | |
135 | This provides a convenient interface to a | |
136 | .I hot-plug | |
137 | system. As each device is detected, | |
138 | .I mdadm | |
139 | has a chance to include it in some array as appropriate. | |
140 | Optionally, when the | |
141 | .I \-\-fail | |
142 | flag is passed in we will remove the device from any active array | |
143 | instead of adding it. | |
144 | ||
145 | If a | |
146 | .B CONTAINER | |
147 | is passed to | |
148 | .I mdadm | |
149 | in this mode, then any arrays within that container will be assembled | |
150 | and started. | |
151 | ||
152 | .TP | |
153 | .B Manage | |
154 | This is for doing things to specific components of an array such as | |
155 | adding new spares and removing faulty devices. | |
156 | ||
157 | .TP | |
158 | .B Misc | |
159 | This is an 'everything else' mode that supports operations on active | |
160 | arrays, operations on component devices such as erasing old superblocks, and | |
161 | information gathering operations. | |
162 | .\"This mode allows operations on independent devices such as examine MD | |
163 | .\"superblocks, erasing old superblocks and stopping active arrays. | |
164 | ||
165 | .TP | |
166 | .B Auto-detect | |
167 | This mode does not act on a specific device or array, but rather it | |
168 | requests the Linux Kernel to activate any auto-detected arrays. | |
169 | .SH OPTIONS | |
170 | ||
171 | .SH Options for selecting a mode are: | |
172 | ||
173 | .TP | |
174 | .BR \-A ", " \-\-assemble | |
175 | Assemble a pre-existing array. | |
176 | ||
177 | .TP | |
178 | .BR \-B ", " \-\-build | |
179 | Build a legacy array without superblocks. | |
180 | ||
181 | .TP | |
182 | .BR \-C ", " \-\-create | |
183 | Create a new array. | |
184 | ||
185 | .TP | |
186 | .BR \-F ", " \-\-follow ", " \-\-monitor | |
187 | Select | |
188 | .B Monitor | |
189 | mode. | |
190 | ||
191 | .TP | |
192 | .BR \-G ", " \-\-grow | |
193 | Change the size or shape of an active array. | |
194 | ||
195 | .TP | |
196 | .BR \-I ", " \-\-incremental | |
197 | Add/remove a single device to/from an appropriate array, and possibly start the array. | |
198 | ||
199 | .TP | |
200 | .B \-\-auto-detect | |
201 | Request that the kernel starts any auto-detected arrays. This can only | |
202 | work if | |
203 | .I md | |
204 | is compiled into the kernel \(em not if it is a module. | |
205 | Arrays can be auto-detected by the kernel if all the components are in | |
206 | primary MS-DOS partitions with partition type | |
207 | .BR FD , | |
208 | and all use v0.90 metadata. | |
209 | In-kernel autodetect is not recommended for new installations. Using | |
210 | .I mdadm | |
211 | to detect and assemble arrays \(em possibly in an | |
212 | .I initrd | |
213 | \(em is substantially more flexible and should be preferred. | |
214 | ||
215 | .P | |
216 | If a device is given before any options, or if the first option is | |
217 | one of | |
218 | .BR \-\-add , | |
219 | .BR \-\-re\-add , | |
220 | .BR \-\-add\-spare , | |
221 | .BR \-\-fail , | |
222 | .BR \-\-remove , | |
223 | or | |
224 | .BR \-\-replace , | |
225 | then the MANAGE mode is assumed. | |
226 | Anything other than these will cause the | |
227 | .B Misc | |
228 | mode to be assumed. | |
229 | ||
230 | .SH Options that are not mode-specific are: | |
231 | ||
232 | .TP | |
233 | .BR \-h ", " \-\-help | |
234 | Display general help message or, after one of the above options, a | |
235 | mode-specific help message. | |
236 | ||
237 | .TP | |
238 | .B \-\-help\-options | |
239 | Display more detailed help about command line parsing and some commonly | |
240 | used options. | |
241 | ||
242 | .TP | |
243 | .BR \-V ", " \-\-version | |
244 | Print version information for mdadm. | |
245 | ||
246 | .TP | |
247 | .BR \-v ", " \-\-verbose | |
248 | Be more verbose about what is happening. This can be used twice to be | |
249 | extra-verbose. | |
250 | The extra verbosity currently only affects | |
251 | .B \-\-detail \-\-scan | |
252 | and | |
253 | .BR "\-\-examine \-\-scan" . | |
254 | ||
255 | .TP | |
256 | .BR \-q ", " \-\-quiet | |
257 | Avoid printing purely informative messages. With this, | |
258 | .I mdadm | |
259 | will be silent unless there is something really important to report. | |
260 | ||
261 | ||
262 | .TP | |
263 | .BR \-f ", " \-\-force | |
264 | Be more forceful about certain operations. See the various modes for | |
265 | the exact meaning of this option in different contexts. | |
266 | ||
267 | .TP | |
268 | .BR \-c ", " \-\-config= | |
269 | Specify the config file or directory. Default is to use | |
270 | .B /etc/mdadm.conf | |
271 | and | |
272 | .BR /etc/mdadm.conf.d , | |
273 | or if those are missing then | |
274 | .B /etc/mdadm/mdadm.conf | |
275 | and | |
276 | .BR /etc/mdadm/mdadm.conf.d . | |
277 | If the config file given is | |
278 | .B "partitions" | |
279 | then nothing will be read, but | |
280 | .I mdadm | |
281 | will act as though the config file contained exactly | |
282 | .br | |
283 | .B " DEVICE partitions containers" | |
284 | .br | |
285 | and will read | |
286 | .B /proc/partitions | |
287 | to find a list of devices to scan, and | |
288 | .B /proc/mdstat | |
289 | to find a list of containers to examine. | |
290 | If the word | |
291 | .B "none" | |
292 | is given for the config file, then | |
293 | .I mdadm | |
294 | will act as though the config file were empty. | |
295 | ||
296 | If the name given is of a directory, then | |
297 | .I mdadm | |
298 | will collect all the files contained in the directory with a name ending | |
299 | in | |
300 | .BR .conf , | |
301 | sort them lexically, and process all of those files as config files. | |
302 | ||
303 | .TP | |
304 | .BR \-s ", " \-\-scan | |
305 | Scan config file or | |
306 | .B /proc/mdstat | |
307 | for missing information. | |
308 | In general, this option gives | |
309 | .I mdadm | |
310 | permission to get any missing information (like component devices, | |
311 | array devices, array identities, and alert destination) from the | |
312 | configuration file (see previous option); | |
313 | one exception is MISC mode when using | |
314 | .B \-\-detail | |
315 | or | |
316 | .B \-\-stop, | |
317 | in which case | |
318 | .B \-\-scan | |
319 | says to get a list of array devices from | |
320 | .BR /proc/mdstat . | |
321 | ||
322 | .TP | |
323 | .BR \-e ", " \-\-metadata= | |
324 | Declare the style of RAID metadata (superblock) to be used. The | |
325 | default is {DEFAULT_METADATA} for | |
326 | .BR \-\-create , | |
327 | and to guess for other operations. | |
328 | The default can be overridden by setting the | |
329 | .B metadata | |
330 | value for the | |
331 | .B CREATE | |
332 | keyword in | |
333 | .BR mdadm.conf . | |
334 | ||
335 | Options are: | |
336 | .RS | |
337 | .ie '{DEFAULT_METADATA}'0.90' | |
338 | .IP "0, 0.90, default" | |
339 | .el | |
340 | .IP "0, 0.90" | |
341 | Use the original 0.90 format superblock. This format limits arrays to | |
342 | 28 component devices and limits component devices of levels 1 and | |
343 | greater to 2 terabytes. It is also possible for there to be confusion | |
344 | about whether the superblock applies to a whole device or just the | |
345 | last partition, if that partition starts on a 64K boundary. | |
346 | .ie '{DEFAULT_METADATA}'0.90' | |
347 | .IP "1, 1.0, 1.1, 1.2" | |
348 | .el | |
349 | .IP "1, 1.0, 1.1, 1.2 default" | |
350 | Use the new version-1 format superblock. This has fewer restrictions. | |
351 | It can easily be moved between hosts with different endian-ness, and a | |
352 | recovery operation can be checkpointed and restarted. The different | |
353 | sub-versions store the superblock at different locations on the | |
354 | device, either at the end (for 1.0), at the start (for 1.1) or 4K from | |
355 | the start (for 1.2). "1" is equivalent to "1.2" (the commonly | |
356 | preferred 1.x format). | |
357 | 'if '{DEFAULT_METADATA}'1.2' "default" is equivalent to "1.2". | |
358 | .IP ddf | |
359 | Use the "Industry Standard" DDF (Disk Data Format) format defined by | |
360 | SNIA. | |
361 | When creating a DDF array a | |
362 | .B CONTAINER | |
363 | will be created, and normal arrays can be created in that container. | |
364 | .IP imsm | |
365 | Use the Intel(R) Matrix Storage Manager metadata format. This creates a | |
366 | .B CONTAINER | |
367 | which is managed in a similar manner to DDF, and is supported by an | |
368 | option-rom on some platforms: | |
369 | .IP | |
370 | .B https://www.intel.com/design/chipsets/matrixstorage_sb.htm | |
371 | .PP | |
372 | .RE | |
373 | ||
374 | .TP | |
375 | .B \-\-homehost= | |
376 | This will override any | |
377 | .B HOMEHOST | |
378 | setting in the config file and provides the identity of the host which | |
379 | should be considered the home for any arrays. | |
380 | ||
381 | When creating an array, the | |
382 | .B homehost | |
383 | will be recorded in the metadata. For version-1 superblocks, it will | |
384 | be prefixed to the array name. For version-0.90 superblocks, part of | |
385 | the SHA1 hash of the hostname will be stored in the later half of the | |
386 | UUID. | |
387 | ||
388 | When reporting information about an array, any array which is tagged | |
389 | for the given homehost will be reported as such. | |
390 | ||
391 | When using Auto-Assemble, only arrays tagged for the given homehost | |
392 | will be allowed to use 'local' names (i.e. not ending in '_' followed | |
393 | by a digit string). See below under | |
394 | .BR "Auto Assembly" . | |
395 | ||
396 | The special name "\fBany\fP" can be used as a wild card. If an array | |
397 | is created with | |
398 | .B --homehost=any | |
399 | then the name "\fBany\fP" will be stored in the array and it can be | |
400 | assembled in the same way on any host. If an array is assembled with | |
401 | this option, then the homehost recorded on the array will be ignored. | |
402 | ||
403 | .TP | |
404 | .B \-\-prefer= | |
405 | When | |
406 | .I mdadm | |
407 | needs to print the name for a device it normally finds the name in | |
408 | .B /dev | |
409 | which refers to the device and is shortest. When a path component is | |
410 | given with | |
411 | .B \-\-prefer | |
412 | .I mdadm | |
413 | will prefer a longer name if it contains that component. For example | |
414 | .B \-\-prefer=by-uuid | |
415 | will prefer a name in a subdirectory of | |
416 | .B /dev | |
417 | called | |
418 | .BR by-uuid . | |
419 | ||
420 | This functionality is currently only provided by | |
421 | .B \-\-detail | |
422 | and | |
423 | .BR \-\-monitor . | |
424 | ||
425 | .TP | |
426 | .B \-\-home\-cluster= | |
427 | specifies the cluster name for the md device. The md device can be assembled | |
428 | only on the cluster which matches the name specified. If this option is not | |
429 | provided, mdadm tries to detect the cluster name automatically. | |
430 | ||
431 | .SH For create, build, or grow: | |
432 | ||
433 | .TP | |
434 | .BR \-n ", " \-\-raid\-devices= | |
435 | Specify the number of active devices in the array. This, plus the | |
436 | number of spare devices (see below) must equal the number of | |
437 | .I component-devices | |
438 | (including "\fBmissing\fP" devices) | |
439 | that are listed on the command line for | |
440 | .BR \-\-create . | |
441 | Setting a value of 1 is probably | |
442 | a mistake and so requires that | |
443 | .B \-\-force | |
444 | be specified first. A value of 1 will then be allowed for linear, | |
445 | multipath, RAID0 and RAID1. It is never allowed for RAID4, RAID5 or RAID6. | |
446 | .br | |
447 | This number can only be changed using | |
448 | .B \-\-grow | |
449 | for RAID1, RAID4, RAID5 and RAID6 arrays, and only on kernels which provide | |
450 | the necessary support. | |
451 | ||
452 | .TP | |
453 | .BR \-x ", " \-\-spare\-devices= | |
454 | Specify the number of spare (eXtra) devices in the initial array. | |
455 | Spares can also be added | |
456 | and removed later. The number of component devices listed | |
457 | on the command line must equal the number of RAID devices plus the | |
458 | number of spare devices. | |
459 | ||
460 | .TP | |
461 | .BR \-z ", " \-\-size= | |
462 | Amount (in Kilobytes) of space to use from each drive in RAID levels 1/4/5/6. | |
463 | This must be a multiple of the chunk size, and must leave about 128Kb | |
464 | of space at the end of the drive for the RAID superblock. | |
465 | If this is not specified | |
466 | (as it normally is not) the smallest drive (or partition) sets the | |
467 | size, though if there is a variance among the drives of greater than 1%, a warning is | |
468 | issued. | |
469 | ||
470 | A suffix of 'K', 'M', 'G' or 'T' can be given to indicate Kilobytes, | |
471 | Megabytes, Gigabytes or Terabytes respectively. | |
472 | ||
473 | Sometimes a replacement drive can be a little smaller than the | |
474 | original drives though this should be minimised by IDEMA standards. | |
475 | Such a replacement drive will be rejected by | |
476 | .IR md . | |
477 | To guard against this it can be useful to set the initial size | |
478 | slightly smaller than the smaller device with the aim that it will | |
479 | still be larger than any replacement. | |
480 | ||
481 | This value can be set with | |
482 | .B \-\-grow | |
483 | for RAID level 1/4/5/6 though | |
484 | DDF arrays may not be able to support this. | |
485 | If the array was created with a size smaller than the currently | |
486 | active drives, the extra space can be accessed using | |
487 | .BR \-\-grow . | |
488 | The size can be given as | |
489 | .B max | |
490 | which means to choose the largest size that fits on all current drives. | |
491 | ||
492 | Before reducing the size of the array (with | |
493 | .BR "\-\-grow \-\-size=" ) | |
494 | you should make sure that space isn't needed. If the device holds a | |
495 | filesystem, you would need to resize the filesystem to use less space. | |
496 | ||
497 | After reducing the array size you should check that the data stored in | |
498 | the device is still available. If the device holds a filesystem, then | |
499 | an 'fsck' of the filesystem is a minimum requirement. If there are | |
500 | problems the array can be made bigger again with no loss with another | |
501 | .B "\-\-grow \-\-size=" | |
502 | command. | |
503 | ||
504 | This value cannot be used when creating a | |
505 | .B CONTAINER | |
506 | such as with DDF and IMSM metadata, though it perfectly valid when | |
507 | creating an array inside a container. | |
508 | ||
509 | .TP | |
510 | .BR \-Z ", " \-\-array\-size= | |
511 | This is only meaningful with | |
512 | .B \-\-grow | |
513 | and its effect is not persistent: when the array is stopped and | |
514 | restarted the default array size will be restored. | |
515 | ||
516 | Setting the array-size causes the array to appear smaller to programs | |
517 | that access the data. This is particularly needed before reshaping an | |
518 | array so that it will be smaller. As the reshape is not reversible, | |
519 | but setting the size with | |
520 | .B \-\-array-size | |
521 | is, it is required that the array size is reduced as appropriate | |
522 | before the number of devices in the array is reduced. | |
523 | ||
524 | Before reducing the size of the array you should make sure that space | |
525 | isn't needed. If the device holds a filesystem, you would need to | |
526 | resize the filesystem to use less space. | |
527 | ||
528 | After reducing the array size you should check that the data stored in | |
529 | the device is still available. If the device holds a filesystem, then | |
530 | an 'fsck' of the filesystem is a minimum requirement. If there are | |
531 | problems the array can be made bigger again with no loss with another | |
532 | .B "\-\-grow \-\-array\-size=" | |
533 | command. | |
534 | ||
535 | A suffix of 'K', 'M', 'G' or 'T' can be given to indicate Kilobytes, | |
536 | Megabytes, Gigabytes or Terabytes respectively. | |
537 | A value of | |
538 | .B max | |
539 | restores the apparent size of the array to be whatever the real | |
540 | amount of available space is. | |
541 | ||
542 | Clustered arrays do not support this parameter yet. | |
543 | ||
544 | .TP | |
545 | .BR \-c ", " \-\-chunk= | |
546 | Specify chunk size of kilobytes. The default when creating an | |
547 | array is 512KB. To ensure compatibility with earlier versions, the | |
548 | default when building an array with no persistent metadata is 64KB. | |
549 | This is only meaningful for RAID0, RAID4, RAID5, RAID6, and RAID10. | |
550 | ||
551 | RAID4, RAID5, RAID6, and RAID10 require the chunk size to be a power | |
552 | of 2. In any case it must be a multiple of 4KB. | |
553 | ||
554 | A suffix of 'K', 'M', 'G' or 'T' can be given to indicate Kilobytes, | |
555 | Megabytes, Gigabytes or Terabytes respectively. | |
556 | ||
557 | .TP | |
558 | .BR \-\-rounding= | |
559 | Specify rounding factor for a Linear array. The size of each | |
560 | component will be rounded down to a multiple of this size. | |
561 | This is a synonym for | |
562 | .B \-\-chunk | |
563 | but highlights the different meaning for Linear as compared to other | |
564 | RAID levels. The default is 64K if a kernel earlier than 2.6.16 is in | |
565 | use, and is 0K (i.e. no rounding) in later kernels. | |
566 | ||
567 | .TP | |
568 | .BR \-l ", " \-\-level= | |
569 | Set RAID level. When used with | |
570 | .BR \-\-create , | |
571 | options are: linear, raid0, 0, stripe, raid1, 1, mirror, raid4, 4, | |
572 | raid5, 5, raid6, 6, raid10, 10, multipath, mp, faulty, container. | |
573 | Obviously some of these are synonymous. | |
574 | ||
575 | When a | |
576 | .B CONTAINER | |
577 | metadata type is requested, only the | |
578 | .B container | |
579 | level is permitted, and it does not need to be explicitly given. | |
580 | ||
581 | When used with | |
582 | .BR \-\-build , | |
583 | only linear, stripe, raid0, 0, raid1, multipath, mp, and faulty are valid. | |
584 | ||
585 | Can be used with | |
586 | .B \-\-grow | |
587 | to change the RAID level in some cases. See LEVEL CHANGES below. | |
588 | ||
589 | .TP | |
590 | .BR \-p ", " \-\-layout= | |
591 | This option configures the fine details of data layout for RAID5, RAID6, | |
592 | and RAID10 arrays, and controls the failure modes for | |
593 | .IR faulty . | |
594 | It can also be used for working around a kernel bug with RAID0, but generally | |
595 | doesn't need to be used explicitly. | |
596 | ||
597 | The layout of the RAID5 parity block can be one of | |
598 | .BR left\-asymmetric , | |
599 | .BR left\-symmetric , | |
600 | .BR right\-asymmetric , | |
601 | .BR right\-symmetric , | |
602 | .BR la ", " ra ", " ls ", " rs . | |
603 | The default is | |
604 | .BR left\-symmetric . | |
605 | ||
606 | It is also possible to cause RAID5 to use a RAID4-like layout by | |
607 | choosing | |
608 | .BR parity\-first , | |
609 | or | |
610 | .BR parity\-last . | |
611 | ||
612 | Finally for RAID5 there are DDF\-compatible layouts, | |
613 | .BR ddf\-zero\-restart , | |
614 | .BR ddf\-N\-restart , | |
615 | and | |
616 | .BR ddf\-N\-continue . | |
617 | ||
618 | These same layouts are available for RAID6. There are also 4 layouts | |
619 | that will provide an intermediate stage for converting between RAID5 | |
620 | and RAID6. These provide a layout which is identical to the | |
621 | corresponding RAID5 layout on the first N\-1 devices, and has the 'Q' | |
622 | syndrome (the second 'parity' block used by RAID6) on the last device. | |
623 | These layouts are: | |
624 | .BR left\-symmetric\-6 , | |
625 | .BR right\-symmetric\-6 , | |
626 | .BR left\-asymmetric\-6 , | |
627 | .BR right\-asymmetric\-6 , | |
628 | and | |
629 | .BR parity\-first\-6 . | |
630 | ||
631 | When setting the failure mode for level | |
632 | .I faulty, | |
633 | the options are: | |
634 | .BR write\-transient ", " wt , | |
635 | .BR read\-transient ", " rt , | |
636 | .BR write\-persistent ", " wp , | |
637 | .BR read\-persistent ", " rp , | |
638 | .BR write\-all , | |
639 | .BR read\-fixable ", " rf , | |
640 | .BR clear ", " flush ", " none . | |
641 | ||
642 | Each failure mode can be followed by a number, which is used as a period | |
643 | between fault generation. Without a number, the fault is generated | |
644 | once on the first relevant request. With a number, the fault will be | |
645 | generated after that many requests, and will continue to be generated | |
646 | every time the period elapses. | |
647 | ||
648 | Multiple failure modes can be current simultaneously by using the | |
649 | .B \-\-grow | |
650 | option to set subsequent failure modes. | |
651 | ||
652 | "clear" or "none" will remove any pending or periodic failure modes, | |
653 | and "flush" will clear any persistent faults. | |
654 | ||
655 | The layout options for RAID10 are one of 'n', 'o' or 'f' followed | |
656 | by a small number. The default is 'n2'. The supported options are: | |
657 | ||
658 | .I 'n' | |
659 | signals 'near' copies. Multiple copies of one data block are at | |
660 | similar offsets in different devices. | |
661 | ||
662 | .I 'o' | |
663 | signals 'offset' copies. Rather than the chunks being duplicated | |
664 | within a stripe, whole stripes are duplicated but are rotated by one | |
665 | device so duplicate blocks are on different devices. Thus subsequent | |
666 | copies of a block are in the next drive, and are one chunk further | |
667 | down. | |
668 | ||
669 | .I 'f' | |
670 | signals 'far' copies | |
671 | (multiple copies have very different offsets). | |
672 | See md(4) for more detail about 'near', 'offset', and 'far'. | |
673 | ||
674 | The number is the number of copies of each datablock. 2 is normal, 3 | |
675 | can be useful. This number can be at most equal to the number of | |
676 | devices in the array. It does not need to divide evenly into that | |
677 | number (e.g. it is perfectly legal to have an 'n2' layout for an array | |
678 | with an odd number of devices). | |
679 | ||
680 | A bug introduced in Linux 3.14 means that RAID0 arrays | |
681 | .B "with devices of differing sizes" | |
682 | started using a different layout. This could lead to | |
683 | data corruption. Since Linux 5.4 (and various stable releases that received | |
684 | backports), the kernel will not accept such an array unless | |
685 | a layout is explictly set. It can be set to | |
686 | .RB ' original ' | |
687 | or | |
688 | .RB ' alternate '. | |
689 | When creating a new array, | |
690 | .I mdadm | |
691 | will select | |
692 | .RB ' original ' | |
693 | by default, so the layout does not normally need to be set. | |
694 | An array created for either | |
695 | .RB ' original ' | |
696 | or | |
697 | .RB ' alternate ' | |
698 | will not be recognized by an (unpatched) kernel prior to 5.4. To create | |
699 | a RAID0 array with devices of differing sizes that can be used on an | |
700 | older kernel, you can set the layout to | |
701 | .RB ' dangerous '. | |
702 | This will use whichever layout the running kernel supports, so the data | |
703 | on the array may become corrupt when changing kernel from pre-3.14 to a | |
704 | later kernel. | |
705 | ||
706 | When an array is converted between RAID5 and RAID6 an intermediate | |
707 | RAID6 layout is used in which the second parity block (Q) is always on | |
708 | the last device. To convert a RAID5 to RAID6 and leave it in this new | |
709 | layout (which does not require re-striping) use | |
710 | .BR \-\-layout=preserve . | |
711 | This will try to avoid any restriping. | |
712 | ||
713 | The converse of this is | |
714 | .B \-\-layout=normalise | |
715 | which will change a non-standard RAID6 layout into a more standard | |
716 | arrangement. | |
717 | ||
718 | .TP | |
719 | .BR \-\-parity= | |
720 | same as | |
721 | .B \-\-layout | |
722 | (thus explaining the p of | |
723 | .BR \-p ). | |
724 | ||
725 | .TP | |
726 | .BR \-b ", " \-\-bitmap= | |
727 | Specify a file to store a write-intent bitmap in. The file should not | |
728 | exist unless | |
729 | .B \-\-force | |
730 | is also given. The same file should be provided | |
731 | when assembling the array. If the word | |
732 | .B "internal" | |
733 | is given, then the bitmap is stored with the metadata on the array, | |
734 | and so is replicated on all devices. If the word | |
735 | .B "none" | |
736 | is given with | |
737 | .B \-\-grow | |
738 | mode, then any bitmap that is present is removed. If the word | |
739 | .B "clustered" | |
740 | is given, the array is created for a clustered environment. One bitmap | |
741 | is created for each node as defined by the | |
742 | .B \-\-nodes | |
743 | parameter and are stored internally. | |
744 | ||
745 | To help catch typing errors, the filename must contain at least one | |
746 | slash ('/') if it is a real file (not 'internal' or 'none'). | |
747 | ||
748 | Note: external bitmaps are only known to work on ext2 and ext3. | |
749 | Storing bitmap files on other filesystems may result in serious problems. | |
750 | ||
751 | When creating an array on devices which are 100G or larger, | |
752 | .I mdadm | |
753 | automatically adds an internal bitmap as it will usually be | |
754 | beneficial. This can be suppressed with | |
755 | .B "\-\-bitmap=none" | |
756 | or by selecting a different consistency policy with | |
757 | .BR \-\-consistency\-policy . | |
758 | ||
759 | .TP | |
760 | .BR \-\-bitmap\-chunk= | |
761 | Set the chunksize of the bitmap. Each bit corresponds to that many | |
762 | Kilobytes of storage. | |
763 | When using a file based bitmap, the default is to use the smallest | |
764 | size that is at-least 4 and requires no more than 2^21 chunks. | |
765 | When using an | |
766 | .B internal | |
767 | bitmap, the chunksize defaults to 64Meg, or larger if necessary to | |
768 | fit the bitmap into the available space. | |
769 | ||
770 | A suffix of 'K', 'M', 'G' or 'T' can be given to indicate Kilobytes, | |
771 | Megabytes, Gigabytes or Terabytes respectively. | |
772 | ||
773 | .TP | |
774 | .BR \-W ", " \-\-write\-mostly | |
775 | subsequent devices listed in a | |
776 | .BR \-\-build , | |
777 | .BR \-\-create , | |
778 | or | |
779 | .B \-\-add | |
780 | command will be flagged as 'write\-mostly'. This is valid for RAID1 | |
781 | only and means that the 'md' driver will avoid reading from these | |
782 | devices if at all possible. This can be useful if mirroring over a | |
783 | slow link. | |
784 | ||
785 | .TP | |
786 | .BR \-\-write\-behind= | |
787 | Specify that write-behind mode should be enabled (valid for RAID1 | |
788 | only). If an argument is specified, it will set the maximum number | |
789 | of outstanding writes allowed. The default value is 256. | |
790 | A write-intent bitmap is required in order to use write-behind | |
791 | mode, and write-behind is only attempted on drives marked as | |
792 | .IR write-mostly . | |
793 | ||
794 | .TP | |
795 | .BR \-\-failfast | |
796 | subsequent devices listed in a | |
797 | .B \-\-create | |
798 | or | |
799 | .B \-\-add | |
800 | command will be flagged as 'failfast'. This is valid for RAID1 and | |
801 | RAID10 only. IO requests to these devices will be encouraged to fail | |
802 | quickly rather than cause long delays due to error handling. Also no | |
803 | attempt is made to repair a read error on these devices. | |
804 | ||
805 | If an array becomes degraded so that the 'failfast' device is the only | |
806 | usable device, the 'failfast' flag will then be ignored and extended | |
807 | delays will be preferred to complete failure. | |
808 | ||
809 | The 'failfast' flag is appropriate for storage arrays which have a | |
810 | low probability of true failure, but which may sometimes | |
811 | cause unacceptable delays due to internal maintenance functions. | |
812 | ||
813 | .TP | |
814 | .BR \-\-assume\-clean | |
815 | Tell | |
816 | .I mdadm | |
817 | that the array pre-existed and is known to be clean. It can be useful | |
818 | when trying to recover from a major failure as you can be sure that no | |
819 | data will be affected unless you actually write to the array. It can | |
820 | also be used when creating a RAID1 or RAID10 if you want to avoid the | |
821 | initial resync, however this practice \(em while normally safe \(em is not | |
822 | recommended. Use this only if you really know what you are doing. | |
823 | .IP | |
824 | When the devices that will be part of a new array were filled | |
825 | with zeros before creation the operator knows the array is | |
826 | actually clean. If that is the case, such as after running | |
827 | badblocks, this argument can be used to tell mdadm the | |
828 | facts the operator knows. | |
829 | .IP | |
830 | When an array is resized to a larger size with | |
831 | .B "\-\-grow \-\-size=" | |
832 | the new space is normally resynced in that same way that the whole | |
833 | array is resynced at creation. From Linux version 3.0, | |
834 | .B \-\-assume\-clean | |
835 | can be used with that command to avoid the automatic resync. | |
836 | ||
837 | .TP | |
838 | .BR \-\-backup\-file= | |
839 | This is needed when | |
840 | .B \-\-grow | |
841 | is used to increase the number of raid-devices in a RAID5 or RAID6 if | |
842 | there are no spare devices available, or to shrink, change RAID level | |
843 | or layout. See the GROW MODE section below on RAID\-DEVICES CHANGES. | |
844 | The file must be stored on a separate device, not on the RAID array | |
845 | being reshaped. | |
846 | ||
847 | .TP | |
848 | .B \-\-data\-offset= | |
849 | Arrays with 1.x metadata can leave a gap between the start of the | |
850 | device and the start of array data. This gap can be used for various | |
851 | metadata. The start of data is known as the | |
852 | .IR data\-offset . | |
853 | Normally an appropriate data offset is computed automatically. | |
854 | However it can be useful to set it explicitly such as when re-creating | |
855 | an array which was originally created using a different version of | |
856 | .I mdadm | |
857 | which computed a different offset. | |
858 | ||
859 | Setting the offset explicitly over-rides the default. The value given | |
860 | is in Kilobytes unless a suffix of 'K', 'M', 'G' or 'T' is used to explicitly | |
861 | indicate Kilobytes, Megabytes, Gigabytes or Terabytes respectively. | |
862 | ||
863 | Since Linux 3.4, | |
864 | .B \-\-data\-offset | |
865 | can also be used with | |
866 | .B --grow | |
867 | for some RAID levels (initially on RAID10). This allows the | |
868 | data\-offset to be changed as part of the reshape process. When the | |
869 | data offset is changed, no backup file is required as the difference | |
870 | in offsets is used to provide the same functionality. | |
871 | ||
872 | When the new offset is earlier than the old offset, the number of | |
873 | devices in the array cannot shrink. When it is after the old offset, | |
874 | the number of devices in the array cannot increase. | |
875 | ||
876 | When creating an array, | |
877 | .B \-\-data\-offset | |
878 | can be specified as | |
879 | .BR variable . | |
880 | In the case each member device is expected to have a offset appended | |
881 | to the name, separated by a colon. This makes it possible to recreate | |
882 | exactly an array which has varying data offsets (as can happen when | |
883 | different versions of | |
884 | .I mdadm | |
885 | are used to add different devices). | |
886 | ||
887 | .TP | |
888 | .BR \-\-continue | |
889 | This option is complementary to the | |
890 | .B \-\-freeze-reshape | |
891 | option for assembly. It is needed when | |
892 | .B \-\-grow | |
893 | operation is interrupted and it is not restarted automatically due to | |
894 | .B \-\-freeze-reshape | |
895 | usage during array assembly. This option is used together with | |
896 | .BR \-G | |
897 | , ( | |
898 | .BR \-\-grow | |
899 | ) command and device for a pending reshape to be continued. | |
900 | All parameters required for reshape continuation will be read from array metadata. | |
901 | If initial | |
902 | .BR \-\-grow | |
903 | command had required | |
904 | .BR \-\-backup\-file= | |
905 | option to be set, continuation option will require to have exactly the same | |
906 | backup file given as well. | |
907 | .IP | |
908 | Any other parameter passed together with | |
909 | .BR \-\-continue | |
910 | option will be ignored. | |
911 | ||
912 | .TP | |
913 | .BR \-N ", " \-\-name= | |
914 | Set a | |
915 | .B name | |
916 | for the array. This is currently only effective when creating an | |
917 | array with a version-1 superblock, or an array in a DDF container. | |
918 | The name is a simple textual string that can be used to identify array | |
919 | components when assembling. If name is needed but not specified, it | |
920 | is taken from the basename of the device that is being created. | |
921 | e.g. when creating | |
922 | .I /dev/md/home | |
923 | the | |
924 | .B name | |
925 | will default to | |
926 | .IR home . | |
927 | ||
928 | .TP | |
929 | .BR \-R ", " \-\-run | |
930 | Insist that | |
931 | .I mdadm | |
932 | run the array, even if some of the components | |
933 | appear to be active in another array or filesystem. Normally | |
934 | .I mdadm | |
935 | will ask for confirmation before including such components in an | |
936 | array. This option causes that question to be suppressed. | |
937 | ||
938 | .TP | |
939 | .BR \-f ", " \-\-force | |
940 | Insist that | |
941 | .I mdadm | |
942 | accept the geometry and layout specified without question. Normally | |
943 | .I mdadm | |
944 | will not allow creation of an array with only one device, and will try | |
945 | to create a RAID5 array with one missing drive (as this makes the | |
946 | initial resync work faster). With | |
947 | .BR \-\-force , | |
948 | .I mdadm | |
949 | will not try to be so clever. | |
950 | ||
951 | .TP | |
952 | .BR \-o ", " \-\-readonly | |
953 | Start the array | |
954 | .B read only | |
955 | rather than read-write as normal. No writes will be allowed to the | |
956 | array, and no resync, recovery, or reshape will be started. It works with | |
957 | Create, Assemble, Manage and Misc mode. | |
958 | ||
959 | .TP | |
960 | .BR \-a ", " "\-\-auto{=yes,md,mdp,part,p}{NN}" | |
961 | Instruct mdadm how to create the device file if needed, possibly allocating | |
962 | an unused minor number. "md" causes a non-partitionable array | |
963 | to be used (though since Linux 2.6.28, these array devices are in fact | |
964 | partitionable). "mdp", "part" or "p" causes a partitionable array (2.6 and | |
965 | later) to be used. "yes" requires the named md device to have | |
966 | a 'standard' format, and the type and minor number will be determined | |
967 | from this. With mdadm 3.0, device creation is normally left up to | |
968 | .I udev | |
969 | so this option is unlikely to be needed. | |
970 | See DEVICE NAMES below. | |
971 | ||
972 | The argument can also come immediately after | |
973 | "\-a". e.g. "\-ap". | |
974 | ||
975 | If | |
976 | .B \-\-auto | |
977 | is not given on the command line or in the config file, then | |
978 | the default will be | |
979 | .BR \-\-auto=yes . | |
980 | ||
981 | If | |
982 | .B \-\-scan | |
983 | is also given, then any | |
984 | .I auto= | |
985 | entries in the config file will override the | |
986 | .B \-\-auto | |
987 | instruction given on the command line. | |
988 | ||
989 | For partitionable arrays, | |
990 | .I mdadm | |
991 | will create the device file for the whole array and for the first 4 | |
992 | partitions. A different number of partitions can be specified at the | |
993 | end of this option (e.g. | |
994 | .BR \-\-auto=p7 ). | |
995 | If the device name ends with a digit, the partition names add a 'p', | |
996 | and a number, e.g. | |
997 | .IR /dev/md/home1p3 . | |
998 | If there is no trailing digit, then the partition names just have a | |
999 | number added, e.g. | |
1000 | .IR /dev/md/scratch3 . | |
1001 | ||
1002 | If the md device name is in a 'standard' format as described in DEVICE | |
1003 | NAMES, then it will be created, if necessary, with the appropriate | |
1004 | device number based on that name. If the device name is not in one of these | |
1005 | formats, then a unused device number will be allocated. The device | |
1006 | number will be considered unused if there is no active array for that | |
1007 | number, and there is no entry in /dev for that number and with a | |
1008 | non-standard name. Names that are not in 'standard' format are only | |
1009 | allowed in "/dev/md/". | |
1010 | ||
1011 | This is meaningful with | |
1012 | .B \-\-create | |
1013 | or | |
1014 | .BR \-\-build . | |
1015 | ||
1016 | .TP | |
1017 | .BR \-a ", " "\-\-add" | |
1018 | This option can be used in Grow mode in two cases. | |
1019 | ||
1020 | If the target array is a Linear array, then | |
1021 | .B \-\-add | |
1022 | can be used to add one or more devices to the array. They | |
1023 | are simply catenated on to the end of the array. Once added, the | |
1024 | devices cannot be removed. | |
1025 | ||
1026 | If the | |
1027 | .B \-\-raid\-disks | |
1028 | option is being used to increase the number of devices in an array, | |
1029 | then | |
1030 | .B \-\-add | |
1031 | can be used to add some extra devices to be included in the array. | |
1032 | In most cases this is not needed as the extra devices can be added as | |
1033 | spares first, and then the number of raid-disks can be changed. | |
1034 | However for RAID0, it is not possible to add spares. So to increase | |
1035 | the number of devices in a RAID0, it is necessary to set the new | |
1036 | number of devices, and to add the new devices, in the same command. | |
1037 | ||
1038 | .TP | |
1039 | .BR \-\-nodes | |
1040 | Only works when the array is for clustered environment. It specifies | |
1041 | the maximum number of nodes in the cluster that will use this device | |
1042 | simultaneously. If not specified, this defaults to 4. | |
1043 | ||
1044 | .TP | |
1045 | .BR \-\-write-journal | |
1046 | Specify journal device for the RAID-4/5/6 array. The journal device | |
1047 | should be a SSD with reasonable lifetime. | |
1048 | ||
1049 | .TP | |
1050 | .BR \-\-symlinks | |
1051 | Auto creation of symlinks in /dev to /dev/md, option --symlinks must | |
1052 | be 'no' or 'yes' and work with --create and --build. | |
1053 | ||
1054 | .TP | |
1055 | .BR \-k ", " \-\-consistency\-policy= | |
1056 | Specify how the array maintains consistency in case of unexpected shutdown. | |
1057 | Only relevant for RAID levels with redundancy. | |
1058 | Currently supported options are: | |
1059 | .RS | |
1060 | ||
1061 | .TP | |
1062 | .B resync | |
1063 | Full resync is performed and all redundancy is regenerated when the array is | |
1064 | started after unclean shutdown. | |
1065 | ||
1066 | .TP | |
1067 | .B bitmap | |
1068 | Resync assisted by a write-intent bitmap. Implicitly selected when using | |
1069 | .BR \-\-bitmap . | |
1070 | ||
1071 | .TP | |
1072 | .B journal | |
1073 | For RAID levels 4/5/6, journal device is used to log transactions and replay | |
1074 | after unclean shutdown. Implicitly selected when using | |
1075 | .BR \-\-write\-journal . | |
1076 | ||
1077 | .TP | |
1078 | .B ppl | |
1079 | For RAID5 only, Partial Parity Log is used to close the write hole and | |
1080 | eliminate resync. PPL is stored in the metadata region of RAID member drives, | |
1081 | no additional journal drive is needed. | |
1082 | ||
1083 | .PP | |
1084 | Can be used with \-\-grow to change the consistency policy of an active array | |
1085 | in some cases. See CONSISTENCY POLICY CHANGES below. | |
1086 | .RE | |
1087 | ||
1088 | ||
1089 | .SH For assemble: | |
1090 | ||
1091 | .TP | |
1092 | .BR \-u ", " \-\-uuid= | |
1093 | uuid of array to assemble. Devices which don't have this uuid are | |
1094 | excluded | |
1095 | ||
1096 | .TP | |
1097 | .BR \-m ", " \-\-super\-minor= | |
1098 | Minor number of device that array was created for. Devices which | |
1099 | don't have this minor number are excluded. If you create an array as | |
1100 | /dev/md1, then all superblocks will contain the minor number 1, even if | |
1101 | the array is later assembled as /dev/md2. | |
1102 | ||
1103 | Giving the literal word "dev" for | |
1104 | .B \-\-super\-minor | |
1105 | will cause | |
1106 | .I mdadm | |
1107 | to use the minor number of the md device that is being assembled. | |
1108 | e.g. when assembling | |
1109 | .BR /dev/md0 , | |
1110 | .B \-\-super\-minor=dev | |
1111 | will look for super blocks with a minor number of 0. | |
1112 | ||
1113 | .B \-\-super\-minor | |
1114 | is only relevant for v0.90 metadata, and should not normally be used. | |
1115 | Using | |
1116 | .B \-\-uuid | |
1117 | is much safer. | |
1118 | ||
1119 | .TP | |
1120 | .BR \-N ", " \-\-name= | |
1121 | Specify the name of the array to assemble. This must be the name | |
1122 | that was specified when creating the array. It must either match | |
1123 | the name stored in the superblock exactly, or it must match | |
1124 | with the current | |
1125 | .I homehost | |
1126 | prefixed to the start of the given name. | |
1127 | ||
1128 | .TP | |
1129 | .BR \-f ", " \-\-force | |
1130 | Assemble the array even if the metadata on some devices appears to be | |
1131 | out-of-date. If | |
1132 | .I mdadm | |
1133 | cannot find enough working devices to start the array, but can find | |
1134 | some devices that are recorded as having failed, then it will mark | |
1135 | those devices as working so that the array can be started. | |
1136 | An array which requires | |
1137 | .B \-\-force | |
1138 | to be started may contain data corruption. Use it carefully. | |
1139 | ||
1140 | .TP | |
1141 | .BR \-R ", " \-\-run | |
1142 | Attempt to start the array even if fewer drives were given than were | |
1143 | present last time the array was active. Normally if not all the | |
1144 | expected drives are found and | |
1145 | .B \-\-scan | |
1146 | is not used, then the array will be assembled but not started. | |
1147 | With | |
1148 | .B \-\-run | |
1149 | an attempt will be made to start it anyway. | |
1150 | ||
1151 | .TP | |
1152 | .B \-\-no\-degraded | |
1153 | This is the reverse of | |
1154 | .B \-\-run | |
1155 | in that it inhibits the startup of array unless all expected drives | |
1156 | are present. This is only needed with | |
1157 | .B \-\-scan, | |
1158 | and can be used if the physical connections to devices are | |
1159 | not as reliable as you would like. | |
1160 | ||
1161 | .TP | |
1162 | .BR \-a ", " "\-\-auto{=no,yes,md,mdp,part}" | |
1163 | See this option under Create and Build options. | |
1164 | ||
1165 | .TP | |
1166 | .BR \-b ", " \-\-bitmap= | |
1167 | Specify the bitmap file that was given when the array was created. If | |
1168 | an array has an | |
1169 | .B internal | |
1170 | bitmap, there is no need to specify this when assembling the array. | |
1171 | ||
1172 | .TP | |
1173 | .BR \-\-backup\-file= | |
1174 | If | |
1175 | .B \-\-backup\-file | |
1176 | was used while reshaping an array (e.g. changing number of devices or | |
1177 | chunk size) and the system crashed during the critical section, then the same | |
1178 | .B \-\-backup\-file | |
1179 | must be presented to | |
1180 | .B \-\-assemble | |
1181 | to allow possibly corrupted data to be restored, and the reshape | |
1182 | to be completed. | |
1183 | ||
1184 | .TP | |
1185 | .BR \-\-invalid\-backup | |
1186 | If the file needed for the above option is not available for any | |
1187 | reason an empty file can be given together with this option to | |
1188 | indicate that the backup file is invalid. In this case the data that | |
1189 | was being rearranged at the time of the crash could be irrecoverably | |
1190 | lost, but the rest of the array may still be recoverable. This option | |
1191 | should only be used as a last resort if there is no way to recover the | |
1192 | backup file. | |
1193 | ||
1194 | ||
1195 | .TP | |
1196 | .BR \-U ", " \-\-update= | |
1197 | Update the superblock on each device while assembling the array. The | |
1198 | argument given to this flag can be one of | |
1199 | .BR sparc2.2 , | |
1200 | .BR summaries , | |
1201 | .BR uuid , | |
1202 | .BR name , | |
1203 | .BR nodes , | |
1204 | .BR homehost , | |
1205 | .BR home-cluster , | |
1206 | .BR resync , | |
1207 | .BR byteorder , | |
1208 | .BR devicesize , | |
1209 | .BR no\-bitmap , | |
1210 | .BR bbl , | |
1211 | .BR no\-bbl , | |
1212 | .BR ppl , | |
1213 | .BR no\-ppl , | |
1214 | .BR layout\-original , | |
1215 | .BR layout\-alternate , | |
1216 | .BR metadata , | |
1217 | or | |
1218 | .BR super\-minor . | |
1219 | ||
1220 | The | |
1221 | .B sparc2.2 | |
1222 | option will adjust the superblock of an array what was created on a Sparc | |
1223 | machine running a patched 2.2 Linux kernel. This kernel got the | |
1224 | alignment of part of the superblock wrong. You can use the | |
1225 | .B "\-\-examine \-\-sparc2.2" | |
1226 | option to | |
1227 | .I mdadm | |
1228 | to see what effect this would have. | |
1229 | ||
1230 | The | |
1231 | .B super\-minor | |
1232 | option will update the | |
1233 | .B "preferred minor" | |
1234 | field on each superblock to match the minor number of the array being | |
1235 | assembled. | |
1236 | This can be useful if | |
1237 | .B \-\-examine | |
1238 | reports a different "Preferred Minor" to | |
1239 | .BR \-\-detail . | |
1240 | In some cases this update will be performed automatically | |
1241 | by the kernel driver. In particular the update happens automatically | |
1242 | at the first write to an array with redundancy (RAID level 1 or | |
1243 | greater) on a 2.6 (or later) kernel. | |
1244 | ||
1245 | The | |
1246 | .B uuid | |
1247 | option will change the uuid of the array. If a UUID is given with the | |
1248 | .B \-\-uuid | |
1249 | option that UUID will be used as a new UUID and will | |
1250 | .B NOT | |
1251 | be used to help identify the devices in the array. | |
1252 | If no | |
1253 | .B \-\-uuid | |
1254 | is given, a random UUID is chosen. | |
1255 | ||
1256 | The | |
1257 | .B name | |
1258 | option will change the | |
1259 | .I name | |
1260 | of the array as stored in the superblock. This is only supported for | |
1261 | version-1 superblocks. | |
1262 | ||
1263 | The | |
1264 | .B nodes | |
1265 | option will change the | |
1266 | .I nodes | |
1267 | of the array as stored in the bitmap superblock. This option only | |
1268 | works for a clustered environment. | |
1269 | ||
1270 | The | |
1271 | .B homehost | |
1272 | option will change the | |
1273 | .I homehost | |
1274 | as recorded in the superblock. For version-0 superblocks, this is the | |
1275 | same as updating the UUID. | |
1276 | For version-1 superblocks, this involves updating the name. | |
1277 | ||
1278 | The | |
1279 | .B home\-cluster | |
1280 | option will change the cluster name as recorded in the superblock and | |
1281 | bitmap. This option only works for clustered environment. | |
1282 | ||
1283 | The | |
1284 | .B resync | |
1285 | option will cause the array to be marked | |
1286 | .I dirty | |
1287 | meaning that any redundancy in the array (e.g. parity for RAID5, | |
1288 | copies for RAID1) may be incorrect. This will cause the RAID system | |
1289 | to perform a "resync" pass to make sure that all redundant information | |
1290 | is correct. | |
1291 | ||
1292 | The | |
1293 | .B byteorder | |
1294 | option allows arrays to be moved between machines with different | |
1295 | byte-order, such as from a big-endian machine like a Sparc or some | |
1296 | MIPS machines, to a little-endian x86_64 machine. | |
1297 | When assembling such an array for the first time after a move, giving | |
1298 | .B "\-\-update=byteorder" | |
1299 | will cause | |
1300 | .I mdadm | |
1301 | to expect superblocks to have their byteorder reversed, and will | |
1302 | correct that order before assembling the array. This is only valid | |
1303 | with original (Version 0.90) superblocks. | |
1304 | ||
1305 | The | |
1306 | .B summaries | |
1307 | option will correct the summaries in the superblock. That is the | |
1308 | counts of total, working, active, failed, and spare devices. | |
1309 | ||
1310 | The | |
1311 | .B devicesize | |
1312 | option will rarely be of use. It applies to version 1.1 and 1.2 metadata | |
1313 | only (where the metadata is at the start of the device) and is only | |
1314 | useful when the component device has changed size (typically become | |
1315 | larger). The version 1 metadata records the amount of the device that | |
1316 | can be used to store data, so if a device in a version 1.1 or 1.2 | |
1317 | array becomes larger, the metadata will still be visible, but the | |
1318 | extra space will not. In this case it might be useful to assemble the | |
1319 | array with | |
1320 | .BR \-\-update=devicesize . | |
1321 | This will cause | |
1322 | .I mdadm | |
1323 | to determine the maximum usable amount of space on each device and | |
1324 | update the relevant field in the metadata. | |
1325 | ||
1326 | The | |
1327 | .B metadata | |
1328 | option only works on v0.90 metadata arrays and will convert them to | |
1329 | v1.0 metadata. The array must not be dirty (i.e. it must not need a | |
1330 | sync) and it must not have a write-intent bitmap. | |
1331 | ||
1332 | The old metadata will remain on the devices, but will appear older | |
1333 | than the new metadata and so will usually be ignored. The old metadata | |
1334 | (or indeed the new metadata) can be removed by giving the appropriate | |
1335 | .B \-\-metadata= | |
1336 | option to | |
1337 | .BR \-\-zero\-superblock . | |
1338 | ||
1339 | The | |
1340 | .B no\-bitmap | |
1341 | option can be used when an array has an internal bitmap which is | |
1342 | corrupt in some way so that assembling the array normally fails. It | |
1343 | will cause any internal bitmap to be ignored. | |
1344 | ||
1345 | The | |
1346 | .B bbl | |
1347 | option will reserve space in each device for a bad block list. This | |
1348 | will be 4K in size and positioned near the end of any free space | |
1349 | between the superblock and the data. | |
1350 | ||
1351 | The | |
1352 | .B no\-bbl | |
1353 | option will cause any reservation of space for a bad block list to be | |
1354 | removed. If the bad block list contains entries, this will fail, as | |
1355 | removing the list could cause data corruption. | |
1356 | ||
1357 | The | |
1358 | .B ppl | |
1359 | option will enable PPL for a RAID5 array and reserve space for PPL on each | |
1360 | device. There must be enough free space between the data and superblock and a | |
1361 | write-intent bitmap or journal must not be used. | |
1362 | ||
1363 | The | |
1364 | .B no\-ppl | |
1365 | option will disable PPL in the superblock. | |
1366 | ||
1367 | The | |
1368 | .B layout\-original | |
1369 | and | |
1370 | .B layout\-alternate | |
1371 | options are for RAID0 arrays in use before Linux 5.4. If the array was being | |
1372 | used with Linux 3.13 or earlier, then to assemble the array on a new kernel, | |
1373 | .B \-\-update=layout\-original | |
1374 | must be given. If the array was created and used with a kernel from Linux 3.14 to | |
1375 | Linux 5.3, then | |
1376 | .B \-\-update=layout\-alternate | |
1377 | must be given. This only needs to be given once. Subsequent assembly of the array | |
1378 | will happen normally. | |
1379 | For more information, see | |
1380 | .IR md (4). | |
1381 | ||
1382 | .TP | |
1383 | .BR \-\-freeze\-reshape | |
1384 | Option is intended to be used in start-up scripts during initrd boot phase. | |
1385 | When array under reshape is assembled during initrd phase, this option | |
1386 | stops reshape after reshape critical section is being restored. This happens | |
1387 | before file system pivot operation and avoids loss of file system context. | |
1388 | Losing file system context would cause reshape to be broken. | |
1389 | ||
1390 | Reshape can be continued later using the | |
1391 | .B \-\-continue | |
1392 | option for the grow command. | |
1393 | ||
1394 | .TP | |
1395 | .BR \-\-symlinks | |
1396 | See this option under Create and Build options. | |
1397 | ||
1398 | .SH For Manage mode: | |
1399 | ||
1400 | .TP | |
1401 | .BR \-t ", " \-\-test | |
1402 | Unless a more serious error occurred, | |
1403 | .I mdadm | |
1404 | will exit with a status of 2 if no changes were made to the array and | |
1405 | 0 if at least one change was made. | |
1406 | This can be useful when an indirect specifier such as | |
1407 | .BR missing , | |
1408 | .B detached | |
1409 | or | |
1410 | .B faulty | |
1411 | is used in requesting an operation on the array. | |
1412 | .B \-\-test | |
1413 | will report failure if these specifiers didn't find any match. | |
1414 | ||
1415 | .TP | |
1416 | .BR \-a ", " \-\-add | |
1417 | hot-add listed devices. | |
1418 | If a device appears to have recently been part of the array | |
1419 | (possibly it failed or was removed) the device is re\-added as described | |
1420 | in the next point. | |
1421 | If that fails or the device was never part of the array, the device is | |
1422 | added as a hot-spare. | |
1423 | If the array is degraded, it will immediately start to rebuild data | |
1424 | onto that spare. | |
1425 | ||
1426 | Note that this and the following options are only meaningful on array | |
1427 | with redundancy. They don't apply to RAID0 or Linear. | |
1428 | ||
1429 | .TP | |
1430 | .BR \-\-re\-add | |
1431 | re\-add a device that was previously removed from an array. | |
1432 | If the metadata on the device reports that it is a member of the | |
1433 | array, and the slot that it used is still vacant, then the device will | |
1434 | be added back to the array in the same position. This will normally | |
1435 | cause the data for that device to be recovered. However based on the | |
1436 | event count on the device, the recovery may only require sections that | |
1437 | are flagged a write-intent bitmap to be recovered or may not require | |
1438 | any recovery at all. | |
1439 | ||
1440 | When used on an array that has no metadata (i.e. it was built with | |
1441 | .BR \-\-build) | |
1442 | it will be assumed that bitmap-based recovery is enough to make the | |
1443 | device fully consistent with the array. | |
1444 | ||
1445 | When used with v1.x metadata, | |
1446 | .B \-\-re\-add | |
1447 | can be accompanied by | |
1448 | .BR \-\-update=devicesize , | |
1449 | .BR \-\-update=bbl ", or" | |
1450 | .BR \-\-update=no\-bbl . | |
1451 | See the description of these option when used in Assemble mode for an | |
1452 | explanation of their use. | |
1453 | ||
1454 | If the device name given is | |
1455 | .B missing | |
1456 | then | |
1457 | .I mdadm | |
1458 | will try to find any device that looks like it should be | |
1459 | part of the array but isn't and will try to re\-add all such devices. | |
1460 | ||
1461 | If the device name given is | |
1462 | .B faulty | |
1463 | then | |
1464 | .I mdadm | |
1465 | will find all devices in the array that are marked | |
1466 | .BR faulty , | |
1467 | remove them and attempt to immediately re\-add them. This can be | |
1468 | useful if you are certain that the reason for failure has been | |
1469 | resolved. | |
1470 | ||
1471 | .TP | |
1472 | .B \-\-add\-spare | |
1473 | Add a device as a spare. This is similar to | |
1474 | .B \-\-add | |
1475 | except that it does not attempt | |
1476 | .B \-\-re\-add | |
1477 | first. The device will be added as a spare even if it looks like it | |
1478 | could be an recent member of the array. | |
1479 | ||
1480 | .TP | |
1481 | .BR \-r ", " \-\-remove | |
1482 | remove listed devices. They must not be active. i.e. they should | |
1483 | be failed or spare devices. | |
1484 | ||
1485 | As well as the name of a device file | |
1486 | (e.g. | |
1487 | .BR /dev/sda1 ) | |
1488 | the words | |
1489 | .BR failed , | |
1490 | .B detached | |
1491 | and names like | |
1492 | .B set-A | |
1493 | can be given to | |
1494 | .BR \-\-remove . | |
1495 | The first causes all failed device to be removed. The second causes | |
1496 | any device which is no longer connected to the system (i.e an 'open' | |
1497 | returns | |
1498 | .BR ENXIO ) | |
1499 | to be removed. | |
1500 | The third will remove a set as describe below under | |
1501 | .BR \-\-fail . | |
1502 | ||
1503 | .TP | |
1504 | .BR \-f ", " \-\-fail | |
1505 | Mark listed devices as faulty. | |
1506 | As well as the name of a device file, the word | |
1507 | .B detached | |
1508 | or a set name like | |
1509 | .B set\-A | |
1510 | can be given. The former will cause any device that has been detached from | |
1511 | the system to be marked as failed. It can then be removed. | |
1512 | ||
1513 | For RAID10 arrays where the number of copies evenly divides the number | |
1514 | of devices, the devices can be conceptually divided into sets where | |
1515 | each set contains a single complete copy of the data on the array. | |
1516 | Sometimes a RAID10 array will be configured so that these sets are on | |
1517 | separate controllers. In this case all the devices in one set can be | |
1518 | failed by giving a name like | |
1519 | .B set\-A | |
1520 | or | |
1521 | .B set\-B | |
1522 | to | |
1523 | .BR \-\-fail . | |
1524 | The appropriate set names are reported by | |
1525 | .BR \-\-detail . | |
1526 | ||
1527 | .TP | |
1528 | .BR \-\-set\-faulty | |
1529 | same as | |
1530 | .BR \-\-fail . | |
1531 | ||
1532 | .TP | |
1533 | .B \-\-replace | |
1534 | Mark listed devices as requiring replacement. As soon as a spare is | |
1535 | available, it will be rebuilt and will replace the marked device. | |
1536 | This is similar to marking a device as faulty, but the device remains | |
1537 | in service during the recovery process to increase resilience against | |
1538 | multiple failures. When the replacement process finishes, the | |
1539 | replaced device will be marked as faulty. | |
1540 | ||
1541 | .TP | |
1542 | .B \-\-with | |
1543 | This can follow a list of | |
1544 | .B \-\-replace | |
1545 | devices. The devices listed after | |
1546 | .B \-\-with | |
1547 | will be preferentially used to replace the devices listed after | |
1548 | .BR \-\-replace . | |
1549 | These device must already be spare devices in the array. | |
1550 | ||
1551 | .TP | |
1552 | .BR \-\-write\-mostly | |
1553 | Subsequent devices that are added or re\-added will have the 'write-mostly' | |
1554 | flag set. This is only valid for RAID1 and means that the 'md' driver | |
1555 | will avoid reading from these devices if possible. | |
1556 | .TP | |
1557 | .BR \-\-readwrite | |
1558 | Subsequent devices that are added or re\-added will have the 'write-mostly' | |
1559 | flag cleared. | |
1560 | .TP | |
1561 | .BR \-\-cluster\-confirm | |
1562 | Confirm the existence of the device. This is issued in response to an \-\-add | |
1563 | request by a node in a cluster. When a node adds a device it sends a message | |
1564 | to all nodes in the cluster to look for a device with a UUID. This translates | |
1565 | to a udev notification with the UUID of the device to be added and the slot | |
1566 | number. The receiving node must acknowledge this message | |
1567 | with \-\-cluster\-confirm. Valid arguments are <slot>:<devicename> in case | |
1568 | the device is found or <slot>:missing in case the device is not found. | |
1569 | ||
1570 | .TP | |
1571 | .BR \-\-add-journal | |
1572 | Add journal to an existing array, or recreate journal for RAID-4/5/6 array | |
1573 | that lost a journal device. To avoid interrupting on-going write opertions, | |
1574 | .B \-\-add-journal | |
1575 | only works for array in Read-Only state. | |
1576 | ||
1577 | .TP | |
1578 | .BR \-\-failfast | |
1579 | Subsequent devices that are added or re\-added will have | |
1580 | the 'failfast' flag set. This is only valid for RAID1 and RAID10 and | |
1581 | means that the 'md' driver will avoid long timeouts on error handling | |
1582 | where possible. | |
1583 | .TP | |
1584 | .BR \-\-nofailfast | |
1585 | Subsequent devices that are re\-added will be re\-added without | |
1586 | the 'failfast' flag set. | |
1587 | ||
1588 | .P | |
1589 | Each of these options requires that the first device listed is the array | |
1590 | to be acted upon, and the remainder are component devices to be added, | |
1591 | removed, marked as faulty, etc. Several different operations can be | |
1592 | specified for different devices, e.g. | |
1593 | .in +5 | |
1594 | mdadm /dev/md0 \-\-add /dev/sda1 \-\-fail /dev/sdb1 \-\-remove /dev/sdb1 | |
1595 | .in -5 | |
1596 | Each operation applies to all devices listed until the next | |
1597 | operation. | |
1598 | ||
1599 | If an array is using a write-intent bitmap, then devices which have | |
1600 | been removed can be re\-added in a way that avoids a full | |
1601 | reconstruction but instead just updates the blocks that have changed | |
1602 | since the device was removed. For arrays with persistent metadata | |
1603 | (superblocks) this is done automatically. For arrays created with | |
1604 | .B \-\-build | |
1605 | mdadm needs to be told that this device we removed recently with | |
1606 | .BR \-\-re\-add . | |
1607 | ||
1608 | Devices can only be removed from an array if they are not in active | |
1609 | use, i.e. that must be spares or failed devices. To remove an active | |
1610 | device, it must first be marked as | |
1611 | .B faulty. | |
1612 | ||
1613 | .SH For Misc mode: | |
1614 | ||
1615 | .TP | |
1616 | .BR \-Q ", " \-\-query | |
1617 | Examine a device to see | |
1618 | (1) if it is an md device and (2) if it is a component of an md | |
1619 | array. | |
1620 | Information about what is discovered is presented. | |
1621 | ||
1622 | .TP | |
1623 | .BR \-D ", " \-\-detail | |
1624 | Print details of one or more md devices. | |
1625 | ||
1626 | .TP | |
1627 | .BR \-\-detail\-platform | |
1628 | Print details of the platform's RAID capabilities (firmware / hardware | |
1629 | topology) for a given metadata format. If used without argument, mdadm | |
1630 | will scan all controllers looking for their capabilities. Otherwise, mdadm | |
1631 | will only look at the controller specified by the argument in form of an | |
1632 | absolute filepath or a link, e.g. | |
1633 | .IR /sys/devices/pci0000:00/0000:00:1f.2 . | |
1634 | ||
1635 | .TP | |
1636 | .BR \-Y ", " \-\-export | |
1637 | When used with | |
1638 | .BR \-\-detail , | |
1639 | .BR \-\-detail-platform , | |
1640 | .BR \-\-examine , | |
1641 | or | |
1642 | .B \-\-incremental | |
1643 | output will be formatted as | |
1644 | .B key=value | |
1645 | pairs for easy import into the environment. | |
1646 | ||
1647 | With | |
1648 | .B \-\-incremental | |
1649 | The value | |
1650 | .B MD_STARTED | |
1651 | indicates whether an array was started | |
1652 | .RB ( yes ) | |
1653 | or not, which may include a reason | |
1654 | .RB ( unsafe ", " nothing ", " no ). | |
1655 | Also the value | |
1656 | .B MD_FOREIGN | |
1657 | indicates if the array is expected on this host | |
1658 | .RB ( no ), | |
1659 | or seems to be from elsewhere | |
1660 | .RB ( yes ). | |
1661 | ||
1662 | .TP | |
1663 | .BR \-E ", " \-\-examine | |
1664 | Print contents of the metadata stored on the named device(s). | |
1665 | Note the contrast between | |
1666 | .B \-\-examine | |
1667 | and | |
1668 | .BR \-\-detail . | |
1669 | .B \-\-examine | |
1670 | applies to devices which are components of an array, while | |
1671 | .B \-\-detail | |
1672 | applies to a whole array which is currently active. | |
1673 | .TP | |
1674 | .B \-\-sparc2.2 | |
1675 | If an array was created on a SPARC machine with a 2.2 Linux kernel | |
1676 | patched with RAID support, the superblock will have been created | |
1677 | incorrectly, or at least incompatibly with 2.4 and later kernels. | |
1678 | Using the | |
1679 | .B \-\-sparc2.2 | |
1680 | flag with | |
1681 | .B \-\-examine | |
1682 | will fix the superblock before displaying it. If this appears to do | |
1683 | the right thing, then the array can be successfully assembled using | |
1684 | .BR "\-\-assemble \-\-update=sparc2.2" . | |
1685 | ||
1686 | .TP | |
1687 | .BR \-X ", " \-\-examine\-bitmap | |
1688 | Report information about a bitmap file. | |
1689 | The argument is either an external bitmap file or an array component | |
1690 | in case of an internal bitmap. Note that running this on an array | |
1691 | device (e.g. | |
1692 | .BR /dev/md0 ) | |
1693 | does not report the bitmap for that array. | |
1694 | ||
1695 | .TP | |
1696 | .B \-\-examine\-badblocks | |
1697 | List the bad-blocks recorded for the device, if a bad-blocks list has | |
1698 | been configured. Currently only | |
1699 | .B 1.x | |
1700 | metadata supports bad-blocks lists. | |
1701 | ||
1702 | .TP | |
1703 | .BI \-\-dump= directory | |
1704 | .TP | |
1705 | .BI \-\-restore= directory | |
1706 | Save metadata from lists devices, or restore metadata to listed devices. | |
1707 | ||
1708 | .TP | |
1709 | .BR \-R ", " \-\-run | |
1710 | start a partially assembled array. If | |
1711 | .B \-\-assemble | |
1712 | did not find enough devices to fully start the array, it might leaving | |
1713 | it partially assembled. If you wish, you can then use | |
1714 | .B \-\-run | |
1715 | to start the array in degraded mode. | |
1716 | ||
1717 | .TP | |
1718 | .BR \-S ", " \-\-stop | |
1719 | deactivate array, releasing all resources. | |
1720 | ||
1721 | .TP | |
1722 | .BR \-o ", " \-\-readonly | |
1723 | mark array as readonly. | |
1724 | ||
1725 | .TP | |
1726 | .BR \-w ", " \-\-readwrite | |
1727 | mark array as readwrite. | |
1728 | ||
1729 | .TP | |
1730 | .B \-\-zero\-superblock | |
1731 | If the device contains a valid md superblock, the block is | |
1732 | overwritten with zeros. With | |
1733 | .B \-\-force | |
1734 | the block where the superblock would be is overwritten even if it | |
1735 | doesn't appear to be valid. | |
1736 | ||
1737 | .B Note: | |
1738 | Be careful to call \-\-zero\-superblock with clustered raid, make sure | |
1739 | array isn't used or assembled in other cluster node before execute it. | |
1740 | ||
1741 | .TP | |
1742 | .B \-\-kill\-subarray= | |
1743 | If the device is a container and the argument to \-\-kill\-subarray | |
1744 | specifies an inactive subarray in the container, then the subarray is | |
1745 | deleted. Deleting all subarrays will leave an 'empty-container' or | |
1746 | spare superblock on the drives. See | |
1747 | .B \-\-zero\-superblock | |
1748 | for completely | |
1749 | removing a superblock. Note that some formats depend on the subarray | |
1750 | index for generating a UUID, this command will fail if it would change | |
1751 | the UUID of an active subarray. | |
1752 | ||
1753 | .TP | |
1754 | .B \-\-update\-subarray= | |
1755 | If the device is a container and the argument to \-\-update\-subarray | |
1756 | specifies a subarray in the container, then attempt to update the given | |
1757 | superblock field in the subarray. See below in | |
1758 | .B MISC MODE | |
1759 | for details. | |
1760 | ||
1761 | .TP | |
1762 | .BR \-t ", " \-\-test | |
1763 | When used with | |
1764 | .BR \-\-detail , | |
1765 | the exit status of | |
1766 | .I mdadm | |
1767 | is set to reflect the status of the device. See below in | |
1768 | .B MISC MODE | |
1769 | for details. | |
1770 | ||
1771 | .TP | |
1772 | .BR \-W ", " \-\-wait | |
1773 | For each md device given, wait for any resync, recovery, or reshape | |
1774 | activity to finish before returning. | |
1775 | .I mdadm | |
1776 | will return with success if it actually waited for every device | |
1777 | listed, otherwise it will return failure. | |
1778 | ||
1779 | .TP | |
1780 | .BR \-\-wait\-clean | |
1781 | For each md device given, or each device in /proc/mdstat if | |
1782 | .B \-\-scan | |
1783 | is given, arrange for the array to be marked clean as soon as possible. | |
1784 | .I mdadm | |
1785 | will return with success if the array uses external metadata and we | |
1786 | successfully waited. For native arrays this returns immediately as the | |
1787 | kernel handles dirty-clean transitions at shutdown. No action is taken | |
1788 | if safe-mode handling is disabled. | |
1789 | ||
1790 | .TP | |
1791 | .B \-\-action= | |
1792 | Set the "sync_action" for all md devices given to one of | |
1793 | .BR idle , | |
1794 | .BR frozen , | |
1795 | .BR check , | |
1796 | .BR repair . | |
1797 | Setting to | |
1798 | .B idle | |
1799 | will abort any currently running action though some actions will | |
1800 | automatically restart. | |
1801 | Setting to | |
1802 | .B frozen | |
1803 | will abort any current action and ensure no other action starts | |
1804 | automatically. | |
1805 | ||
1806 | Details of | |
1807 | .B check | |
1808 | and | |
1809 | .B repair | |
1810 | can be found it | |
1811 | .IR md (4) | |
1812 | under | |
1813 | .BR "SCRUBBING AND MISMATCHES" . | |
1814 | ||
1815 | .SH For Incremental Assembly mode: | |
1816 | .TP | |
1817 | .BR \-\-rebuild\-map ", " \-r | |
1818 | Rebuild the map file | |
1819 | .RB ( {MAP_PATH} ) | |
1820 | that | |
1821 | .I mdadm | |
1822 | uses to help track which arrays are currently being assembled. | |
1823 | ||
1824 | .TP | |
1825 | .BR \-\-run ", " \-R | |
1826 | Run any array assembled as soon as a minimal number of devices are | |
1827 | available, rather than waiting until all expected devices are present. | |
1828 | ||
1829 | .TP | |
1830 | .BR \-\-scan ", " \-s | |
1831 | Only meaningful with | |
1832 | .B \-R | |
1833 | this will scan the | |
1834 | .B map | |
1835 | file for arrays that are being incrementally assembled and will try to | |
1836 | start any that are not already started. If any such array is listed | |
1837 | in | |
1838 | .B mdadm.conf | |
1839 | as requiring an external bitmap, that bitmap will be attached first. | |
1840 | ||
1841 | .TP | |
1842 | .BR \-\-fail ", " \-f | |
1843 | This allows the hot-plug system to remove devices that have fully disappeared | |
1844 | from the kernel. It will first fail and then remove the device from any | |
1845 | array it belongs to. | |
1846 | The device name given should be a kernel device name such as "sda", | |
1847 | not a name in | |
1848 | .IR /dev . | |
1849 | ||
1850 | .TP | |
1851 | .BR \-\-path= | |
1852 | Only used with \-\-fail. The 'path' given will be recorded so that if | |
1853 | a new device appears at the same location it can be automatically | |
1854 | added to the same array. This allows the failed device to be | |
1855 | automatically replaced by a new device without metadata if it appears | |
1856 | at specified path. This option is normally only set by a | |
1857 | .I udev | |
1858 | script. | |
1859 | ||
1860 | .SH For Monitor mode: | |
1861 | .TP | |
1862 | .BR \-m ", " \-\-mail | |
1863 | Give a mail address to send alerts to. | |
1864 | ||
1865 | .TP | |
1866 | .BR \-p ", " \-\-program ", " \-\-alert | |
1867 | Give a program to be run whenever an event is detected. | |
1868 | ||
1869 | .TP | |
1870 | .BR \-y ", " \-\-syslog | |
1871 | Cause all events to be reported through 'syslog'. The messages have | |
1872 | facility of 'daemon' and varying priorities. | |
1873 | ||
1874 | .TP | |
1875 | .BR \-d ", " \-\-delay | |
1876 | Give a delay in seconds. | |
1877 | .I mdadm | |
1878 | polls the md arrays and then waits this many seconds before polling | |
1879 | again. The default is 60 seconds. Since 2.6.16, there is no need to | |
1880 | reduce this as the kernel alerts | |
1881 | .I mdadm | |
1882 | immediately when there is any change. | |
1883 | ||
1884 | .TP | |
1885 | .BR \-r ", " \-\-increment | |
1886 | Give a percentage increment. | |
1887 | .I mdadm | |
1888 | will generate RebuildNN events with the given percentage increment. | |
1889 | ||
1890 | .TP | |
1891 | .BR \-f ", " \-\-daemonise | |
1892 | Tell | |
1893 | .I mdadm | |
1894 | to run as a background daemon if it decides to monitor anything. This | |
1895 | causes it to fork and run in the child, and to disconnect from the | |
1896 | terminal. The process id of the child is written to stdout. | |
1897 | This is useful with | |
1898 | .B \-\-scan | |
1899 | which will only continue monitoring if a mail address or alert program | |
1900 | is found in the config file. | |
1901 | ||
1902 | .TP | |
1903 | .BR \-i ", " \-\-pid\-file | |
1904 | When | |
1905 | .I mdadm | |
1906 | is running in daemon mode, write the pid of the daemon process to | |
1907 | the specified file, instead of printing it on standard output. | |
1908 | ||
1909 | .TP | |
1910 | .BR \-1 ", " \-\-oneshot | |
1911 | Check arrays only once. This will generate | |
1912 | .B NewArray | |
1913 | events and more significantly | |
1914 | .B DegradedArray | |
1915 | and | |
1916 | .B SparesMissing | |
1917 | events. Running | |
1918 | .in +5 | |
1919 | .B " mdadm \-\-monitor \-\-scan \-1" | |
1920 | .in -5 | |
1921 | from a cron script will ensure regular notification of any degraded arrays. | |
1922 | ||
1923 | .TP | |
1924 | .BR \-t ", " \-\-test | |
1925 | Generate a | |
1926 | .B TestMessage | |
1927 | alert for every array found at startup. This alert gets mailed and | |
1928 | passed to the alert program. This can be used for testing that alert | |
1929 | message do get through successfully. | |
1930 | ||
1931 | .TP | |
1932 | .BR \-\-no\-sharing | |
1933 | This inhibits the functionality for moving spares between arrays. | |
1934 | Only one monitoring process started with | |
1935 | .B \-\-scan | |
1936 | but without this flag is allowed, otherwise the two could interfere | |
1937 | with each other. | |
1938 | ||
1939 | .SH ASSEMBLE MODE | |
1940 | ||
1941 | .HP 12 | |
1942 | Usage: | |
1943 | .B mdadm \-\-assemble | |
1944 | .I md-device options-and-component-devices... | |
1945 | .HP 12 | |
1946 | Usage: | |
1947 | .B mdadm \-\-assemble \-\-scan | |
1948 | .I md-devices-and-options... | |
1949 | .HP 12 | |
1950 | Usage: | |
1951 | .B mdadm \-\-assemble \-\-scan | |
1952 | .I options... | |
1953 | ||
1954 | .PP | |
1955 | This usage assembles one or more RAID arrays from pre-existing components. | |
1956 | For each array, mdadm needs to know the md device, the identity of the | |
1957 | array, and a number of component-devices. These can be found in a number of ways. | |
1958 | ||
1959 | In the first usage example (without the | |
1960 | .BR \-\-scan ) | |
1961 | the first device given is the md device. | |
1962 | In the second usage example, all devices listed are treated as md | |
1963 | devices and assembly is attempted. | |
1964 | In the third (where no devices are listed) all md devices that are | |
1965 | listed in the configuration file are assembled. If no arrays are | |
1966 | described by the configuration file, then any arrays that | |
1967 | can be found on unused devices will be assembled. | |
1968 | ||
1969 | If precisely one device is listed, but | |
1970 | .B \-\-scan | |
1971 | is not given, then | |
1972 | .I mdadm | |
1973 | acts as though | |
1974 | .B \-\-scan | |
1975 | was given and identity information is extracted from the configuration file. | |
1976 | ||
1977 | The identity can be given with the | |
1978 | .B \-\-uuid | |
1979 | option, the | |
1980 | .B \-\-name | |
1981 | option, or the | |
1982 | .B \-\-super\-minor | |
1983 | option, will be taken from the md-device record in the config file, or | |
1984 | will be taken from the super block of the first component-device | |
1985 | listed on the command line. | |
1986 | ||
1987 | Devices can be given on the | |
1988 | .B \-\-assemble | |
1989 | command line or in the config file. Only devices which have an md | |
1990 | superblock which contains the right identity will be considered for | |
1991 | any array. | |
1992 | ||
1993 | The config file is only used if explicitly named with | |
1994 | .B \-\-config | |
1995 | or requested with (a possibly implicit) | |
1996 | .BR \-\-scan . | |
1997 | In the later case, | |
1998 | .B /etc/mdadm.conf | |
1999 | or | |
2000 | .B /etc/mdadm/mdadm.conf | |
2001 | is used. | |
2002 | ||
2003 | If | |
2004 | .B \-\-scan | |
2005 | is not given, then the config file will only be used to find the | |
2006 | identity of md arrays. | |
2007 | ||
2008 | Normally the array will be started after it is assembled. However if | |
2009 | .B \-\-scan | |
2010 | is not given and not all expected drives were listed, then the array | |
2011 | is not started (to guard against usage errors). To insist that the | |
2012 | array be started in this case (as may work for RAID1, 4, 5, 6, or 10), | |
2013 | give the | |
2014 | .B \-\-run | |
2015 | flag. | |
2016 | ||
2017 | If | |
2018 | .I udev | |
2019 | is active, | |
2020 | .I mdadm | |
2021 | does not create any entries in | |
2022 | .B /dev | |
2023 | but leaves that to | |
2024 | .IR udev . | |
2025 | It does record information in | |
2026 | .B {MAP_PATH} | |
2027 | which will allow | |
2028 | .I udev | |
2029 | to choose the correct name. | |
2030 | ||
2031 | If | |
2032 | .I mdadm | |
2033 | detects that udev is not configured, it will create the devices in | |
2034 | .B /dev | |
2035 | itself. | |
2036 | ||
2037 | In Linux kernels prior to version 2.6.28 there were two distinctly | |
2038 | different types of md devices that could be created: one that could be | |
2039 | partitioned using standard partitioning tools and one that could not. | |
2040 | Since 2.6.28 that distinction is no longer relevant as both type of | |
2041 | devices can be partitioned. | |
2042 | .I mdadm | |
2043 | will normally create the type that originally could not be partitioned | |
2044 | as it has a well defined major number (9). | |
2045 | ||
2046 | Prior to 2.6.28, it is important that mdadm chooses the correct type | |
2047 | of array device to use. This can be controlled with the | |
2048 | .B \-\-auto | |
2049 | option. In particular, a value of "mdp" or "part" or "p" tells mdadm | |
2050 | to use a partitionable device rather than the default. | |
2051 | ||
2052 | In the no-udev case, the value given to | |
2053 | .B \-\-auto | |
2054 | can be suffixed by a number. This tells | |
2055 | .I mdadm | |
2056 | to create that number of partition devices rather than the default of 4. | |
2057 | ||
2058 | The value given to | |
2059 | .B \-\-auto | |
2060 | can also be given in the configuration file as a word starting | |
2061 | .B auto= | |
2062 | on the ARRAY line for the relevant array. | |
2063 | ||
2064 | .SS Auto Assembly | |
2065 | When | |
2066 | .B \-\-assemble | |
2067 | is used with | |
2068 | .B \-\-scan | |
2069 | and no devices are listed, | |
2070 | .I mdadm | |
2071 | will first attempt to assemble all the arrays listed in the config | |
2072 | file. | |
2073 | ||
2074 | If no arrays are listed in the config (other than those marked | |
2075 | .BR <ignore> ) | |
2076 | it will look through the available devices for possible arrays and | |
2077 | will try to assemble anything that it finds. Arrays which are tagged | |
2078 | as belonging to the given homehost will be assembled and started | |
2079 | normally. Arrays which do not obviously belong to this host are given | |
2080 | names that are expected not to conflict with anything local, and are | |
2081 | started "read-auto" so that nothing is written to any device until the | |
2082 | array is written to. i.e. automatic resync etc is delayed. | |
2083 | ||
2084 | If | |
2085 | .I mdadm | |
2086 | finds a consistent set of devices that look like they should comprise | |
2087 | an array, and if the superblock is tagged as belonging to the given | |
2088 | home host, it will automatically choose a device name and try to | |
2089 | assemble the array. If the array uses version-0.90 metadata, then the | |
2090 | .B minor | |
2091 | number as recorded in the superblock is used to create a name in | |
2092 | .B /dev/md/ | |
2093 | so for example | |
2094 | .BR /dev/md/3 . | |
2095 | If the array uses version-1 metadata, then the | |
2096 | .B name | |
2097 | from the superblock is used to similarly create a name in | |
2098 | .B /dev/md/ | |
2099 | (the name will have any 'host' prefix stripped first). | |
2100 | ||
2101 | This behaviour can be modified by the | |
2102 | .I AUTO | |
2103 | line in the | |
2104 | .I mdadm.conf | |
2105 | configuration file. This line can indicate that specific metadata | |
2106 | type should, or should not, be automatically assembled. If an array | |
2107 | is found which is not listed in | |
2108 | .I mdadm.conf | |
2109 | and has a metadata format that is denied by the | |
2110 | .I AUTO | |
2111 | line, then it will not be assembled. | |
2112 | The | |
2113 | .I AUTO | |
2114 | line can also request that all arrays identified as being for this | |
2115 | homehost should be assembled regardless of their metadata type. | |
2116 | See | |
2117 | .IR mdadm.conf (5) | |
2118 | for further details. | |
2119 | ||
2120 | Note: Auto assembly cannot be used for assembling and activating some | |
2121 | arrays which are undergoing reshape. In particular as the | |
2122 | .B backup\-file | |
2123 | cannot be given, any reshape which requires a backup-file to continue | |
2124 | cannot be started by auto assembly. An array which is growing to more | |
2125 | devices and has passed the critical section can be assembled using | |
2126 | auto-assembly. | |
2127 | ||
2128 | .SH BUILD MODE | |
2129 | ||
2130 | .HP 12 | |
2131 | Usage: | |
2132 | .B mdadm \-\-build | |
2133 | .I md-device | |
2134 | .BI \-\-chunk= X | |
2135 | .BI \-\-level= Y | |
2136 | .BI \-\-raid\-devices= Z | |
2137 | .I devices | |
2138 | ||
2139 | .PP | |
2140 | This usage is similar to | |
2141 | .BR \-\-create . | |
2142 | The difference is that it creates an array without a superblock. With | |
2143 | these arrays there is no difference between initially creating the array and | |
2144 | subsequently assembling the array, except that hopefully there is useful | |
2145 | data there in the second case. | |
2146 | ||
2147 | The level may raid0, linear, raid1, raid10, multipath, or faulty, or | |
2148 | one of their synonyms. All devices must be listed and the array will | |
2149 | be started once complete. It will often be appropriate to use | |
2150 | .B \-\-assume\-clean | |
2151 | with levels raid1 or raid10. | |
2152 | ||
2153 | .SH CREATE MODE | |
2154 | ||
2155 | .HP 12 | |
2156 | Usage: | |
2157 | .B mdadm \-\-create | |
2158 | .I md-device | |
2159 | .BI \-\-chunk= X | |
2160 | .BI \-\-level= Y | |
2161 | .br | |
2162 | .BI \-\-raid\-devices= Z | |
2163 | .I devices | |
2164 | ||
2165 | .PP | |
2166 | This usage will initialise a new md array, associate some devices with | |
2167 | it, and activate the array. | |
2168 | ||
2169 | The named device will normally not exist when | |
2170 | .I "mdadm \-\-create" | |
2171 | is run, but will be created by | |
2172 | .I udev | |
2173 | once the array becomes active. | |
2174 | ||
2175 | As devices are added, they are checked to see if they contain RAID | |
2176 | superblocks or filesystems. They are also checked to see if the variance in | |
2177 | device size exceeds 1%. | |
2178 | ||
2179 | If any discrepancy is found, the array will not automatically be run, though | |
2180 | the presence of a | |
2181 | .B \-\-run | |
2182 | can override this caution. | |
2183 | ||
2184 | To create a "degraded" array in which some devices are missing, simply | |
2185 | give the word "\fBmissing\fP" | |
2186 | in place of a device name. This will cause | |
2187 | .I mdadm | |
2188 | to leave the corresponding slot in the array empty. | |
2189 | For a RAID4 or RAID5 array at most one slot can be | |
2190 | "\fBmissing\fP"; for a RAID6 array at most two slots. | |
2191 | For a RAID1 array, only one real device needs to be given. All of the | |
2192 | others can be | |
2193 | "\fBmissing\fP". | |
2194 | ||
2195 | When creating a RAID5 array, | |
2196 | .I mdadm | |
2197 | will automatically create a degraded array with an extra spare drive. | |
2198 | This is because building the spare into a degraded array is in general | |
2199 | faster than resyncing the parity on a non-degraded, but not clean, | |
2200 | array. This feature can be overridden with the | |
2201 | .B \-\-force | |
2202 | option. | |
2203 | ||
2204 | When creating an array with version-1 metadata a name for the array is | |
2205 | required. | |
2206 | If this is not given with the | |
2207 | .B \-\-name | |
2208 | option, | |
2209 | .I mdadm | |
2210 | will choose a name based on the last component of the name of the | |
2211 | device being created. So if | |
2212 | .B /dev/md3 | |
2213 | is being created, then the name | |
2214 | .B 3 | |
2215 | will be chosen. | |
2216 | If | |
2217 | .B /dev/md/home | |
2218 | is being created, then the name | |
2219 | .B home | |
2220 | will be used. | |
2221 | ||
2222 | When creating a partition based array, using | |
2223 | .I mdadm | |
2224 | with version-1.x metadata, the partition type should be set to | |
2225 | .B 0xDA | |
2226 | (non fs-data). This type selection allows for greater precision since | |
2227 | using any other [RAID auto-detect (0xFD) or a GNU/Linux partition (0x83)], | |
2228 | might create problems in the event of array recovery through a live cdrom. | |
2229 | ||
2230 | A new array will normally get a randomly assigned 128bit UUID which is | |
2231 | very likely to be unique. If you have a specific need, you can choose | |
2232 | a UUID for the array by giving the | |
2233 | .B \-\-uuid= | |
2234 | option. Be warned that creating two arrays with the same UUID is a | |
2235 | recipe for disaster. Also, using | |
2236 | .B \-\-uuid= | |
2237 | when creating a v0.90 array will silently override any | |
2238 | .B \-\-homehost= | |
2239 | setting. | |
2240 | .\"If the | |
2241 | .\".B \-\-size | |
2242 | .\"option is given, it is not necessary to list any component-devices in this command. | |
2243 | .\"They can be added later, before a | |
2244 | .\".B \-\-run. | |
2245 | .\"If no | |
2246 | .\".B \-\-size | |
2247 | .\"is given, the apparent size of the smallest drive given is used. | |
2248 | ||
2249 | If the array type supports a write-intent bitmap, and if the devices | |
2250 | in the array exceed 100G is size, an internal write-intent bitmap | |
2251 | will automatically be added unless some other option is explicitly | |
2252 | requested with the | |
2253 | .B \-\-bitmap | |
2254 | option or a different consistency policy is selected with the | |
2255 | .B \-\-consistency\-policy | |
2256 | option. In any case space for a bitmap will be reserved so that one | |
2257 | can be added later with | |
2258 | .BR "\-\-grow \-\-bitmap=internal" . | |
2259 | ||
2260 | If the metadata type supports it (currently only 1.x and IMSM metadata), | |
2261 | space will be allocated to store a bad block list. This allows a modest | |
2262 | number of bad blocks to be recorded, allowing the drive to remain in | |
2263 | service while only partially functional. | |
2264 | ||
2265 | When creating an array within a | |
2266 | .B CONTAINER | |
2267 | .I mdadm | |
2268 | can be given either the list of devices to use, or simply the name of | |
2269 | the container. The former case gives control over which devices in | |
2270 | the container will be used for the array. The latter case allows | |
2271 | .I mdadm | |
2272 | to automatically choose which devices to use based on how much spare | |
2273 | space is available. | |
2274 | ||
2275 | The General Management options that are valid with | |
2276 | .B \-\-create | |
2277 | are: | |
2278 | .TP | |
2279 | .B \-\-run | |
2280 | insist on running the array even if some devices look like they might | |
2281 | be in use. | |
2282 | ||
2283 | .TP | |
2284 | .B \-\-readonly | |
2285 | start the array in readonly mode. | |
2286 | ||
2287 | .SH MANAGE MODE | |
2288 | .HP 12 | |
2289 | Usage: | |
2290 | .B mdadm | |
2291 | .I device | |
2292 | .I options... devices... | |
2293 | .PP | |
2294 | ||
2295 | This usage will allow individual devices in an array to be failed, | |
2296 | removed or added. It is possible to perform multiple operations with | |
2297 | on command. For example: | |
2298 | .br | |
2299 | .B " mdadm /dev/md0 \-f /dev/hda1 \-r /dev/hda1 \-a /dev/hda1" | |
2300 | .br | |
2301 | will firstly mark | |
2302 | .B /dev/hda1 | |
2303 | as faulty in | |
2304 | .B /dev/md0 | |
2305 | and will then remove it from the array and finally add it back | |
2306 | in as a spare. However only one md array can be affected by a single | |
2307 | command. | |
2308 | ||
2309 | When a device is added to an active array, mdadm checks to see if it | |
2310 | has metadata on it which suggests that it was recently a member of the | |
2311 | array. If it does, it tries to "re\-add" the device. If there have | |
2312 | been no changes since the device was removed, or if the array has a | |
2313 | write-intent bitmap which has recorded whatever changes there were, | |
2314 | then the device will immediately become a full member of the array and | |
2315 | those differences recorded in the bitmap will be resolved. | |
2316 | ||
2317 | .SH MISC MODE | |
2318 | .HP 12 | |
2319 | Usage: | |
2320 | .B mdadm | |
2321 | .I options ... | |
2322 | .I devices ... | |
2323 | .PP | |
2324 | ||
2325 | MISC mode includes a number of distinct operations that | |
2326 | operate on distinct devices. The operations are: | |
2327 | .TP | |
2328 | .B \-\-query | |
2329 | The device is examined to see if it is | |
2330 | (1) an active md array, or | |
2331 | (2) a component of an md array. | |
2332 | The information discovered is reported. | |
2333 | ||
2334 | .TP | |
2335 | .B \-\-detail | |
2336 | The device should be an active md device. | |
2337 | .B mdadm | |
2338 | will display a detailed description of the array. | |
2339 | .B \-\-brief | |
2340 | or | |
2341 | .B \-\-scan | |
2342 | will cause the output to be less detailed and the format to be | |
2343 | suitable for inclusion in | |
2344 | .BR mdadm.conf . | |
2345 | The exit status of | |
2346 | .I mdadm | |
2347 | will normally be 0 unless | |
2348 | .I mdadm | |
2349 | failed to get useful information about the device(s); however, if the | |
2350 | .B \-\-test | |
2351 | option is given, then the exit status will be: | |
2352 | .RS | |
2353 | .TP | |
2354 | 0 | |
2355 | The array is functioning normally. | |
2356 | .TP | |
2357 | 1 | |
2358 | The array has at least one failed device. | |
2359 | .TP | |
2360 | 2 | |
2361 | The array has multiple failed devices such that it is unusable. | |
2362 | .TP | |
2363 | 4 | |
2364 | There was an error while trying to get information about the device. | |
2365 | .RE | |
2366 | ||
2367 | .TP | |
2368 | .B \-\-detail\-platform | |
2369 | Print detail of the platform's RAID capabilities (firmware / hardware | |
2370 | topology). If the metadata is specified with | |
2371 | .B \-e | |
2372 | or | |
2373 | .B \-\-metadata= | |
2374 | then the return status will be: | |
2375 | .RS | |
2376 | .TP | |
2377 | 0 | |
2378 | metadata successfully enumerated its platform components on this system | |
2379 | .TP | |
2380 | 1 | |
2381 | metadata is platform independent | |
2382 | .TP | |
2383 | 2 | |
2384 | metadata failed to find its platform components on this system | |
2385 | .RE | |
2386 | ||
2387 | .TP | |
2388 | .B \-\-update\-subarray= | |
2389 | If the device is a container and the argument to \-\-update\-subarray | |
2390 | specifies a subarray in the container, then attempt to update the given | |
2391 | superblock field in the subarray. Similar to updating an array in | |
2392 | "assemble" mode, the field to update is selected by | |
2393 | .B \-U | |
2394 | or | |
2395 | .B \-\-update= | |
2396 | option. The supported options are | |
2397 | .BR name , | |
2398 | .B ppl | |
2399 | and | |
2400 | .BR no\-ppl . | |
2401 | ||
2402 | The | |
2403 | .B name | |
2404 | option updates the subarray name in the metadata, it may not affect the | |
2405 | device node name or the device node symlink until the subarray is | |
2406 | re\-assembled. If updating | |
2407 | .B name | |
2408 | would change the UUID of an active subarray this operation is blocked, | |
2409 | and the command will end in an error. | |
2410 | ||
2411 | The | |
2412 | .B ppl | |
2413 | and | |
2414 | .B no\-ppl | |
2415 | options enable and disable PPL in the metadata. Currently supported only for | |
2416 | IMSM subarrays. | |
2417 | ||
2418 | .TP | |
2419 | .B \-\-examine | |
2420 | The device should be a component of an md array. | |
2421 | .I mdadm | |
2422 | will read the md superblock of the device and display the contents. | |
2423 | If | |
2424 | .B \-\-brief | |
2425 | or | |
2426 | .B \-\-scan | |
2427 | is given, then multiple devices that are components of the one array | |
2428 | are grouped together and reported in a single entry suitable | |
2429 | for inclusion in | |
2430 | .BR mdadm.conf . | |
2431 | ||
2432 | Having | |
2433 | .B \-\-scan | |
2434 | without listing any devices will cause all devices listed in the | |
2435 | config file to be examined. | |
2436 | ||
2437 | .TP | |
2438 | .BI \-\-dump= directory | |
2439 | If the device contains RAID metadata, a file will be created in the | |
2440 | .I directory | |
2441 | and the metadata will be written to it. The file will be the same | |
2442 | size as the device and have the metadata written in the file at the | |
2443 | same locate that it exists in the device. However the file will be "sparse" so | |
2444 | that only those blocks containing metadata will be allocated. The | |
2445 | total space used will be small. | |
2446 | ||
2447 | The file name used in the | |
2448 | .I directory | |
2449 | will be the base name of the device. Further if any links appear in | |
2450 | .I /dev/disk/by-id | |
2451 | which point to the device, then hard links to the file will be created | |
2452 | in | |
2453 | .I directory | |
2454 | based on these | |
2455 | .I by-id | |
2456 | names. | |
2457 | ||
2458 | Multiple devices can be listed and their metadata will all be stored | |
2459 | in the one directory. | |
2460 | ||
2461 | .TP | |
2462 | .BI \-\-restore= directory | |
2463 | This is the reverse of | |
2464 | .BR \-\-dump . | |
2465 | .I mdadm | |
2466 | will locate a file in the directory that has a name appropriate for | |
2467 | the given device and will restore metadata from it. Names that match | |
2468 | .I /dev/disk/by-id | |
2469 | names are preferred, however if two of those refer to different files, | |
2470 | .I mdadm | |
2471 | will not choose between them but will abort the operation. | |
2472 | ||
2473 | If a file name is given instead of a | |
2474 | .I directory | |
2475 | then | |
2476 | .I mdadm | |
2477 | will restore from that file to a single device, always provided the | |
2478 | size of the file matches that of the device, and the file contains | |
2479 | valid metadata. | |
2480 | .TP | |
2481 | .B \-\-stop | |
2482 | The devices should be active md arrays which will be deactivated, as | |
2483 | long as they are not currently in use. | |
2484 | ||
2485 | .TP | |
2486 | .B \-\-run | |
2487 | This will fully activate a partially assembled md array. | |
2488 | ||
2489 | .TP | |
2490 | .B \-\-readonly | |
2491 | This will mark an active array as read-only, providing that it is | |
2492 | not currently being used. | |
2493 | ||
2494 | .TP | |
2495 | .B \-\-readwrite | |
2496 | This will change a | |
2497 | .B readonly | |
2498 | array back to being read/write. | |
2499 | ||
2500 | .TP | |
2501 | .B \-\-scan | |
2502 | For all operations except | |
2503 | .BR \-\-examine , | |
2504 | .B \-\-scan | |
2505 | will cause the operation to be applied to all arrays listed in | |
2506 | .BR /proc/mdstat . | |
2507 | For | |
2508 | .BR \-\-examine, | |
2509 | .B \-\-scan | |
2510 | causes all devices listed in the config file to be examined. | |
2511 | ||
2512 | .TP | |
2513 | .BR \-b ", " \-\-brief | |
2514 | Be less verbose. This is used with | |
2515 | .B \-\-detail | |
2516 | and | |
2517 | .BR \-\-examine . | |
2518 | Using | |
2519 | .B \-\-brief | |
2520 | with | |
2521 | .B \-\-verbose | |
2522 | gives an intermediate level of verbosity. | |
2523 | ||
2524 | .SH MONITOR MODE | |
2525 | ||
2526 | .HP 12 | |
2527 | Usage: | |
2528 | .B mdadm \-\-monitor | |
2529 | .I options... devices... | |
2530 | ||
2531 | .PP | |
2532 | This usage causes | |
2533 | .I mdadm | |
2534 | to periodically poll a number of md arrays and to report on any events | |
2535 | noticed. | |
2536 | .I mdadm | |
2537 | will never exit once it decides that there are arrays to be checked, | |
2538 | so it should normally be run in the background. | |
2539 | ||
2540 | As well as reporting events, | |
2541 | .I mdadm | |
2542 | may move a spare drive from one array to another if they are in the | |
2543 | same | |
2544 | .B spare-group | |
2545 | or | |
2546 | .B domain | |
2547 | and if the destination array has a failed drive but no spares. | |
2548 | ||
2549 | If any devices are listed on the command line, | |
2550 | .I mdadm | |
2551 | will only monitor those devices. Otherwise all arrays listed in the | |
2552 | configuration file will be monitored. Further, if | |
2553 | .B \-\-scan | |
2554 | is given, then any other md devices that appear in | |
2555 | .B /proc/mdstat | |
2556 | will also be monitored. | |
2557 | ||
2558 | The result of monitoring the arrays is the generation of events. | |
2559 | These events are passed to a separate program (if specified) and may | |
2560 | be mailed to a given E-mail address. | |
2561 | ||
2562 | When passing events to a program, the program is run once for each event, | |
2563 | and is given 2 or 3 command-line arguments: the first is the | |
2564 | name of the event (see below), the second is the name of the | |
2565 | md device which is affected, and the third is the name of a related | |
2566 | device if relevant (such as a component device that has failed). | |
2567 | ||
2568 | If | |
2569 | .B \-\-scan | |
2570 | is given, then a program or an E-mail address must be specified on the | |
2571 | command line or in the config file. If neither are available, then | |
2572 | .I mdadm | |
2573 | will not monitor anything. | |
2574 | Without | |
2575 | .B \-\-scan, | |
2576 | .I mdadm | |
2577 | will continue monitoring as long as something was found to monitor. If | |
2578 | no program or email is given, then each event is reported to | |
2579 | .BR stdout . | |
2580 | ||
2581 | The different events are: | |
2582 | ||
2583 | .RS 4 | |
2584 | .TP | |
2585 | .B DeviceDisappeared | |
2586 | An md array which previously was configured appears to no longer be | |
2587 | configured. (syslog priority: Critical) | |
2588 | ||
2589 | If | |
2590 | .I mdadm | |
2591 | was told to monitor an array which is RAID0 or Linear, then it will | |
2592 | report | |
2593 | .B DeviceDisappeared | |
2594 | with the extra information | |
2595 | .BR Wrong-Level . | |
2596 | This is because RAID0 and Linear do not support the device-failed, | |
2597 | hot-spare and resync operations which are monitored. | |
2598 | ||
2599 | .TP | |
2600 | .B RebuildStarted | |
2601 | An md array started reconstruction (e.g. recovery, resync, reshape, | |
2602 | check, repair). (syslog priority: Warning) | |
2603 | ||
2604 | .TP | |
2605 | .BI Rebuild NN | |
2606 | Where | |
2607 | .I NN | |
2608 | is a two-digit number (ie. 05, 48). This indicates that rebuild | |
2609 | has passed that many percent of the total. The events are generated | |
2610 | with fixed increment since 0. Increment size may be specified with | |
2611 | a commandline option (default is 20). (syslog priority: Warning) | |
2612 | ||
2613 | .TP | |
2614 | .B RebuildFinished | |
2615 | An md array that was rebuilding, isn't any more, either because it | |
2616 | finished normally or was aborted. (syslog priority: Warning) | |
2617 | ||
2618 | .TP | |
2619 | .B Fail | |
2620 | An active component device of an array has been marked as | |
2621 | faulty. (syslog priority: Critical) | |
2622 | ||
2623 | .TP | |
2624 | .B FailSpare | |
2625 | A spare component device which was being rebuilt to replace a faulty | |
2626 | device has failed. (syslog priority: Critical) | |
2627 | ||
2628 | .TP | |
2629 | .B SpareActive | |
2630 | A spare component device which was being rebuilt to replace a faulty | |
2631 | device has been successfully rebuilt and has been made active. | |
2632 | (syslog priority: Info) | |
2633 | ||
2634 | .TP | |
2635 | .B NewArray | |
2636 | A new md array has been detected in the | |
2637 | .B /proc/mdstat | |
2638 | file. (syslog priority: Info) | |
2639 | ||
2640 | .TP | |
2641 | .B DegradedArray | |
2642 | A newly noticed array appears to be degraded. This message is not | |
2643 | generated when | |
2644 | .I mdadm | |
2645 | notices a drive failure which causes degradation, but only when | |
2646 | .I mdadm | |
2647 | notices that an array is degraded when it first sees the array. | |
2648 | (syslog priority: Critical) | |
2649 | ||
2650 | .TP | |
2651 | .B MoveSpare | |
2652 | A spare drive has been moved from one array in a | |
2653 | .B spare-group | |
2654 | or | |
2655 | .B domain | |
2656 | to another to allow a failed drive to be replaced. | |
2657 | (syslog priority: Info) | |
2658 | ||
2659 | .TP | |
2660 | .B SparesMissing | |
2661 | If | |
2662 | .I mdadm | |
2663 | has been told, via the config file, that an array should have a certain | |
2664 | number of spare devices, and | |
2665 | .I mdadm | |
2666 | detects that it has fewer than this number when it first sees the | |
2667 | array, it will report a | |
2668 | .B SparesMissing | |
2669 | message. | |
2670 | (syslog priority: Warning) | |
2671 | ||
2672 | .TP | |
2673 | .B TestMessage | |
2674 | An array was found at startup, and the | |
2675 | .B \-\-test | |
2676 | flag was given. | |
2677 | (syslog priority: Info) | |
2678 | .RE | |
2679 | ||
2680 | Only | |
2681 | .B Fail, | |
2682 | .B FailSpare, | |
2683 | .B DegradedArray, | |
2684 | .B SparesMissing | |
2685 | and | |
2686 | .B TestMessage | |
2687 | cause Email to be sent. All events cause the program to be run. | |
2688 | The program is run with two or three arguments: the event | |
2689 | name, the array device and possibly a second device. | |
2690 | ||
2691 | Each event has an associated array device (e.g. | |
2692 | .BR /dev/md1 ) | |
2693 | and possibly a second device. For | |
2694 | .BR Fail , | |
2695 | .BR FailSpare , | |
2696 | and | |
2697 | .B SpareActive | |
2698 | the second device is the relevant component device. | |
2699 | For | |
2700 | .B MoveSpare | |
2701 | the second device is the array that the spare was moved from. | |
2702 | ||
2703 | For | |
2704 | .I mdadm | |
2705 | to move spares from one array to another, the different arrays need to | |
2706 | be labeled with the same | |
2707 | .B spare-group | |
2708 | or the spares must be allowed to migrate through matching POLICY domains | |
2709 | in the configuration file. The | |
2710 | .B spare-group | |
2711 | name can be any string; it is only necessary that different spare | |
2712 | groups use different names. | |
2713 | ||
2714 | When | |
2715 | .I mdadm | |
2716 | detects that an array in a spare group has fewer active | |
2717 | devices than necessary for the complete array, and has no spare | |
2718 | devices, it will look for another array in the same spare group that | |
2719 | has a full complement of working drive and a spare. It will then | |
2720 | attempt to remove the spare from the second drive and add it to the | |
2721 | first. | |
2722 | If the removal succeeds but the adding fails, then it is added back to | |
2723 | the original array. | |
2724 | ||
2725 | If the spare group for a degraded array is not defined, | |
2726 | .I mdadm | |
2727 | will look at the rules of spare migration specified by POLICY lines in | |
2728 | .B mdadm.conf | |
2729 | and then follow similar steps as above if a matching spare is found. | |
2730 | ||
2731 | .SH GROW MODE | |
2732 | The GROW mode is used for changing the size or shape of an active | |
2733 | array. | |
2734 | For this to work, the kernel must support the necessary change. | |
2735 | Various types of growth are being added during 2.6 development. | |
2736 | ||
2737 | Currently the supported changes include | |
2738 | .IP \(bu 4 | |
2739 | change the "size" attribute for RAID1, RAID4, RAID5 and RAID6. | |
2740 | .IP \(bu 4 | |
2741 | increase or decrease the "raid\-devices" attribute of RAID0, RAID1, RAID4, | |
2742 | RAID5, and RAID6. | |
2743 | .IP \(bu 4 | |
2744 | change the chunk-size and layout of RAID0, RAID4, RAID5, RAID6 and RAID10. | |
2745 | .IP \(bu 4 | |
2746 | convert between RAID1 and RAID5, between RAID5 and RAID6, between | |
2747 | RAID0, RAID4, and RAID5, and between RAID0 and RAID10 (in the near-2 mode). | |
2748 | .IP \(bu 4 | |
2749 | add a write-intent bitmap to any array which supports these bitmaps, or | |
2750 | remove a write-intent bitmap from such an array. | |
2751 | .IP \(bu 4 | |
2752 | change the array's consistency policy. | |
2753 | .PP | |
2754 | ||
2755 | Using GROW on containers is currently supported only for Intel's IMSM | |
2756 | container format. The number of devices in a container can be | |
2757 | increased - which affects all arrays in the container - or an array | |
2758 | in a container can be converted between levels where those levels are | |
2759 | supported by the container, and the conversion is on of those listed | |
2760 | above. | |
2761 | ||
2762 | .PP | |
2763 | Notes: | |
2764 | .IP \(bu 4 | |
2765 | Intel's native checkpointing doesn't use | |
2766 | .B --backup-file | |
2767 | option and it is transparent for assembly feature. | |
2768 | .IP \(bu 4 | |
2769 | Roaming between Windows(R) and Linux systems for IMSM metadata is not | |
2770 | supported during grow process. | |
2771 | .IP \(bu 4 | |
2772 | When growing a raid0 device, the new component disk size (or external | |
2773 | backup size) should be larger than LCM(old, new) * chunk-size * 2, | |
2774 | where LCM() is the least common multiple of the old and new count of | |
2775 | component disks, and "* 2" comes from the fact that mdadm refuses to | |
2776 | use more than half of a spare device for backup space. | |
2777 | ||
2778 | .SS SIZE CHANGES | |
2779 | Normally when an array is built the "size" is taken from the smallest | |
2780 | of the drives. If all the small drives in an arrays are, one at a | |
2781 | time, removed and replaced with larger drives, then you could have an | |
2782 | array of large drives with only a small amount used. In this | |
2783 | situation, changing the "size" with "GROW" mode will allow the extra | |
2784 | space to start being used. If the size is increased in this way, a | |
2785 | "resync" process will start to make sure the new parts of the array | |
2786 | are synchronised. | |
2787 | ||
2788 | Note that when an array changes size, any filesystem that may be | |
2789 | stored in the array will not automatically grow or shrink to use or | |
2790 | vacate the space. The | |
2791 | filesystem will need to be explicitly told to use the extra space | |
2792 | after growing, or to reduce its size | |
2793 | .B prior | |
2794 | to shrinking the array. | |
2795 | ||
2796 | Also the size of an array cannot be changed while it has an active | |
2797 | bitmap. If an array has a bitmap, it must be removed before the size | |
2798 | can be changed. Once the change is complete a new bitmap can be created. | |
2799 | ||
2800 | .PP | |
2801 | Note: | |
2802 | .B "--grow --size" | |
2803 | is not yet supported for external file bitmap. | |
2804 | ||
2805 | .SS RAID\-DEVICES CHANGES | |
2806 | ||
2807 | A RAID1 array can work with any number of devices from 1 upwards | |
2808 | (though 1 is not very useful). There may be times which you want to | |
2809 | increase or decrease the number of active devices. Note that this is | |
2810 | different to hot-add or hot-remove which changes the number of | |
2811 | inactive devices. | |
2812 | ||
2813 | When reducing the number of devices in a RAID1 array, the slots which | |
2814 | are to be removed from the array must already be vacant. That is, the | |
2815 | devices which were in those slots must be failed and removed. | |
2816 | ||
2817 | When the number of devices is increased, any hot spares that are | |
2818 | present will be activated immediately. | |
2819 | ||
2820 | Changing the number of active devices in a RAID5 or RAID6 is much more | |
2821 | effort. Every block in the array will need to be read and written | |
2822 | back to a new location. From 2.6.17, the Linux Kernel is able to | |
2823 | increase the number of devices in a RAID5 safely, including restarting | |
2824 | an interrupted "reshape". From 2.6.31, the Linux Kernel is able to | |
2825 | increase or decrease the number of devices in a RAID5 or RAID6. | |
2826 | ||
2827 | From 2.6.35, the Linux Kernel is able to convert a RAID0 in to a RAID4 | |
2828 | or RAID5. | |
2829 | .I mdadm | |
2830 | uses this functionality and the ability to add | |
2831 | devices to a RAID4 to allow devices to be added to a RAID0. When | |
2832 | requested to do this, | |
2833 | .I mdadm | |
2834 | will convert the RAID0 to a RAID4, add the necessary disks and make | |
2835 | the reshape happen, and then convert the RAID4 back to RAID0. | |
2836 | ||
2837 | When decreasing the number of devices, the size of the array will also | |
2838 | decrease. If there was data in the array, it could get destroyed and | |
2839 | this is not reversible, so you should firstly shrink the filesystem on | |
2840 | the array to fit within the new size. To help prevent accidents, | |
2841 | .I mdadm | |
2842 | requires that the size of the array be decreased first with | |
2843 | .BR "mdadm --grow --array-size" . | |
2844 | This is a reversible change which simply makes the end of the array | |
2845 | inaccessible. The integrity of any data can then be checked before | |
2846 | the non-reversible reduction in the number of devices is request. | |
2847 | ||
2848 | When relocating the first few stripes on a RAID5 or RAID6, it is not | |
2849 | possible to keep the data on disk completely consistent and | |
2850 | crash-proof. To provide the required safety, mdadm disables writes to | |
2851 | the array while this "critical section" is reshaped, and takes a | |
2852 | backup of the data that is in that section. For grows, this backup may be | |
2853 | stored in any spare devices that the array has, however it can also be | |
2854 | stored in a separate file specified with the | |
2855 | .B \-\-backup\-file | |
2856 | option, and is required to be specified for shrinks, RAID level | |
2857 | changes and layout changes. If this option is used, and the system | |
2858 | does crash during the critical period, the same file must be passed to | |
2859 | .B \-\-assemble | |
2860 | to restore the backup and reassemble the array. When shrinking rather | |
2861 | than growing the array, the reshape is done from the end towards the | |
2862 | beginning, so the "critical section" is at the end of the reshape. | |
2863 | ||
2864 | .SS LEVEL CHANGES | |
2865 | ||
2866 | Changing the RAID level of any array happens instantaneously. However | |
2867 | in the RAID5 to RAID6 case this requires a non-standard layout of the | |
2868 | RAID6 data, and in the RAID6 to RAID5 case that non-standard layout is | |
2869 | required before the change can be accomplished. So while the level | |
2870 | change is instant, the accompanying layout change can take quite a | |
2871 | long time. A | |
2872 | .B \-\-backup\-file | |
2873 | is required. If the array is not simultaneously being grown or | |
2874 | shrunk, so that the array size will remain the same - for example, | |
2875 | reshaping a 3-drive RAID5 into a 4-drive RAID6 - the backup file will | |
2876 | be used not just for a "cricital section" but throughout the reshape | |
2877 | operation, as described below under LAYOUT CHANGES. | |
2878 | ||
2879 | .SS CHUNK-SIZE AND LAYOUT CHANGES | |
2880 | ||
2881 | Changing the chunk-size or layout without also changing the number of | |
2882 | devices as the same time will involve re-writing all blocks in-place. | |
2883 | To ensure against data loss in the case of a crash, a | |
2884 | .B --backup-file | |
2885 | must be provided for these changes. Small sections of the array will | |
2886 | be copied to the backup file while they are being rearranged. This | |
2887 | means that all the data is copied twice, once to the backup and once | |
2888 | to the new layout on the array, so this type of reshape will go very | |
2889 | slowly. | |
2890 | ||
2891 | If the reshape is interrupted for any reason, this backup file must be | |
2892 | made available to | |
2893 | .B "mdadm --assemble" | |
2894 | so the array can be reassembled. Consequently the file cannot be | |
2895 | stored on the device being reshaped. | |
2896 | ||
2897 | ||
2898 | .SS BITMAP CHANGES | |
2899 | ||
2900 | A write-intent bitmap can be added to, or removed from, an active | |
2901 | array. Either internal bitmaps, or bitmaps stored in a separate file, | |
2902 | can be added. Note that if you add a bitmap stored in a file which is | |
2903 | in a filesystem that is on the RAID array being affected, the system | |
2904 | will deadlock. The bitmap must be on a separate filesystem. | |
2905 | ||
2906 | .SS CONSISTENCY POLICY CHANGES | |
2907 | ||
2908 | The consistency policy of an active array can be changed by using the | |
2909 | .B \-\-consistency\-policy | |
2910 | option in Grow mode. Currently this works only for the | |
2911 | .B ppl | |
2912 | and | |
2913 | .B resync | |
2914 | policies and allows to enable or disable the RAID5 Partial Parity Log (PPL). | |
2915 | ||
2916 | .SH INCREMENTAL MODE | |
2917 | ||
2918 | .HP 12 | |
2919 | Usage: | |
2920 | .B mdadm \-\-incremental | |
2921 | .RB [ \-\-run ] | |
2922 | .RB [ \-\-quiet ] | |
2923 | .I component-device | |
2924 | .RI [ optional-aliases-for-device ] | |
2925 | .HP 12 | |
2926 | Usage: | |
2927 | .B mdadm \-\-incremental \-\-fail | |
2928 | .I component-device | |
2929 | .HP 12 | |
2930 | Usage: | |
2931 | .B mdadm \-\-incremental \-\-rebuild\-map | |
2932 | .HP 12 | |
2933 | Usage: | |
2934 | .B mdadm \-\-incremental \-\-run \-\-scan | |
2935 | ||
2936 | .PP | |
2937 | This mode is designed to be used in conjunction with a device | |
2938 | discovery system. As devices are found in a system, they can be | |
2939 | passed to | |
2940 | .B "mdadm \-\-incremental" | |
2941 | to be conditionally added to an appropriate array. | |
2942 | ||
2943 | Conversely, it can also be used with the | |
2944 | .B \-\-fail | |
2945 | flag to do just the opposite and find whatever array a particular device | |
2946 | is part of and remove the device from that array. | |
2947 | ||
2948 | If the device passed is a | |
2949 | .B CONTAINER | |
2950 | device created by a previous call to | |
2951 | .IR mdadm , | |
2952 | then rather than trying to add that device to an array, all the arrays | |
2953 | described by the metadata of the container will be started. | |
2954 | ||
2955 | .I mdadm | |
2956 | performs a number of tests to determine if the device is part of an | |
2957 | array, and which array it should be part of. If an appropriate array | |
2958 | is found, or can be created, | |
2959 | .I mdadm | |
2960 | adds the device to the array and conditionally starts the array. | |
2961 | ||
2962 | Note that | |
2963 | .I mdadm | |
2964 | will normally only add devices to an array which were previously working | |
2965 | (active or spare) parts of that array. The support for automatic | |
2966 | inclusion of a new drive as a spare in some array requires | |
2967 | a configuration through POLICY in config file. | |
2968 | ||
2969 | The tests that | |
2970 | .I mdadm | |
2971 | makes are as follow: | |
2972 | .IP + | |
2973 | Is the device permitted by | |
2974 | .BR mdadm.conf ? | |
2975 | That is, is it listed in a | |
2976 | .B DEVICES | |
2977 | line in that file. If | |
2978 | .B DEVICES | |
2979 | is absent then the default it to allow any device. Similarly if | |
2980 | .B DEVICES | |
2981 | contains the special word | |
2982 | .B partitions | |
2983 | then any device is allowed. Otherwise the device name given to | |
2984 | .IR mdadm , | |
2985 | or one of the aliases given, or an alias found in the filesystem, | |
2986 | must match one of the names or patterns in a | |
2987 | .B DEVICES | |
2988 | line. | |
2989 | ||
2990 | This is the only context where the aliases are used. They are | |
2991 | usually provided by a | |
2992 | .I udev | |
2993 | rules mentioning | |
2994 | .BR $env{DEVLINKS} . | |
2995 | ||
2996 | .IP + | |
2997 | Does the device have a valid md superblock? If a specific metadata | |
2998 | version is requested with | |
2999 | .B \-\-metadata | |
3000 | or | |
3001 | .B \-e | |
3002 | then only that style of metadata is accepted, otherwise | |
3003 | .I mdadm | |
3004 | finds any known version of metadata. If no | |
3005 | .I md | |
3006 | metadata is found, the device may be still added to an array | |
3007 | as a spare if POLICY allows. | |
3008 | ||
3009 | .ig | |
3010 | .IP + | |
3011 | Does the metadata match an expected array? | |
3012 | The metadata can match in two ways. Either there is an array listed | |
3013 | in | |
3014 | .B mdadm.conf | |
3015 | which identifies the array (either by UUID, by name, by device list, | |
3016 | or by minor-number), or the array was created with a | |
3017 | .B homehost | |
3018 | specified and that | |
3019 | .B homehost | |
3020 | matches the one in | |
3021 | .B mdadm.conf | |
3022 | or on the command line. | |
3023 | If | |
3024 | .I mdadm | |
3025 | is not able to positively identify the array as belonging to the | |
3026 | current host, the device will be rejected. | |
3027 | .. | |
3028 | ||
3029 | .PP | |
3030 | .I mdadm | |
3031 | keeps a list of arrays that it has partially assembled in | |
3032 | .BR {MAP_PATH} . | |
3033 | If no array exists which matches | |
3034 | the metadata on the new device, | |
3035 | .I mdadm | |
3036 | must choose a device name and unit number. It does this based on any | |
3037 | name given in | |
3038 | .B mdadm.conf | |
3039 | or any name information stored in the metadata. If this name | |
3040 | suggests a unit number, that number will be used, otherwise a free | |
3041 | unit number will be chosen. Normally | |
3042 | .I mdadm | |
3043 | will prefer to create a partitionable array, however if the | |
3044 | .B CREATE | |
3045 | line in | |
3046 | .B mdadm.conf | |
3047 | suggests that a non-partitionable array is preferred, that will be | |
3048 | honoured. | |
3049 | ||
3050 | If the array is not found in the config file and its metadata does not | |
3051 | identify it as belonging to the "homehost", then | |
3052 | .I mdadm | |
3053 | will choose a name for the array which is certain not to conflict with | |
3054 | any array which does belong to this host. It does this be adding an | |
3055 | underscore and a small number to the name preferred by the metadata. | |
3056 | ||
3057 | Once an appropriate array is found or created and the device is added, | |
3058 | .I mdadm | |
3059 | must decide if the array is ready to be started. It will | |
3060 | normally compare the number of available (non-spare) devices to the | |
3061 | number of devices that the metadata suggests need to be active. If | |
3062 | there are at least that many, the array will be started. This means | |
3063 | that if any devices are missing the array will not be restarted. | |
3064 | ||
3065 | As an alternative, | |
3066 | .B \-\-run | |
3067 | may be passed to | |
3068 | .I mdadm | |
3069 | in which case the array will be run as soon as there are enough | |
3070 | devices present for the data to be accessible. For a RAID1, that | |
3071 | means one device will start the array. For a clean RAID5, the array | |
3072 | will be started as soon as all but one drive is present. | |
3073 | ||
3074 | Note that neither of these approaches is really ideal. If it can | |
3075 | be known that all device discovery has completed, then | |
3076 | .br | |
3077 | .B " mdadm \-IRs" | |
3078 | .br | |
3079 | can be run which will try to start all arrays that are being | |
3080 | incrementally assembled. They are started in "read-auto" mode in | |
3081 | which they are read-only until the first write request. This means | |
3082 | that no metadata updates are made and no attempt at resync or recovery | |
3083 | happens. Further devices that are found before the first write can | |
3084 | still be added safely. | |
3085 | ||
3086 | .SH ENVIRONMENT | |
3087 | This section describes environment variables that affect how mdadm | |
3088 | operates. | |
3089 | ||
3090 | .TP | |
3091 | .B MDADM_NO_MDMON | |
3092 | Setting this value to 1 will prevent mdadm from automatically launching | |
3093 | mdmon. This variable is intended primarily for debugging mdadm/mdmon. | |
3094 | ||
3095 | .TP | |
3096 | .B MDADM_NO_UDEV | |
3097 | Normally, | |
3098 | .I mdadm | |
3099 | does not create any device nodes in /dev, but leaves that task to | |
3100 | .IR udev . | |
3101 | If | |
3102 | .I udev | |
3103 | appears not to be configured, or if this environment variable is set | |
3104 | to '1', the | |
3105 | .I mdadm | |
3106 | will create and devices that are needed. | |
3107 | ||
3108 | .TP | |
3109 | .B MDADM_NO_SYSTEMCTL | |
3110 | If | |
3111 | .I mdadm | |
3112 | detects that | |
3113 | .I systemd | |
3114 | is in use it will normally request | |
3115 | .I systemd | |
3116 | to start various background tasks (particularly | |
3117 | .IR mdmon ) | |
3118 | rather than forking and running them in the background. This can be | |
3119 | suppressed by setting | |
3120 | .BR MDADM_NO_SYSTEMCTL=1 . | |
3121 | ||
3122 | .TP | |
3123 | .B IMSM_NO_PLATFORM | |
3124 | A key value of IMSM metadata is that it allows interoperability with | |
3125 | boot ROMs on Intel platforms, and with other major operating systems. | |
3126 | Consequently, | |
3127 | .I mdadm | |
3128 | will only allow an IMSM array to be created or modified if detects | |
3129 | that it is running on an Intel platform which supports IMSM, and | |
3130 | supports the particular configuration of IMSM that is being requested | |
3131 | (some functionality requires newer OROM support). | |
3132 | ||
3133 | These checks can be suppressed by setting IMSM_NO_PLATFORM=1 in the | |
3134 | environment. This can be useful for testing or for disaster | |
3135 | recovery. You should be aware that interoperability may be | |
3136 | compromised by setting this value. | |
3137 | ||
3138 | .TP | |
3139 | .B MDADM_GROW_ALLOW_OLD | |
3140 | If an array is stopped while it is performing a reshape and that | |
3141 | reshape was making use of a backup file, then when the array is | |
3142 | re-assembled | |
3143 | .I mdadm | |
3144 | will sometimes complain that the backup file is too old. If this | |
3145 | happens and you are certain it is the right backup file, you can | |
3146 | over-ride this check by setting | |
3147 | .B MDADM_GROW_ALLOW_OLD=1 | |
3148 | in the environment. | |
3149 | ||
3150 | .TP | |
3151 | .B MDADM_CONF_AUTO | |
3152 | Any string given in this variable is added to the start of the | |
3153 | .B AUTO | |
3154 | line in the config file, or treated as the whole | |
3155 | .B AUTO | |
3156 | line if none is given. It can be used to disable certain metadata | |
3157 | types when | |
3158 | .I mdadm | |
3159 | is called from a boot script. For example | |
3160 | .br | |
3161 | .B " export MDADM_CONF_AUTO='-ddf -imsm' | |
3162 | .br | |
3163 | will make sure that | |
3164 | .I mdadm | |
3165 | does not automatically assemble any DDF or | |
3166 | IMSM arrays that are found. This can be useful on systems configured | |
3167 | to manage such arrays with | |
3168 | .BR dmraid . | |
3169 | ||
3170 | ||
3171 | .SH EXAMPLES | |
3172 | ||
3173 | .B " mdadm \-\-query /dev/name-of-device" | |
3174 | .br | |
3175 | This will find out if a given device is a RAID array, or is part of | |
3176 | one, and will provide brief information about the device. | |
3177 | ||
3178 | .B " mdadm \-\-assemble \-\-scan" | |
3179 | .br | |
3180 | This will assemble and start all arrays listed in the standard config | |
3181 | file. This command will typically go in a system startup file. | |
3182 | ||
3183 | .B " mdadm \-\-stop \-\-scan" | |
3184 | .br | |
3185 | This will shut down all arrays that can be shut down (i.e. are not | |
3186 | currently in use). This will typically go in a system shutdown script. | |
3187 | ||
3188 | .B " mdadm \-\-follow \-\-scan \-\-delay=120" | |
3189 | .br | |
3190 | If (and only if) there is an Email address or program given in the | |
3191 | standard config file, then | |
3192 | monitor the status of all arrays listed in that file by | |
3193 | polling them ever 2 minutes. | |
3194 | ||
3195 | .B " mdadm \-\-create /dev/md0 \-\-level=1 \-\-raid\-devices=2 /dev/hd[ac]1" | |
3196 | .br | |
3197 | Create /dev/md0 as a RAID1 array consisting of /dev/hda1 and /dev/hdc1. | |
3198 | ||
3199 | .br | |
3200 | .B " echo 'DEVICE /dev/hd*[0\-9] /dev/sd*[0\-9]' > mdadm.conf" | |
3201 | .br | |
3202 | .B " mdadm \-\-detail \-\-scan >> mdadm.conf" | |
3203 | .br | |
3204 | This will create a prototype config file that describes currently | |
3205 | active arrays that are known to be made from partitions of IDE or SCSI drives. | |
3206 | This file should be reviewed before being used as it may | |
3207 | contain unwanted detail. | |
3208 | ||
3209 | .B " echo 'DEVICE /dev/hd[a\-z] /dev/sd*[a\-z]' > mdadm.conf" | |
3210 | .br | |
3211 | .B " mdadm \-\-examine \-\-scan \-\-config=mdadm.conf >> mdadm.conf" | |
3212 | .br | |
3213 | This will find arrays which could be assembled from existing IDE and | |
3214 | SCSI whole drives (not partitions), and store the information in the | |
3215 | format of a config file. | |
3216 | This file is very likely to contain unwanted detail, particularly | |
3217 | the | |
3218 | .B devices= | |
3219 | entries. It should be reviewed and edited before being used as an | |
3220 | actual config file. | |
3221 | ||
3222 | .B " mdadm \-\-examine \-\-brief \-\-scan \-\-config=partitions" | |
3223 | .br | |
3224 | .B " mdadm \-Ebsc partitions" | |
3225 | .br | |
3226 | Create a list of devices by reading | |
3227 | .BR /proc/partitions , | |
3228 | scan these for RAID superblocks, and printout a brief listing of all | |
3229 | that were found. | |
3230 | ||
3231 | .B " mdadm \-Ac partitions \-m 0 /dev/md0" | |
3232 | .br | |
3233 | Scan all partitions and devices listed in | |
3234 | .BR /proc/partitions | |
3235 | and assemble | |
3236 | .B /dev/md0 | |
3237 | out of all such devices with a RAID superblock with a minor number of 0. | |
3238 | ||
3239 | .B " mdadm \-\-monitor \-\-scan \-\-daemonise > /run/mdadm/mon.pid" | |
3240 | .br | |
3241 | If config file contains a mail address or alert program, run mdadm in | |
3242 | the background in monitor mode monitoring all md devices. Also write | |
3243 | pid of mdadm daemon to | |
3244 | .BR /run/mdadm/mon.pid . | |
3245 | ||
3246 | .B " mdadm \-Iq /dev/somedevice" | |
3247 | .br | |
3248 | Try to incorporate newly discovered device into some array as | |
3249 | appropriate. | |
3250 | ||
3251 | .B " mdadm \-\-incremental \-\-rebuild\-map \-\-run \-\-scan" | |
3252 | .br | |
3253 | Rebuild the array map from any current arrays, and then start any that | |
3254 | can be started. | |
3255 | ||
3256 | .B " mdadm /dev/md4 --fail detached --remove detached" | |
3257 | .br | |
3258 | Any devices which are components of /dev/md4 will be marked as faulty | |
3259 | and then remove from the array. | |
3260 | ||
3261 | .B " mdadm --grow /dev/md4 --level=6 --backup-file=/root/backup-md4" | |
3262 | .br | |
3263 | The array | |
3264 | .B /dev/md4 | |
3265 | which is currently a RAID5 array will be converted to RAID6. There | |
3266 | should normally already be a spare drive attached to the array as a | |
3267 | RAID6 needs one more drive than a matching RAID5. | |
3268 | ||
3269 | .B " mdadm --create /dev/md/ddf --metadata=ddf --raid-disks 6 /dev/sd[a-f]" | |
3270 | .br | |
3271 | Create a DDF array over 6 devices. | |
3272 | ||
3273 | .B " mdadm --create /dev/md/home -n3 -l5 -z 30000000 /dev/md/ddf" | |
3274 | .br | |
3275 | Create a RAID5 array over any 3 devices in the given DDF set. Use | |
3276 | only 30 gigabytes of each device. | |
3277 | ||
3278 | .B " mdadm -A /dev/md/ddf1 /dev/sd[a-f]" | |
3279 | .br | |
3280 | Assemble a pre-exist ddf array. | |
3281 | ||
3282 | .B " mdadm -I /dev/md/ddf1" | |
3283 | .br | |
3284 | Assemble all arrays contained in the ddf array, assigning names as | |
3285 | appropriate. | |
3286 | ||
3287 | .B " mdadm \-\-create \-\-help" | |
3288 | .br | |
3289 | Provide help about the Create mode. | |
3290 | ||
3291 | .B " mdadm \-\-config \-\-help" | |
3292 | .br | |
3293 | Provide help about the format of the config file. | |
3294 | ||
3295 | .B " mdadm \-\-help" | |
3296 | .br | |
3297 | Provide general help. | |
3298 | ||
3299 | .SH FILES | |
3300 | ||
3301 | .SS /proc/mdstat | |
3302 | ||
3303 | If you're using the | |
3304 | .B /proc | |
3305 | filesystem, | |
3306 | .B /proc/mdstat | |
3307 | lists all active md devices with information about them. | |
3308 | .I mdadm | |
3309 | uses this to find arrays when | |
3310 | .B \-\-scan | |
3311 | is given in Misc mode, and to monitor array reconstruction | |
3312 | on Monitor mode. | |
3313 | ||
3314 | .SS /etc/mdadm.conf | |
3315 | ||
3316 | The config file lists which devices may be scanned to see if | |
3317 | they contain MD super block, and gives identifying information | |
3318 | (e.g. UUID) about known MD arrays. See | |
3319 | .BR mdadm.conf (5) | |
3320 | for more details. | |
3321 | ||
3322 | .SS /etc/mdadm.conf.d | |
3323 | ||
3324 | A directory containing configuration files which are read in lexical | |
3325 | order. | |
3326 | ||
3327 | .SS {MAP_PATH} | |
3328 | When | |
3329 | .B \-\-incremental | |
3330 | mode is used, this file gets a list of arrays currently being created. | |
3331 | ||
3332 | .SH DEVICE NAMES | |
3333 | ||
3334 | .I mdadm | |
3335 | understand two sorts of names for array devices. | |
3336 | ||
3337 | The first is the so-called 'standard' format name, which matches the | |
3338 | names used by the kernel and which appear in | |
3339 | .IR /proc/mdstat . | |
3340 | ||
3341 | The second sort can be freely chosen, but must reside in | |
3342 | .IR /dev/md/ . | |
3343 | When giving a device name to | |
3344 | .I mdadm | |
3345 | to create or assemble an array, either full path name such as | |
3346 | .I /dev/md0 | |
3347 | or | |
3348 | .I /dev/md/home | |
3349 | can be given, or just the suffix of the second sort of name, such as | |
3350 | .I home | |
3351 | can be given. | |
3352 | ||
3353 | When | |
3354 | .I mdadm | |
3355 | chooses device names during auto-assembly or incremental assembly, it | |
3356 | will sometimes add a small sequence number to the end of the name to | |
3357 | avoid conflicted between multiple arrays that have the same name. If | |
3358 | .I mdadm | |
3359 | can reasonably determine that the array really is meant for this host, | |
3360 | either by a hostname in the metadata, or by the presence of the array | |
3361 | in | |
3362 | .BR mdadm.conf , | |
3363 | then it will leave off the suffix if possible. | |
3364 | Also if the homehost is specified as | |
3365 | .B <ignore> | |
3366 | .I mdadm | |
3367 | will only use a suffix if a different array of the same name already | |
3368 | exists or is listed in the config file. | |
3369 | ||
3370 | The standard names for non-partitioned arrays (the only sort of md | |
3371 | array available in 2.4 and earlier) are of the form | |
3372 | .IP | |
3373 | .RB /dev/md NN | |
3374 | .PP | |
3375 | where NN is a number. | |
3376 | The standard names for partitionable arrays (as available from 2.6 | |
3377 | onwards) are of the form: | |
3378 | .IP | |
3379 | .RB /dev/md_d NN | |
3380 | .PP | |
3381 | Partition numbers should be indicated by adding "pMM" to these, thus "/dev/md/d1p2". | |
3382 | .PP | |
3383 | From kernel version 2.6.28 the "non-partitioned array" can actually | |
3384 | be partitioned. So the "md_d\fBNN\fP" | |
3385 | names are no longer needed, and | |
3386 | partitions such as "/dev/md\fBNN\fPp\fBXX\fP" | |
3387 | are possible. | |
3388 | .PP | |
3389 | From kernel version 2.6.29 standard names can be non-numeric following | |
3390 | the form: | |
3391 | .IP | |
3392 | .RB /dev/md_ XXX | |
3393 | .PP | |
3394 | where | |
3395 | .B XXX | |
3396 | is any string. These names are supported by | |
3397 | .I mdadm | |
3398 | since version 3.3 provided they are enabled in | |
3399 | .IR mdadm.conf . | |
3400 | ||
3401 | .SH NOTE | |
3402 | .I mdadm | |
3403 | was previously known as | |
3404 | .IR mdctl . | |
3405 | ||
3406 | .SH SEE ALSO | |
3407 | For further information on mdadm usage, MD and the various levels of | |
3408 | RAID, see: | |
3409 | .IP | |
3410 | .B https://raid.wiki.kernel.org/ | |
3411 | .PP | |
3412 | (based upon Jakob \(/Ostergaard's Software\-RAID.HOWTO) | |
3413 | .PP | |
3414 | The latest version of | |
3415 | .I mdadm | |
3416 | should always be available from | |
3417 | .IP | |
3418 | .B https://www.kernel.org/pub/linux/utils/raid/mdadm/ | |
3419 | .PP | |
3420 | Related man pages: | |
3421 | .PP | |
3422 | .IR mdmon (8), | |
3423 | .IR mdadm.conf (5), | |
3424 | .IR md (4). |