If upgrading from 1.x:
* Some people have had problems getting statistics in previous versions.
  This is/was because of a number of problems which should be fixed in
  version 1.3 (this version).

  IF you've had problems before, got to 'Graph Management', search for
  'Any' host and with the search string 'bind9'. Make sure you delete
  (!) all entries found. Also make sure to delete all RRD files in the
  '<path_rra>' directory - the command 'rm <path_rra>/*_bind9_*.rrd'
  should do that. Be sure to replace '<path_rra>' with the full path
  to your RRD files (you can find that in the 'Configuration->Settings->Paths'
  section...

If upgrading from 1.1:
* The script is now a pass-through PERSISTS instead of a
  simple pass-through. This because I noticed that SNMPd
  calls the script ONCE FOR EACH OID requested. For a walk,
  that's a lot of calls!
  Every time the script is/was called, it dumped the Bind9
  statistics, THEN processed and returned the information...
  Lead to a LOT of dumping of statistics!

  Instead use a 'pass_persist' in the snmpd.conf file, so
  that the script is called only ONCE. The SNMPd will kill
  the script if not used 'for a couple of minutes'. Don't
  know how long, but IF the script is keept running for
  more than five minutes, it will issue a reload/redump
  of the Bind9 statistics by itself...
* The script now removes the information dump file (dns-stats.log)
  after each dump. This to keep the file in reasonable sizes.
  The file grows with a couple of kilobytes _every five minutes_
  and over time (not MUCH time!), it's over 2Gb in size!