[thirdparty/mdadm.git] / ChangeLog
1Changes Prior to this release
2 - Support "--grow --add" to add a device to a linear array, if the
3 kernel supports it. Not documented yet.
4 - Restore support for uclibc which was broken recently.
5 - Several improvements to the output of --detail, including
6 reporting "resyncing" or "recovering" in the state.
7 - Close filedescriptor at end of --detail (exit would have closed it
8 anyway, so this isn't abig deal).
9 - Report "Sync checkpoint" in --examine output if appropriate.
10 - Add --update=resync for --assemble mode to for a resync when the
11 array is assembled.
12 - Add support for "raid10", which is under development in 2.6.
13 Not documented yet.
14 - --monitor now reads spare-group and spares info from config file
15 even when names of arrays to scan are given on the command line
17Changes Prior to 1.6.0 release
18 - Device name given in -Eb is determined by examining /dev rather
19 than assuming /dev/md%d
20 - Fix bug in --monitor where an array could be held open an so
21 could not be stopped without killing mdadm.
22 - Add --grow mode. Currently only --size and --raid-disks can be
23 changed. Both require kernel support which, at the time of
24 writing, is not in a release kernel yet.
25 - Don't print out "errors" or "no-errors" in -D and -E, as the bit
26 is never set or used.
27 - Use md event notification in 2.6.??? to make --monitor mode
28 respond instantly to events.
29 - Add --auto= option and auto= configfile entry to tell mdadm to
30 create device files as needed. This is particularly useful
31 with partitioned arrays where the major device number can change.
32 - When generating --brief listing, if the standard name doesn't
33 exist, search /dev for one rather than using a temp name.
34 - Allow --build to build raid1 and multipath arrays.
35 - Add "--assume-clean" for Create and Build, particularly for raid1
36 Note: this is dangerous. Only use it if you are certain.
37 - Fix bug so that Rebuild status monitoring works again.
38 - Add "degraded" and "recovering" options to the "Status:"
39 entry for --detail
41Changes Prior to 1.5.0 release
42 - new commands "mdassemble" which is a stripped-down equivalent of
43 "mdadm -As", that can be compiled with dietlibc.
44 Thanks to Luca Berra <>.
45 It can be using in an initramfs or initrd.
46 - Fix compiling error with BLKGETSIZE64 and some signed/unsigned
47 comparison warnings.
48 - Add Rebuild Status (% complete) to --detail output.
49 - Support "--monitor --test" which will generate a test alert
50 for each array once, to test notification paths.
51 - Generate RebuildFinished event when rebuild finishes.
52 - Support for raid6 as found in 2.6.2 - thanks to
53 H. Peter Anvin <>
54 - Support partitioned md arrays with a different major number and
55 naming scheme (md_dX in /proc/mdstat, /dev/md/dXpY in /dev).
57Changes Prior to 1.4.0 release
58 - Document fact that creating a raid5 array really creates a
59 degraded array with a spare.
60 - Add "spares=" tag to config file and generate it wit --detail and
61 --examine
62 - Add "SparesMissing" event when --monitor first sees an array and
63 it doesn't have the enough spare devices.
64 - Add --update=summaries for --assemble to update summary
65 information in superblock, and correct other inconsistancies in
66 the superblock.
67 - Add --test option to --detail to set a meaningful exit status.
69Changes Prior to 1.3.0 release
70 - Make 'size' and unsigned long in Create to allow creation of
71 larger arrays.
72 - Explicitly flag spare devices as 'spare' in --detail and --examine
73 output. Previously they simply had no flags lists.
74 - Make MailCmd (for monitor) configurable in Makefile, and default
75 to "/usr/sbin/sendmail -t". Also split out the warning related
76 flags into CWFLAGS for easier build configurability.
77 - Minor bugfix in Manage code.
78 - --monitor now notices and reports degraded arrays at startup using
79 "DegradedArray" event, and also has a --oneshot option to only
80 report DegradedArrays, and then exit.
81 - Small man-page clarification w.r.t. raid levels and raid4 in
82 particular.
83 - Disallow creation of arrays with only one device as this is
84 probably a mistake. --force will override this check.
85 - Correct some misleading documentation in the "mdadm --create --help"
86 message.
87 - Ignore chunksize if raid1 or multipath.
88 - Explicit statement in man page that raid-disks cannot be changed
89 after array is created.
90 - Improve message when attempting to start an array with
91 insufficient devices. Instead of required the array to be full,
92 we only require it has as many active devices as last time.
94Changes Prior to 1.2.0 release
95 - Fix bug where --daemonise required an argument.
96 - In --assemble --verbose, print appropriate message if device is
97 not in devices= list
98 - Updated mdadm.conf.5 to reflect fact that device= takes wildcards
99 - Typos: componenet -> component
100 - Reduce size of "--help" message put excess into "--help-options"
101 - Fix bug introduced when MD_SB_DISKS dependancy removed, and which
102 caused spares not be assembled properly.
103 - Print appropriate message if --monitor --scan decides not to
104 monitor anything.
105Changes Prior to 1.1.0 release
106 - add --deamonise flag for --monitor - forks and prints pid to stdout
107 - Fix bug so we REALLY clear dirty flag with -Af
108 - -Db now prints a 'devices=' word for each array.
109 - "mdadm -A /dev/md0" will get info from configfile, even without scan
110 - When assembling multipath arrays, ignore devices which are flagged
111 as having errors.
112 - take --super-minor=dev to mean "use the minor number of the mddev
113 being assembled.
114 - take --config=none to mean "completely ignore config file"
115 - Make --monitor require --scan or a device list.
116Changes Prior to 1.0.9 release
117 - Documentation updates including kernel parameters documented
118 in md.4
119 - --assemble --force for raid4/5 will mark clean, needed for 2.5
120 - --detail prints out the events counter as well
121 - flush device before reading superblock to be sure to get
122 current data
123 - added mdadm.static target to makefile for static linking
124 - --monitor was ignoring /dev/md0 due to off-by-one error
125 - Fix assorted typos
126 - Fix printing of Gibibytes - calc was wrong.
127 - Fix printing of Array Size in --detail when very big.
128 - --monitor no longer tries to work for raid0 or linear as these
d013a55e 129 have nothing to be monitored.
130 - The word 'partitions' on a DEVICE line will cause all partitions
131 listed in /proc/partitions to be considered
132 - If the config file is called 'partitions' then it will be treated
133 as though it contained exactly 'device partitions' so e.g.
134 mdadm -Ebsc partitions
d013a55e 135 will find all raid partitions easily.
136 - successfully assemble multipath devices by ignoring raid_disk
137 value from superblock (it is always the same).
138 - --assemble not tied to MD_SB_DISKS limit quite so much
139 - Support compiling with tcc
140 - Support compiling with uclibc - just skip scan of /dev
141 - Add --update= option for Assemble mode. Either sparc2.2
142 or super-minor updates are possible. See mdadm.8
144Changes Prior to 1.0.1 release
145 - Round off MB/GiB etc values instead of round down.
146 - Add --sparc2.2 option to examine to shift superblock around
147 and --sparc2.2update to rewrite the superblock
5787fa49 148 - Fix assorted typos in online help
150Changes Prior to 1.0.0 release
151 - Allow --config with Misc mode (for --examine --scan)
152 - Add $(CXFLAGS) to end of CFLAGS in makefile
153 - When making an N disk raid5 array, the Nth drive
154 is moved to the end of the array as a spare rather than
155 being shifted up one place. This means that when the
156 kernel builds onto the last spare and inserts it,
157 the devices will be in the expected order.
158 - Man page improvements
159Changes Prior to 0.8.2 release
160 - Correct spelling of persist[ae]nce/persist[ae]nt.
161 - Change "disk" to "device" in options and config file
162 - convert array size to "long long" *before* shift-left in -D and -Q
164Changes Prior to 0.8.1 release
165 - Add "INSTALL" file.
166 - Fix some "i" variables that were not being set properly
167 - Initialise minsize and maxsize so that compilers don't complain.
168 - Tidy up Makefile and mdadm.spec installations
169 - Add "multipath" to documentation of valid levels
171Changes Prior to 0.8 release
172 - Fix another bug in Assemble.c due to confusing 'i' with 'j'
173 - Minimal, untested, support for multipath
174 - re-write of argument parsing to have more coherent modes,
175 - add --query,-Q option
176 - Update mdadm.8 to reflect arg processing change and --query
177 - Change "long" to "unsigned long" for device sizes
178 - Handle "mailaddr" and "program" lines in config file for follow/scan mode.
179 - --follow --scan will exit if no program or mail found
180 - Add MAILADDR and PROGRAM to mdadm.conf-example
181 - Spell check man pages
182 - consistently use "component devices" instead of "subdevices"
183 - Make -Wall -Werror really work and fix lots of errors.
184 - --detail and --stop can have --scan which chooses devices from /proc/mdstat
185 - --monitor detects 20% changes in resync, failed spares,
186 disappearing arrays,
187 - --monitor --scan will automatically add any devices found in /proc/mdstat
188 - --monitor will move spares between arrays with same spare-group if necessary
189 - Documentation for Monitor Mode
190 - --query notes if the array containing the given device is active or not
191 - Finished md.4 man page.
193Changes Prior to 0.7.2 release
194 - mdadm.spec updates and ifdef BLKGETSIZE64 from Luca Berra --
195 - more mdadm.spec updates from Gregory Leblanc <>
196 - make directory for mdadm.conf configurable in Makefile
197 - Finished mdadm.conf.5. Removed details of conf file from
198 mdadm.8 leaving a reference to mdadm.conf.5.
199 - Fix bug in Assemble.c, thanks to Junaid Rizvi <>
200 - Get --assemble --force to make sure old major/minor numbers are
201 consistant, as md.c worries about this :-(
204Changes Prior to 0.7.1 release
205 - update mdadm.spec
206 - use BLKGETSIZE64 if available for array size
207 - give human readable as GiB/MiB and GB and MB, with 2 decimal point precision
208 - Only warn about size variation for raid1/4/5.
209 - Started md.4 man page
210 - Started mdadm.conf.5 man page
212Changes Prior to 0.7 release
214 - Fix makefile to install binary at /sbin and not /sbin/sbin
215 Also install man page.
216 - Add --zero-superblock based on --destroywithextremeprejudice
217 from Dale Stephenson <>
218 - change name to mdadm. It is palandromic, and much nicer to pronouce.
220Changes Prior to 0.6 release
9a9dab36 222 - Remove the limit on the number of device names that can be
223 given on the command line.
224 - Fix bug in --assemble --force where it would only update a
225 single superblock.
226 - Fix bogus printing of big numbers not being block devices
227 when given names of devices that don't exist.
228 - When --assemble --force, consider superblocks with an event
229 count that is 1 behind as out-of-date. Normally they are
230 considered up-to-date (as the kernel assumes this too).
231 - When marking drives as not-failed in the superblock,
232 we also mark them as ACTIVE and SYNC.
233 - Don't start arrays for which not all drives are available unless:
234 --scan which implies that all drives were found automatically
235 --run which means the user knows what they want
236 --force which means that we are fixing something broken
237 - Make sure all device numbers passed as 3rd arg of ioctl
238 are passed as unsigned lock, so that it works on SPARC
239 - If HOT_ADD_DISK failes for -a, then only try ADD_NEW_DISK
240 if we cannot read from the array, i.e. if the array is
241 not started yet.
242 - man page update
243 - Taught Examine to handle --scan. It examines all devices listed
244 on DEVICE lines in the config file.
245 - Added --brief (-b) flag for Examine and Detail to print out
246 and mdctl.conf compatible description with uuid=, level=,
247 disks= and - for Examine - devices=
248 --examine --brief collects all devices the make the one array and
249 list them as one entry.
250 - Added level= and disks= options to ARRAY lines in config files
251 so --brief output could be used as-is.
252 - Make parity style ({left,right}-{,a}symmetric) consistantly use -,
253 never _.
254 - Add "Array Size" to --detail output
255 - Change "Size" to "Device Size" and exclude from Detail of arrays
256 that do not have a consistent device size.
257 - Add Human readable MiB or GiB value on size lines of Detail and Examine
258 - --assemble --scan doesn't complain about active drives
259 - require number of spares given in -x to be listed.
260 - Made --build actually work.
261Changes Prior to 0.5 release
263 --assemble:
264 spare drives are handled properly.
266 --force can be used to recover from 2-drive failures on RAID5
267 If you belive that /dev/hda1 /dev/hdb1 /dev/hdc1 /dev/hdd1 should
268 make a raid5 array, but it has experienced multiple failures and
269 wont start, then
271 mdctl --assemble --force /dev/md0 /dev/hd[abcd]1
273 Should update the superblock on the newest failed drive and
274 restart the array in degraded mode. You should then remove the
275 remaining failed drive and re-add it (if you are happy that it
276 might work).
278 Ofcourse whenever you have a 2-drive failure, you have a risk
279 of corruption in data that hasn't be changed for a long time. So
280 this doesn't give you your array back all nice and happy, but it
281 does allow you to recover data that might not be corrupt.
283 More flexibility in identifying a RAID array in the mdctl.conf
284 e.g.
285 array /dev/md4 super-minor=4
287 assembles /dev/md4 from all devices found that have a raid
288 superblock that says the minor number of the array is 4.
289 If the blocks with the right minor number do not all have the
290 same UUID, an error is flags and no assembly happens.
292 array /dev/md3 devices=/dev/hd[abc]2
294 Assembles /dev/md3 drom /dev/hda2 /dev/hdb2 and/dev/hdc2. All
295 devices must exist and have raid superblock with the same uuid.
297 If two identity specifiers are used, only devices that match all
298 of them are considered, so
300 array /dev/md2 devices=/dev/hd?2 super-minor=2
302 will assemble /dev/md2 using all /dev/hd?2 devices which have a
303 raid superblock with minor number 2.
305 --create:
306 When listing devices for --create, the word "missing" can be
307 used to indicate that the respective slot does not have a
308 working drive currently. This is similar to the "failed-disk"
309 directive in mkraid/raidtab.
310 e.g.
311 mdctl --create --level=5 -raid-disks=4 --spare-disks=2
312 /dev/md0 /dev/sda /dev/sdb missing /dev/sdc /dev/sdd /dev/sde
314 will create a raid5 array with the third slot empty, and two
315 spares.
317 By default, raid5 arrays are created with the last slot empty
318 and drive listed for the last slot added as a spare. If a
319 "missing" slot is given, or if --force is given, then --create
320 does exactly what you ask and doesn't try to be clever.
323 --follow / --monitor:
325 This is a new mode. I couldn't stop my self from picking a name
326 starting with F (as current modes start A,B,C,D,E) but I
327 relented and provided an alternate name that is somewhat more
328 meaningful.
330 In this mode, mdctl does not exit, but runs continuously and
331 periodically polls all the md devices to see if they have had
332 any interested state change.
333 The changes that it currently notices are:
334 Fail - an active disc fails
335 FailSpare - a spare, that was presumably being build, fails
336 ActiveSpare - a spare becomes active, presumably after a rebuild.
338 Options:
339 --mail mailaddress - send Email on any Fail* event
340 --program program - run the program on any event.
341 Args are: eventname mddevice subdevice(if-known)
342 --delay seconds - change from the default 60second pause
343 between polls.
345 I plan to add functionality to this mode to allow sharing of
346 spare drives. If an array is marks "spare-group=fred", and it
347 has a failed drive and no spares, and if some other array is
348 also "spare-group=fred" and it has no failed drives, but does
349 have a spare drive that is big enough, the spare will be moved
350 to the first array.
352 I also have the idea of adding a --grow mode which will re-organise
353 the data on an N disk raid0/4/5 array to be on an N+M disk array.
354 I have no concrete plans for this though.
356 I got rid of the "v" in the archive file name, and include the
357 version number in the directory created by the archive.
359 There is now a man page and mdctl.spec (for rpm) thanks to
360 Danilo Godec <>.
362 Ofcourse, the man page is now out of date and despite being based on
363 the --help output, is not wholy correct. After I get --follow
364 working properly, I plan to revise the various documentation and/or
365 the code to make sure the two match.