TODO

   1 * --assemble could have a --update option.
   2   following word can be:
   3         sparc2.2
   4         super-minor
   5
   6 * mdadm /dev/md11, where md11 is raid0 can segfault, particularly when looking in the
   7    [UU_UUU] string ... which doesn't exist !
   8 It should be more sensible.  DONE
   9
  10 Example:
  11
  12 from  Raimund Sacherer <raimund.sacherer@ngit.at>
  13
  14 mke2fs -m0 -q /dev/ram1 300
  15 mount -n -t ext2 /dev/ram1 /tmp
  16 echo DEVICE /dev/[sh]* >> /tmp/mdadm.conf
  17 mdadm -Esb /dev/[sh]* 2>/dev/null >> /tmp/mdadm.conf
  18 mdadm -ARsc /tmp/mdadm.conf
  19 umount /tmp
  20
  21
  22 ?? Allow -S /dev/md? - current complains subsequent not a/d/r - DONE
  23
  24 * new "Query" mode to subsume --detail and --examine.
  25    --query or -Q, takes a device and tells if it is an MD device,
  26    and also tells in a raid superblock is found.
  27  DONE
  28
  29 * write mdstat.c to parse /proc/mdstat file
  30    Build list of arrays:  name, rebuild-percent
  31   DONE
  32
  33 * parse /proc/partitions and map major/minor into /dev/* names,
  34   and use that for default DEVICE list ????
  35
  36 * --detail --scan to read /proc/mdstat, and then iterate over these,
  37     but assume --brief.  --verbose can override
  38     check each subdevice to see if it is in conf_get_devs.
  39     Warn if not.
  40   DONE, but don't warn yet...
  41
  42 * Support multipath ... maybe...
  43   maybe DONE
  44
  45 * --follow to syslog
  46
  47 * --follow to move spares around DONE
  48
  49 * --follow to notice other events: DONE
  50      rebuild started
  51      spare activated
  52      spare removed
  53      spare added
  54
  55 ------------------------------------
  56 - --examine --scan scans all drives and build an mdadm.conf file DONE
  57
  58 - check superblock checksum in examine DONE
  59 - report "chunk" or "rounding" depending on raid level DONE
  60 - report "linear" instead of "-1" for raid level DONE
  61 - decode ayout depending on raid level DONE
  62 - --verbose and --force flags. DONE
  63
  64 - set md_minor, *_disks for Create  - DONE
  65 - for create raid5, how to choose between
  66    all working, but not insync
  67    one missing, one spare, insync  DONE (--force)
  68 - and for raid1 - some failed drives...  (missing)
  69
  70 - when RUN_ARRAY, make sure *_disks counts are right
  71
  72 - get --detail to extract extra stuff from superblock,
  73    like uuid  DONE
  74 - --detail --brief to give a config file line DONE
  75 - parse config file. DONE
  76 - test...
  77
  78 - when --assemble --scan, if an underlying device is an md device,
  79   then try to assemble that device first.
  80
  81
  82 - mdadm -S /dev/md0 /dev/md1 gives internal error FIXED
  83
  84 - mdadm --detail --scan print summary of what it can find?
  85
  86
  87 ---------
  88 Assemble doesn't add spares. - DONE
  89 Create to allow "missing" name for devices.
  90 Create to accept "--force" for do exactly what is requested
  91 - get Assemble to upgrade devices if force flag.
  92 ARRAY lines in config file to have super_minor=n
  93 ARRAY lines in config file to have device=pattern, and only accept
  94    those devices
  95    If UUID given, insist on that
  96    If not, but super_minor given, require all found with that minor
  97     to have same uuid
  98    If only device given, all valid supers on those devices must have
  99     same uuid
 100 allow /dev/mdX as first argument before any options
 101 Possible --dry-run option for create and assemble--force
 102
 103 Assemble to check that all devices mentioned in superblock
 104   are present.
 105
 106 New mode: --Monitor (or --Follow)
 107   Periodically check status of all arrays (listed in config file).
 108   Log every event and apparent cause - or differences
 109   Email and alert - or run a program - for important events
 110   Move spares around if necessary.
 111
 112   An Array line can have a spare-group= field that indicates that
 113    the array shares spares with other arrays with the same
 114    spare-group name.
 115    If an array has a failed and no spares, then check all other
 116      arrays in the spare group.  If one has no failures and a spare,
 117      then consider that spare.
 118     Choose the smallest considered spare that is large enough.
 119     If there is one, then hot-remove it from it's home, and
 120     hot-add it to the array in question.
 121
 122   --mail-to address
 123   --alert-handler program
 124
 125   Will also extract information from /proc/mdstat if present,
 126   and consider 20% marks in rebuild as events.
 127
 128   Events are:
 129      drive fails  - causes mail to be sent
 130      rebuild started
 131      spare activated
 132      spare removed
 133      spare added