| blob | eabcdffca53669c6d79b65f4f196fa3fb38e6866 |
1 Changes Prior to 2.6 release
2 - Fixed UUID printing in "--detail --brief" for version1 metadata.
3 - --update=resync did exactly the wrong thing for version1 metadata.
4 It caused a resync to not happen, rather than to happen.
5 - Allow --assemble --force to mark a raid6 clean when it has two
6 missing devices (which is needed else if won't assemble.
7 Without this fix it would only assemble if one or zero
8 missing devices.
9 - Support --update=devicesize for cases where the underlying device
10 can change size.
11 - Default to --auto=yes so the array devices with 'standard' names
12 get created automatically, as this is almost always what is wanted.
13 - Give useful message if raid4/5/6 cannot be started because it is
14 not clean and is also degraded.
15 - Increase raid456 stripe cache size if needed to --grow the array.
16 The setting used unfortunately requires intimate knowledge of the
17 kernel, and it not reset when the reshape finishes.
18 - Change 'Device Size' to 'Used Dev Size' because it only shows how
19 much of each device is actually used, not how big they are.
20 - --wait or -W will wait for resync activity to finish on the given
21 devices.
22 - Fix some problems with --update=uuid and add a test.
23 - If two drives in a raid5 disappear at the same time, then "-Af"
24 will add them both in rather than just one and forcing the array
25 to 'clean'. This is slightly safer in some cases.
26 - Check device is large enough before hot-add: this improves quality
27 of error message.
28 - Don't hold md device open for so long in --monitor mode - map_dev
29 can be slow and interferes with trying to stop the array.
30 - Support --uuid= with --create to choose your own UUID.
31 - New major more "--incremental" for incremental assemble of arrays,
32 intended for use with udev.
34 Changes Prior to 2.5.6 release
35 - Fix bug which meant "bitmap=xxx" in mdadm.conf was not handled
36 properly.
37 - Documentation updates.
38 - Fix bug that caused infinite loop when doing auto-assembly,
39 in certain cases where arrays couldn't be assembled.
41 Changes Prior to 2.5.5 release
42 - Don't #include linux/blkpg.h as that isn't safe. Just
43 include the content literally.
44 - Reduce maximum bitmap usage when working with bitmap files,
45 so that a only single-page allocations are made, even on
46 64bit hosts with 4K pages.
47 - Close stray fd in mdassemble so that it can assemble stacked
48 devices
49 - If mdassemble finds an array already assembled, it marks it
50 read-write.
51 - Remove error in md_open if array is already active. This isn't
52 needed and gets in the ways if an array was created e.g. in
53 initramfs, but device doesn't yet exist in /dev.
54 - When --assemble --scan is run, if all arrays that could be found
55 have already been started, don't report an error.
56 - Fix a couple of bugs related to raid10 and the new 'offset' layout.
57 - Improve error message when a wrong '--update' option is given.
59 Changes Prior to 2.5.4 release
60 - When creating devices in /dev/md/ create matching symlinks
61 from /dev. e.g. /dev/md0 -> /dev/md/0.
62 Allow this to be disabled in mdadm.conf or on command line.
63 - Fix some endian-ness issues with version-1 superblocks (affects
64 bigendian only).
65 - Fix endian problem with 'bitmap' metadata
66 - Allow a number (of partitions) after the 'yes' option to --auto=
67 This is particularly useful in the 'create' line in mdadm.conf.
68 - Remove partitions from any whole device that is made part of
69 an md array. This is a work-around for annoying messages
70 when the first block on some drive accidentally looks like a
71 partition table.
73 Changes Prior to 2.5.3 release
74 - Document v0.91 superblocks in md.4
75 - Make GPL explicit in man pages.
76 - Fix recent breakage of starting degraded arrays.
77 - Tidyup automatic name choice for v-1 arrays:
78 /dev/md_d0 now becomes '0', not '_d0'.
80 Changes Prior to 2.5.2 release
81 - Fix problem with compiling with gcc-2 compilers
82 - Fix compile problem of post-incrmenting a variable in a macro arg.
83 - Stop map_dev from returning [0:0], as that breaks things.
84 - Add 'Array Slot' line to --examine for version-1 superblocks
85 to make it a bit easier to see what is happening.
86 - Work around bug in --add handling for version-1 superblocks
87 in 2.6.17 (and prior).
88 - Make -assemble a bit more resilient to finding strange
89 information in superblocks.
90 - Don't claim newly added spares are InSync!! (don't know why that
91 code was ever in there)
92 - Work better when no 'ftw' is available, and check to see
93 if current uclibc provides ftw.
94 - Never use /etc/mdadm.conf if --config file is given (previously
95 some code used one, some used the other).
97 Changes Prior to 2.5.1 release
98 - Various fixes for gcc warnings
99 - uclibc warnings
100 - Makefile improvements for static linking/intalling
101 - Makefile uninstall target
102 - Really fix return status of --examine
103 - Typos
104 - Byteorder stuff (again)
105 - Don't try to create devices with --manage or --grow
106 - allow default metadata (superblock) type to be specified
107 in mdadm.conf
108 - Get --stop to list devices stopped but honour --quiet
109 - remove libssl dependency
110 - Avoid some misdetection of overlapping partitions
111 - Fix memory leak in --monitor mode
113 Changes Prior to 2.5 release
114 - Support 'mailfrom' line in mdadm.conf so the From: line in alert
115 emails can be explicitly set.
116 - Arrange that SparesMissing (which is similar in import to
117 DegradedArray) generates an Email.
118 - Assume "DEVICE partitions" if no DEVICE line is given.
119 - Support new 'offset' layout for raid10.
120 - When creating a bitmap file, choose a chunksize to limit number
121 of bitmap chunks to 2 million. More than this can cause kmalloc
122 failure.
123 - New 'CREATE' line in mdadm.conf for defaults such as owner, group,
124 mode and auto-flag
125 - --detail checks if array has been started or not and includes that
126 in report.
127 - When using --update=uuid on an array with a bitmap, update the
128 bitmap's uuid too.
129 - Add a copy of /proc/mdstat to the mail message sent by mdadm
130 --monitor.
131 - New flag --no-degraded to avoid starting arrays if there are
132 fewer devices available than last time the array was started.
133 This is only needed with --scan, as with --scan, that behaviour
134 is the default.
135 - Support for 'homehost' concept. This is a fairly major update.
136 It includes a configfile option and a command line option for
137 specifying a homehost, records that host in the superblock,
138 and reports the homehost where possible.
139 - Support for Auto Assembly. "mdadm -As" will, if provided with
140 the name of a homehost, try to assemble all arrays it can find
141 that were created for that homehost. See man pages for more details.
143 Changes Prior to 2.4.1 release
144 - Honour --write-mostly when adding to an array without persistent
145 superblocks.
146 - Fix alignment problem in version-1 superblocks.
147 NOTE: This is an incompatable change affecting raid5 reshape.
148 If you want to reshape a raid5 using version-1 superblocks,
149 use 2.6.17-rc2 or later, and mdadm-2.4.1 or later.
151 Changes Prior to 2.4 release
152 - Rewrite 'reshape' support including performing a backup
153 of the critical region for a raid5 growth, and restoring that
154 backup after a crash.
155 - Put a 'canary' at each end of the backup so a corruption
156 can be more easily detected.
157 - Remove useless 'ident' arguement from ->getinfo_super method.
158 - Support --backup-file for backing-up critical section during
159 growth.
160 - Erase old superblocks (of different versions) when creating new
161 array.
162 - Allow --monitor to work with arrays with >28 devices
163 - Report reshape information in --detail
164 - Handle symlinks in /dev better
165 - Fix mess in --detail output which a device is missing.
166 - Manpage tidyup
167 - Support 'bitmap=' in mdadm.conf for auto-assembling arrays with
168 write-intent bitmaps in separate files.
169 - Updates to md.4 man page including section on RESTRIPING and SYSFS
171 Changes Prior to 2.3.1 release
172 - Fixed -O2 compile so I could make and RPM.
173 - Type cast number to be printed %llu so it compiles on 64bit
174 machines. (Thanks Luca).
175 - Stop using asm/byteorder.h - to make Redhat happy :-(
176 - Require bitmap files to have a '/' in their name.
177 - Error-check a few syscalls - code from SuSE package.
179 Changes Prior to 2.3 release
180 - Try /etc/mdadm/mdadm.conf if /etc/mdadm.conf doesn't exist.
181 This provided compatability for Debian.
182 - Fixed for version-1 superblock:
183 report chunksize for raid6 and raid10
184 make sure device size used is a multiple of chunksize
185 - Fix "--assemble --scan" crash.
186 - Fix completely failure to create array on ppc64
187 - Fix memcmp in place of memcpy
188 - A few minor improvements to online help
189 - Clean up usage of 'long long' for used-size of devices, so
190 that it is possible to create a raid1 of 7TB devices!
191 - Make internal bitmaps work on 7TB raid1 arrays.
192 - Provide error message if --examine doesn't find any superblock.
193 - Report 'reshape' status in --examine - this depends on kernel
194 patches that are not yet finalised.
195 - Report bitmap status in --detail and --examine
196 - Default to v1 superblocks instead of v0.90 if the array
197 is too big for 0.90 to handle.
198 - Sort the output of "mdadm --detail --scan" so that it is
199 in a suitable order for assembling arrays. i.e. components come
200 before an array that they are part of.
201 - Print size of large reiserfs array properly went warning of
202 possible confilcts.
204 Changes Prior to 2.2 release
205 - Assorted bug fixes
206 - Support write-intent-bitmaps on raid10
207 - Support little-endian (Rather than hostendian) bitmaps.
208 - Return correct error code from 'mdadm -S'
209 - Remove extra blank line from 'mdadm -Eb' output.
210 - Improve option parsing so that -a and -b do not have
211 optional arguements: the arg is either required or not
212 depending on context.
213 - Allow scanning of devices listed in /proc/partitions even
214 if they don't appear in /dev.
215 - Support --assume-clean in --create mode as well as --build
216 - Add support for --monitor to report to syslog: -y or --syslog.
217 Thanks to Ross Vandegrift
218 - --monitor now reports which device failed in a 'Fail' message
219 This broke with 2.6
220 - Improve chance of array starting properly after a crash.
221 mdadm was insisting the event numbers were identical, but this
222 isn't needed, and is a problem if the crash was while the metadata
223 was being updated.
224 - Support --update==uuid
225 - Added README.initramfs and mkinitramfs to help people use an
226 initram for starting md arrays at boot.
228 Changes Prior to 2.1 release
229 - Fix assembling of raid10 array when devices are missing.
230 mdadm now correctly detects if a array is workable or not
231 depending on which devices are present, and so will correctly
232 handle "--assemble --force" if multiple devices have failed.
233 - Report raid10 layout in --examine output.
234 - Fix assembling of arrays that use the version-1 superblock and
235 have spares. Previously the spares would be ignored.
236 - Fix bug so that multiple drives can be re-added at once.
237 - Fix problem with hot-adding a bitmap to version-1-superblock
238 arrays.
240 Changes Prior to 2.0
241 - Support assembling from byte-swapped superblocks
242 metadata type "0.swap" and --update=byteorder
243 - write-mostly and write-behind support for raid1.
244 - Support --name= and 'name=' config entry for identifying
245 arrays be name.
246 - RAID10 in man pages.
247 - Lot of minor manpage updates
249 Changes Prior to 2.0-devel-3 release
250 - Assorted fixes for multiple bugs...
251 - Add test suite
253 Changes Prior to 1.12.0 release
254 Several of these are backported from the Debian package
255 - Don't use 'lstat' to check for blockdevices, use stat.
256 - Document --size=max option for --grow
257 - Document SparesMissing event and DeviceDisappeared/WrongLevel
258 - --stop --scan repeatly cycles until no more progress can be made
259 so that stack devices are stopped properly
260 - Minor type rationalisation for ident->uuid - now always 'int[]'
261 - Fix type in online help for --grow (was -F, now -G)
262 - Allow --auto command line option to set default auto=
263 value when running "--assemble --scan". Previously
264 --auto was ignored if --scan was given
265 - Fix a few type casts
266 - Fix parsing of /dev/md/N in is_standard
267 - Fix rounding errors in human_size()
268 - Fix silly example in mdadm.conf-examples
269 - When finding a /dev name for a device, prefer shorter names
270 - Suppress listing of devices= in --brief output of -D or -E,
271 unless -v is given (-vv gives the old behaviour with -Dsv).
272 This is because the device list can change and so is not a
273 stable aspect of the array
274 - Allow --force with --grow so '-Gfn1' works (on raid1)
275 - Replace sprintf calls with snprintf (to quiet diet-libc)
276 - Add man page for mdassemble
277 - Fix compilation with tinyc
279 Changes Prior to 1.11.0 release
280 - Fix embarassing bug which causes --add to always fail.
282 Changes Prior to 1.10.0 release
283 - Fix bug with --config=partitions
284 - Open sub-devices with O_EXCL to detect if already in use
285 - Make sure superblock updates are flushed directly to disk.
287 Changes Prior to 2.0-devel-1 release
288 - Support for version-1 superblock. See --metadata option.
289 - Support for bitmap based intent logging.
290 - Minor fixes.
292 Changes Prior to 1.9.0 release
293 - Fix rpm build problem (stray %)
294 - Minor manpage updates
295 - Change "dirty" status to "active" as it was confusing people.
296 - --assemble --auto recognises 'standard' name and insists on using
297 the appropriate major/minor number for them.
298 - Remove underscore from partition names, so partitions of
299 "foo" are "foo1", "foo2" etc (unchanged) and partitions of
300 "f00" are "f00p1", "f00p2" etc rather than "f00_p1"...
301 - Use "major", "minor", "makedev" macros instead of
302 "MAJOR", "MINOR", "MKDEV" so that large device numbers work
303 on 2.6 (providing you have glibc 2.3.3 or later).
304 - Add some missing closes of open file descriptors.
305 - Reread /proc/partition for every array assembled when using
306 it to find devices, rather than only once.
307 - Make "mdadm -Ss" stop stacked devices properly, by reversing the
308 order in which arrays are stopped.
309 - Improve some error messages.
310 - Allow device name to appear before first option, so e.g.
311 mdadm /dev/md0 -A /dev/sd[ab]
312 works.
313 - Assume '-Q' if just a device is given, rather than being silent.
314 - Change "dirty" status to "active" as it was confusing people.
316 Changes Prior to 1.8.0 release
317 - Makefile cleanup from Luca Berra <bluca@comedia.it>
318 - --pid-file (-i) to set a pid file to use with --monitor --daemonise
319 - Fix typo in mdadm man page
320 - Fix coredump when "-s" used with no config file present.
321 - Support new "faulty" personality which can inject synthetic
322 faults. (Not in kernel.org yet at 1Nov2004)
323 - Support raid0/linear on devices > 2 Terabytes
324 - Make sure raid6 resyncs when created with one missing device
326 Changes Prior to 1.7.0 release
327 - Support "--grow --add" to add a device to a linear array, if the
328 kernel supports it. Not documented yet.
329 - Restore support for uclibc which was broken recently.
330 - Several improvements to the output of --detail, including
331 reporting "resyncing" or "recovering" in the state.
332 - Close filedescriptor at end of --detail (exit would have closed it
333 anyway, so this isn't abig deal).
334 - Report "Sync checkpoint" in --examine output if appropriate.
335 - Add --update=resync for --assemble mode to for a resync when the
336 array is assembled.
337 - Add support for "raid10", which is under development in 2.6.
338 Not documented yet.
339 - --monitor now reads spare-group and spares info from config file
340 even when names of arrays to scan are given on the command line
342 Changes Prior to 1.6.0 release
343 - Device name given in -Eb is determined by examining /dev rather
344 than assuming /dev/md%d
345 - Fix bug in --monitor where an array could be held open an so
346 could not be stopped without killing mdadm.
347 - Add --grow mode. Currently only --size and --raid-disks can be
348 changed. Both require kernel support which, at the time of
349 writing, is not in a release kernel yet.
350 - Don't print out "errors" or "no-errors" in -D and -E, as the bit
351 is never set or used.
352 - Use md event notification in 2.6.??? to make --monitor mode
353 respond instantly to events.
354 - Add --auto= option and auto= configfile entry to tell mdadm to
355 create device files as needed. This is particularly useful
356 with partitioned arrays where the major device number can change.
357 - When generating --brief listing, if the standard name doesn't
358 exist, search /dev for one rather than using a temp name.
359 - Allow --build to build raid1 and multipath arrays.
360 - Add "--assume-clean" for Create and Build, particularly for raid1
361 Note: this is dangerous. Only use it if you are certain.
362 - Fix bug so that Rebuild status monitoring works again.
363 - Add "degraded" and "recovering" options to the "Status:"
364 entry for --detail
366 Changes Prior to 1.5.0 release
367 - new commands "mdassemble" which is a stripped-down equivalent of
368 "mdadm -As", that can be compiled with dietlibc.
369 Thanks to Luca Berra <bluca@comedia.it>.
370 It can be using in an initramfs or initrd.
371 - Fix compiling error with BLKGETSIZE64 and some signed/unsigned
372 comparison warnings.
373 - Add Rebuild Status (% complete) to --detail output.
374 - Support "--monitor --test" which will generate a test alert
375 for each array once, to test notification paths.
376 - Generate RebuildFinished event when rebuild finishes.
377 - Support for raid6 as found in 2.6.2 - thanks to
378 H. Peter Anvin <hpa@zytor.com>
379 - Support partitioned md arrays with a different major number and
380 naming scheme (md_dX in /proc/mdstat, /dev/md/dXpY in /dev).
382 Changes Prior to 1.4.0 release
383 - Document fact that creating a raid5 array really creates a
384 degraded array with a spare.
385 - Add "spares=" tag to config file and generate it wit --detail and
386 --examine
387 - Add "SparesMissing" event when --monitor first sees an array and
388 it doesn't have the enough spare devices.
389 - Add --update=summaries for --assemble to update summary
390 information in superblock, and correct other inconsistancies in
391 the superblock.
392 - Add --test option to --detail to set a meaningful exit status.
394 Changes Prior to 1.3.0 release
395 - Make 'size' and unsigned long in Create to allow creation of
396 larger arrays.
397 - Explicitly flag spare devices as 'spare' in --detail and --examine
398 output. Previously they simply had no flags lists.
399 - Make MailCmd (for monitor) configurable in Makefile, and default
400 to "/usr/sbin/sendmail -t". Also split out the warning related
401 flags into CWFLAGS for easier build configurability.
402 - Minor bugfix in Manage code.
403 - --monitor now notices and reports degraded arrays at startup using
404 "DegradedArray" event, and also has a --oneshot option to only
405 report DegradedArrays, and then exit.
406 - Small man-page clarification w.r.t. raid levels and raid4 in
407 particular.
408 - Disallow creation of arrays with only one device as this is
409 probably a mistake. --force will override this check.
410 - Correct some misleading documentation in the "mdadm --create --help"
411 message.
412 - Ignore chunksize if raid1 or multipath.
413 - Explicit statement in man page that raid-disks cannot be changed
414 after array is created.
415 - Improve message when attempting to start an array with
416 insufficient devices. Instead of required the array to be full,
417 we only require it has as many active devices as last time.
419 Changes Prior to 1.2.0 release
420 - Fix bug where --daemonise required an argument.
421 - In --assemble --verbose, print appropriate message if device is
422 not in devices= list
423 - Updated mdadm.conf.5 to reflect fact that device= takes wildcards
424 - Typos: componenet -> component
425 - Reduce size of "--help" message put excess into "--help-options"
426 - Fix bug introduced when MD_SB_DISKS dependancy removed, and which
427 caused spares not be assembled properly.
428 - Print appropriate message if --monitor --scan decides not to
429 monitor anything.
430 Changes Prior to 1.1.0 release
431 - add --deamonise flag for --monitor - forks and prints pid to stdout
432 - Fix bug so we REALLY clear dirty flag with -Af
433 - -Db now prints a 'devices=' word for each array.
434 - "mdadm -A /dev/md0" will get info from configfile, even without scan
435 - When assembling multipath arrays, ignore devices which are flagged
436 as having errors.
437 - take --super-minor=dev to mean "use the minor number of the mddev
438 being assembled.
439 - take --config=none to mean "completely ignore config file"
440 - Make --monitor require --scan or a device list.
441 Changes Prior to 1.0.9 release
442 - Documentation updates including kernel parameters documented
443 in md.4
444 - --assemble --force for raid4/5 will mark clean, needed for 2.5
445 - --detail prints out the events counter as well
446 - flush device before reading superblock to be sure to get
447 current data
448 - added mdadm.static target to makefile for static linking
449 - --monitor was ignoring /dev/md0 due to off-by-one error
450 - Fix assorted typos
451 - Fix printing of Gibibytes - calc was wrong.
452 - Fix printing of Array Size in --detail when very big.
453 - --monitor no longer tries to work for raid0 or linear as these
454 have nothing to be monitored.
455 - The word 'partitions' on a DEVICE line will cause all partitions
456 listed in /proc/partitions to be considered
457 - If the config file is called 'partitions' then it will be treated
458 as though it contained exactly 'device partitions' so e.g.
459 mdadm -Ebsc partitions
460 will find all raid partitions easily.
461 - successfully assemble multipath devices by ignoring raid_disk
462 value from superblock (it is always the same).
463 - --assemble not tied to MD_SB_DISKS limit quite so much
464 - Support compiling with tcc
465 - Support compiling with uclibc - just skip scan of /dev
466 - Add --update= option for Assemble mode. Either sparc2.2
467 or super-minor updates are possible. See mdadm.8
469 Changes Prior to 1.0.1 release
470 - Round off MB/GiB etc values instead of round down.
471 - Add --sparc2.2 option to examine to shift superblock around
472 and --sparc2.2update to rewrite the superblock
473 - Fix assorted typos in online help
475 Changes Prior to 1.0.0 release
476 - Allow --config with Misc mode (for --examine --scan)
477 - Add $(CXFLAGS) to end of CFLAGS in makefile
478 - When making an N disk raid5 array, the Nth drive
479 is moved to the end of the array as a spare rather than
480 being shifted up one place. This means that when the
481 kernel builds onto the last spare and inserts it,
482 the devices will be in the expected order.
483 - Man page improvements
484 Changes Prior to 0.8.2 release
485 - Correct spelling of persist[ae]nce/persist[ae]nt.
486 - Change "disk" to "device" in options and config file
487 - convert array size to "long long" *before* shift-left in -D and -Q
489 Changes Prior to 0.8.1 release
490 - Add "INSTALL" file.
491 - Fix some "i" variables that were not being set properly
492 - Initialise minsize and maxsize so that compilers don't complain.
493 - Tidy up Makefile and mdadm.spec installations
494 - Add "multipath" to documentation of valid levels
496 Changes Prior to 0.8 release
497 - Fix another bug in Assemble.c due to confusing 'i' with 'j'
498 - Minimal, untested, support for multipath
499 - re-write of argument parsing to have more coherent modes,
500 - add --query,-Q option
501 - Update mdadm.8 to reflect arg processing change and --query
502 - Change "long" to "unsigned long" for device sizes
503 - Handle "mailaddr" and "program" lines in config file for follow/scan mode.
504 - --follow --scan will exit if no program or mail found
505 - Add MAILADDR and PROGRAM to mdadm.conf-example
506 - Spell check man pages
507 - consistently use "component devices" instead of "subdevices"
508 - Make -Wall -Werror really work and fix lots of errors.
509 - --detail and --stop can have --scan which chooses devices from /proc/mdstat
510 - --monitor detects 20% changes in resync, failed spares,
511 disappearing arrays,
512 - --monitor --scan will automatically add any devices found in /proc/mdstat
513 - --monitor will move spares between arrays with same spare-group if necessary
514 - Documentation for Monitor Mode
515 - --query notes if the array containing the given device is active or not
516 - Finished md.4 man page.
518 Changes Prior to 0.7.2 release
519 - mdadm.spec updates and ifdef BLKGETSIZE64 from Luca Berra -- bluca@comedia.it
520 - more mdadm.spec updates from Gregory Leblanc <gleblanc@linuxweasel.com>
521 - make directory for mdadm.conf configurable in Makefile
522 - Finished mdadm.conf.5. Removed details of conf file from
523 mdadm.8 leaving a reference to mdadm.conf.5.
524 - Fix bug in Assemble.c, thanks to Junaid Rizvi <domdev@sat.net.pk>
525 - Get --assemble --force to make sure old major/minor numbers are
526 consistant, as md.c worries about this :-(
529 Changes Prior to 0.7.1 release
530 - update mdadm.spec
531 - use BLKGETSIZE64 if available for array size
532 - give human readable as GiB/MiB and GB and MB, with 2 decimal point precision
533 - Only warn about size variation for raid1/4/5.
534 - Started md.4 man page
535 - Started mdadm.conf.5 man page
537 Changes Prior to 0.7 release
539 - Fix makefile to install binary at /sbin and not /sbin/sbin
540 Also install man page.
541 - Add --zero-superblock based on --destroywithextremeprejudice
542 from Dale Stephenson <steph@snapserver.com>
543 - change name to mdadm. It is palandromic, and much nicer to pronouce.
545 Changes Prior to 0.6 release
547 - Remove the limit on the number of device names that can be
548 given on the command line.
549 - Fix bug in --assemble --force where it would only update a
550 single superblock.
551 - Fix bogus printing of big numbers not being block devices
552 when given names of devices that don't exist.
553 - When --assemble --force, consider superblocks with an event
554 count that is 1 behind as out-of-date. Normally they are
555 considered up-to-date (as the kernel assumes this too).
556 - When marking drives as not-failed in the superblock,
557 we also mark them as ACTIVE and SYNC.
558 - Don't start arrays for which not all drives are available unless:
559 --scan which implies that all drives were found automatically
560 --run which means the user knows what they want
561 --force which means that we are fixing something broken
562 - Make sure all device numbers passed as 3rd arg of ioctl
563 are passed as unsigned lock, so that it works on SPARC
564 - If HOT_ADD_DISK failes for -a, then only try ADD_NEW_DISK
565 if we cannot read from the array, i.e. if the array is
566 not started yet.
567 - man page update
568 - Taught Examine to handle --scan. It examines all devices listed
569 on DEVICE lines in the config file.
570 - Added --brief (-b) flag for Examine and Detail to print out
571 and mdctl.conf compatible description with uuid=, level=,
572 disks= and - for Examine - devices=
573 --examine --brief collects all devices the make the one array and
574 list them as one entry.
575 - Added level= and disks= options to ARRAY lines in config files
576 so --brief output could be used as-is.
577 - Make parity style ({left,right}-{,a}symmetric) consistantly use -,
578 never _.
579 - Add "Array Size" to --detail output
580 - Change "Size" to "Device Size" and exclude from Detail of arrays
581 that do not have a consistent device size.
582 - Add Human readable MiB or GiB value on size lines of Detail and Examine
583 - --assemble --scan doesn't complain about active drives
584 - require number of spares given in -x to be listed.
585 - Made --build actually work.
586 Changes Prior to 0.5 release
588 --assemble:
589 spare drives are handled properly.
591 --force can be used to recover from 2-drive failures on RAID5
592 If you belive that /dev/hda1 /dev/hdb1 /dev/hdc1 /dev/hdd1 should
593 make a raid5 array, but it has experienced multiple failures and
594 wont start, then
596 mdctl --assemble --force /dev/md0 /dev/hd[abcd]1
598 Should update the superblock on the newest failed drive and
599 restart the array in degraded mode. You should then remove the
600 remaining failed drive and re-add it (if you are happy that it
601 might work).
603 Ofcourse whenever you have a 2-drive failure, you have a risk
604 of corruption in data that hasn't be changed for a long time. So
605 this doesn't give you your array back all nice and happy, but it
606 does allow you to recover data that might not be corrupt.
608 More flexibility in identifying a RAID array in the mdctl.conf
609 e.g.
610 array /dev/md4 super-minor=4
612 assembles /dev/md4 from all devices found that have a raid
613 superblock that says the minor number of the array is 4.
614 If the blocks with the right minor number do not all have the
615 same UUID, an error is flags and no assembly happens.
617 array /dev/md3 devices=/dev/hd[abc]2
619 Assembles /dev/md3 drom /dev/hda2 /dev/hdb2 and/dev/hdc2. All
620 devices must exist and have raid superblock with the same uuid.
622 If two identity specifiers are used, only devices that match all
623 of them are considered, so
625 array /dev/md2 devices=/dev/hd?2 super-minor=2
627 will assemble /dev/md2 using all /dev/hd?2 devices which have a
628 raid superblock with minor number 2.
630 --create:
631 When listing devices for --create, the word "missing" can be
632 used to indicate that the respective slot does not have a
633 working drive currently. This is similar to the "failed-disk"
634 directive in mkraid/raidtab.
635 e.g.
636 mdctl --create --level=5 -raid-disks=4 --spare-disks=2
637 /dev/md0 /dev/sda /dev/sdb missing /dev/sdc /dev/sdd /dev/sde
639 will create a raid5 array with the third slot empty, and two
640 spares.
642 By default, raid5 arrays are created with the last slot empty
643 and drive listed for the last slot added as a spare. If a
644 "missing" slot is given, or if --force is given, then --create
645 does exactly what you ask and doesn't try to be clever.
648 --follow / --monitor:
650 This is a new mode. I couldn't stop my self from picking a name
651 starting with F (as current modes start A,B,C,D,E) but I
652 relented and provided an alternate name that is somewhat more
653 meaningful.
655 In this mode, mdctl does not exit, but runs continuously and
656 periodically polls all the md devices to see if they have had
657 any interested state change.
658 The changes that it currently notices are:
659 Fail - an active disc fails
660 FailSpare - a spare, that was presumably being build, fails
661 ActiveSpare - a spare becomes active, presumably after a rebuild.
663 Options:
664 --mail mailaddress - send Email on any Fail* event
665 --program program - run the program on any event.
666 Args are: eventname mddevice subdevice(if-known)
667 --delay seconds - change from the default 60second pause
668 between polls.
670 I plan to add functionality to this mode to allow sharing of
671 spare drives. If an array is marks "spare-group=fred", and it
672 has a failed drive and no spares, and if some other array is
673 also "spare-group=fred" and it has no failed drives, but does
674 have a spare drive that is big enough, the spare will be moved
675 to the first array.
677 I also have the idea of adding a --grow mode which will re-organise
678 the data on an N disk raid0/4/5 array to be on an N+M disk array.
679 I have no concrete plans for this though.
681 I got rid of the "v" in the archive file name, and include the
682 version number in the directory created by the archive.
684 There is now a man page and mdctl.spec (for rpm) thanks to
685 Danilo Godec <danci@agenda.si>.
687 Ofcourse, the man page is now out of date and despite being based on
688 the --help output, is not wholy correct. After I get --follow
689 working properly, I plan to revise the various documentation and/or
690 the code to make sure the two match.