mdadm-1.3.0
[thirdparty/mdadm.git] / ChangeLog
CommitLineData
aa88f531
NB
1Changes Prior to 1.3.0 release
2 - Make 'size' and unsigned long in Create to allow creation of
3 larger arrays.
4 - Explicitly flag spare devices as 'spare' in --detail and --examine
5 output. Previously they simply had no flags lists.
6 - Make MailCmd (for monitor) configurable in Makefile, and default
7 to "/usr/sbin/sendmail -t". Also split out the warning related
8 flags into CWFLAGS for easier build configurability.
9 - Minor bugfix in Manage code.
10 - --monitor now notices and reports degraded arrays at startup using
11 "DegradedArray" event, and also has a --oneshot option to only
12 report DegradedArrays, and then exit.
13 - Small man-page clarification w.r.t. raid levels and raid4 in
14 particular.
15 - Disallow creation of arrays with only one device as this is
16 probably a mistake. --force will override this check.
17 - Correct some misleading documentation in the "mdadm --create --help"
18 message.
19 - Ignore chunksize if raid1 or multipath.
20 - Explicit statement in man page that raid-disks cannot be changed
21 after array is created.
22 - Improve message when attempting to start an array with
23 insufficient devices. Instead of required the array to be full,
24 we only require it has as many active devices as last time.
25
56eedc1a
NB
26Changes Prior to 1.2.0 release
27 - Fix bug where --daemonise required an argument.
28 - In --assemble --verbose, print appropriate message if device is
29 not in devices= list
30 - Updated mdadm.conf.5 to reflect fact that device= takes wildcards
31 - Typos: componenet -> component
32 - Reduce size of "--help" message put excess into "--help-options"
33 - Fix bug introduced when MD_SB_DISKS dependancy removed, and which
34 caused spares not be assembled properly.
35 - Print appropriate message if --monitor --scan decides not to
36 monitor anything.
d013a55e
NB
37Changes Prior to 1.1.0 release
38 - add --deamonise flag for --monitor - forks and prints pid to stdout
39 - Fix bug so we REALLY clear dirty flag with -Af
40 - -Db now prints a 'devices=' word for each array.
41 - "mdadm -A /dev/md0" will get info from configfile, even without scan
42 - When assembling multipath arrays, ignore devices which are flagged
43 as having errors.
44 - take --super-minor=dev to mean "use the minor number of the mddev
45 being assembled.
46 - take --config=none to mean "completely ignore config file"
47 - Make --monitor require --scan or a device list.
5787fa49
NB
48Changes Prior to 1.0.9 release
49 - Documentation updates including kernel parameters documented
50 in md.4
51 - --assemble --force for raid4/5 will mark clean, needed for 2.5
52 - --detail prints out the events counter as well
53 - flush device before reading superblock to be sure to get
54 current data
55 - added mdadm.static target to makefile for static linking
56 - --monitor was ignoring /dev/md0 due to off-by-one error
57 - Fix assorted typos
58 - Fix printing of Gibibytes - calc was wrong.
59 - Fix printing of Array Size in --detail when very big.
60 - --monitor no longer tries to work for raid0 or linear as these
d013a55e 61 have nothing to be monitored.
5787fa49
NB
62 - The word 'partitions' on a DEVICE line will cause all partitions
63 listed in /proc/partitions to be considered
64 - If the config file is called 'partitions' then it will be treated
65 as though it contained exactly 'device partitions' so e.g.
66 mdadm -Ebsc partitions
d013a55e 67 will find all raid partitions easily.
5787fa49
NB
68 - successfully assemble multipath devices by ignoring raid_disk
69 value from superblock (it is always the same).
70 - --assemble not tied to MD_SB_DISKS limit quite so much
71 - Support compiling with tcc
72 - Support compiling with uclibc - just skip scan of /dev
73 - Add --update= option for Assemble mode. Either sparc2.2
74 or super-minor updates are possible. See mdadm.8
75
76Changes Prior to 1.0.1 release
bd526cee
NB
77 - Round off MB/GiB etc values instead of round down.
78 - Add --sparc2.2 option to examine to shift superblock around
79 and --sparc2.2update to rewrite the superblock
5787fa49 80 - Fix assorted typos in online help
bd526cee
NB
81
82Changes Prior to 1.0.0 release
2d465520
NB
83 - Allow --config with Misc mode (for --examine --scan)
84 - Add $(CXFLAGS) to end of CFLAGS in makefile
85 - When making an N disk raid5 array, the Nth drive
86 is moved to the end of the array as a spare rather than
87 being shifted up one place. This means that when the
88 kernel builds onto the last spare and inserts it,
89 the devices will be in the expected order.
90 - Man page improvements
91Changes Prior to 0.8.2 release
b83d95f3
NB
92 - Correct spelling of persist[ae]nce/persist[ae]nt.
93 - Change "disk" to "device" in options and config file
94 - convert array size to "long long" *before* shift-left in -D and -Q
95
96Changes Prior to 0.8.1 release
c913b90e
NB
97 - Add "INSTALL" file.
98 - Fix some "i" variables that were not being set properly
99 - Initialise minsize and maxsize so that compilers don't complain.
100 - Tidy up Makefile and mdadm.spec installations
101 - Add "multipath" to documentation of valid levels
102
103Changes Prior to 0.8 release
e0d19036
NB
104 - Fix another bug in Assemble.c due to confusing 'i' with 'j'
105 - Minimal, untested, support for multipath
106 - re-write of argument parsing to have more coherent modes,
107 - add --query,-Q option
108 - Update mdadm.8 to reflect arg processing change and --query
109 - Change "long" to "unsigned long" for device sizes
110 - Handle "mailaddr" and "program" lines in config file for follow/scan mode.
111 - --follow --scan will exit if no program or mail found
112 - Add MAILADDR and PROGRAM to mdadm.conf-example
113 - Spell check man pages
114 - consistently use "component devices" instead of "subdevices"
115 - Make -Wall -Werror really work and fix lots of errors.
116 - --detail and --stop can have --scan which chooses devices from /proc/mdstat
117 - --monitor detects 20% changes in resync, failed spares,
118 disappearing arrays,
119 - --monitor --scan will automatically add any devices found in /proc/mdstat
120 - --monitor will move spares between arrays with same spare-group if necessary
121 - Documentation for Monitor Mode
122 - --query notes if the array containing the given device is active or not
123 - Finished md.4 man page.
124
125Changes Prior to 0.7.2 release
11a3e71d
NB
126 - mdadm.spec updates and ifdef BLKGETSIZE64 from Luca Berra -- bluca@comedia.it
127 - more mdadm.spec updates from Gregory Leblanc <gleblanc@linuxweasel.com>
128 - make directory for mdadm.conf configurable in Makefile
129 - Finished mdadm.conf.5. Removed details of conf file from
130 mdadm.8 leaving a reference to mdadm.conf.5.
131 - Fix bug in Assemble.c, thanks to Junaid Rizvi <domdev@sat.net.pk>
132 - Get --assemble --force to make sure old major/minor numbers are
133 consistant, as md.c worries about this :-(
134
135
136Changes Prior to 0.7.1 release
56eb10c0
NB
137 - update mdadm.spec
138 - use BLKGETSIZE64 if available for array size
139 - give human readable as GiB/MiB and GB and MB, with 2 decimal point precision
140 - Only warn about size variation for raid1/4/5.
141 - Started md.4 man page
142 - Started mdadm.conf.5 man page
143
9a9dab36
NB
144Changes Prior to 0.7 release
145
146 - Fix makefile to install binary at /sbin and not /sbin/sbin
147 Also install man page.
148 - Add --zero-superblock based on --destroywithextremeprejudice
149 from Dale Stephenson <steph@snapserver.com>
150 - change name to mdadm. It is palandromic, and much nicer to pronouce.
151
cd29a5c8
NB
152Changes Prior to 0.6 release
153
9a9dab36 154 - Remove the limit on the number of device names that can be
cd29a5c8
NB
155 given on the command line.
156 - Fix bug in --assemble --force where it would only update a
157 single superblock.
158 - Fix bogus printing of big numbers not being block devices
159 when given names of devices that don't exist.
160 - When --assemble --force, consider superblocks with an event
161 count that is 1 behind as out-of-date. Normally they are
162 considered up-to-date (as the kernel assumes this too).
163 - When marking drives as not-failed in the superblock,
164 we also mark them as ACTIVE and SYNC.
165 - Don't start arrays for which not all drives are available unless:
166 --scan which implies that all drives were found automatically
167 --run which means the user knows what they want
168 --force which means that we are fixing something broken
169 - Make sure all device numbers passed as 3rd arg of ioctl
170 are passed as unsigned lock, so that it works on SPARC
171 - If HOT_ADD_DISK failes for -a, then only try ADD_NEW_DISK
172 if we cannot read from the array, i.e. if the array is
173 not started yet.
174 - man page update
175 - Taught Examine to handle --scan. It examines all devices listed
176 on DEVICE lines in the config file.
177 - Added --brief (-b) flag for Examine and Detail to print out
178 and mdctl.conf compatible description with uuid=, level=,
179 disks= and - for Examine - devices=
180 --examine --brief collects all devices the make the one array and
181 list them as one entry.
182 - Added level= and disks= options to ARRAY lines in config files
183 so --brief output could be used as-is.
184 - Make parity style ({left,right}-{,a}symmetric) consistantly use -,
185 never _.
186 - Add "Array Size" to --detail output
187 - Change "Size" to "Device Size" and exclude from Detail of arrays
188 that do not have a consistent device size.
189 - Add Human readable MiB or GiB value on size lines of Detail and Examine
190 - --assemble --scan doesn't complain about active drives
191 - require number of spares given in -x to be listed.
192 - Made --build actually work.
193Changes Prior to 0.5 release
194
195 --assemble:
196 spare drives are handled properly.
197
198 --force can be used to recover from 2-drive failures on RAID5
199 If you belive that /dev/hda1 /dev/hdb1 /dev/hdc1 /dev/hdd1 should
200 make a raid5 array, but it has experienced multiple failures and
201 wont start, then
202
203 mdctl --assemble --force /dev/md0 /dev/hd[abcd]1
204
205 Should update the superblock on the newest failed drive and
206 restart the array in degraded mode. You should then remove the
207 remaining failed drive and re-add it (if you are happy that it
208 might work).
209
210 Ofcourse whenever you have a 2-drive failure, you have a risk
211 of corruption in data that hasn't be changed for a long time. So
212 this doesn't give you your array back all nice and happy, but it
213 does allow you to recover data that might not be corrupt.
214
215 More flexibility in identifying a RAID array in the mdctl.conf
216 e.g.
217 array /dev/md4 super-minor=4
218
219 assembles /dev/md4 from all devices found that have a raid
220 superblock that says the minor number of the array is 4.
221 If the blocks with the right minor number do not all have the
222 same UUID, an error is flags and no assembly happens.
223
224 array /dev/md3 devices=/dev/hd[abc]2
225
226 Assembles /dev/md3 drom /dev/hda2 /dev/hdb2 and/dev/hdc2. All
227 devices must exist and have raid superblock with the same uuid.
228
229 If two identity specifiers are used, only devices that match all
230 of them are considered, so
231
232 array /dev/md2 devices=/dev/hd?2 super-minor=2
233
234 will assemble /dev/md2 using all /dev/hd?2 devices which have a
235 raid superblock with minor number 2.
236
237 --create:
238 When listing devices for --create, the word "missing" can be
239 used to indicate that the respective slot does not have a
240 working drive currently. This is similar to the "failed-disk"
241 directive in mkraid/raidtab.
242 e.g.
243 mdctl --create --level=5 -raid-disks=4 --spare-disks=2
244 /dev/md0 /dev/sda /dev/sdb missing /dev/sdc /dev/sdd /dev/sde
245
246 will create a raid5 array with the third slot empty, and two
247 spares.
248
249 By default, raid5 arrays are created with the last slot empty
250 and drive listed for the last slot added as a spare. If a
251 "missing" slot is given, or if --force is given, then --create
252 does exactly what you ask and doesn't try to be clever.
253
254
255 --follow / --monitor:
256
257 This is a new mode. I couldn't stop my self from picking a name
258 starting with F (as current modes start A,B,C,D,E) but I
259 relented and provided an alternate name that is somewhat more
260 meaningful.
261
262 In this mode, mdctl does not exit, but runs continuously and
263 periodically polls all the md devices to see if they have had
264 any interested state change.
265 The changes that it currently notices are:
266 Fail - an active disc fails
267 FailSpare - a spare, that was presumably being build, fails
268 ActiveSpare - a spare becomes active, presumably after a rebuild.
269
270 Options:
271 --mail mailaddress - send Email on any Fail* event
272 --program program - run the program on any event.
273 Args are: eventname mddevice subdevice(if-known)
274 --delay seconds - change from the default 60second pause
275 between polls.
276
277 I plan to add functionality to this mode to allow sharing of
278 spare drives. If an array is marks "spare-group=fred", and it
279 has a failed drive and no spares, and if some other array is
280 also "spare-group=fred" and it has no failed drives, but does
281 have a spare drive that is big enough, the spare will be moved
282 to the first array.
283
284 I also have the idea of adding a --grow mode which will re-organise
285 the data on an N disk raid0/4/5 array to be on an N+M disk array.
286 I have no concrete plans for this though.
287
288 I got rid of the "v" in the archive file name, and include the
289 version number in the directory created by the archive.
290
291 There is now a man page and mdctl.spec (for rpm) thanks to
292 Danilo Godec <danci@agenda.si>.
293
294 Ofcourse, the man page is now out of date and despite being based on
295 the --help output, is not wholy correct. After I get --follow
296 working properly, I plan to revise the various documentation and/or
297 the code to make sure the two match.
298