2013-10-12.

/usr/sbin/raid-check
WARNING: mismatch_cnt is not 0 on /dev/md0

Looking this up on 
  http://www.fractalizer.ru/frpost_350/warning-mismatch_cnt-is-not-0-on-devmd0/

we see the actual value can be found from
  cat /sys/block/md0/md/mismatch_cnt
(which showed 8 blocks mismatched), then
  echo repair >/sys/block/md0/md/sync_action
  watch cat /proc/mdstat	
and 
  echo check >/sys/block/md0/md/sync_action
  watch cat /proc/mdstat
and finally
  cat /sys/block/md0/md/mismatch_cnt

So: new things: "repair" and "check" into the sync_actions.
The mismatch is claimed to be no trouble on RAID1[0], but 
worrying on RAID5 (our case!).  Haven't fully understood yet
what it implies.  

The same source stated (as an update) that the array with
that problem had later had a disk trouble (smart report of 
relocated sectors) and then problems with applications using
the data: so, recommended (of course), 
  smartctl --test long /dev/sda