blackhole: read/delete/write counts

- health daemon: niet rebooten binnen 't uur
- hd: bericht sturen waarom we rebooten

- replacing & delete old hash & new hash
  - get current hash H1 for this sector number
  - get number of references to this hash C1
  - delete hash H1/data pair
  - delete usage counter C1
  - get number of references to the new hash C2
  - store hash H2 /data pair
  - set new usage count C2
  = 3 x read, 2 x delete, 3 x write

- replacing & delete old hash & existing hash
  - get current hash H1 for this sector number
  - get number of references to this hash C1
  - delete hash H1/data pair
  - delete usage counter C1
  - get number of references to the new hash C2
  - set new usage count C2
  = 3 x read, 2 x delete, 1 x write

- replacing & keep old hash & new hash
  - get current hash H1 for this sector number
  - get number of references to this hash C1
  - set new usage count C1
  - get number of references to the new hash C2
  - store hash H2 /data pair
  - set new usage count C2
  = 3 x read, 3 x write

- replacing & keep old hash & existing hash
  - get current hash H1 for this sector number
  - get number of references to this hash C1
  - set new usage count C1
  - get number of references to the new hash C2
  - store hash H2 /data pair
  - set new usage count C2
  - get number of references to the new hash C2
  - set new usage count C2
  = 4 x read, 4 x write