- NDB table objects (logfile groups, tablespaces, and tables)
  submitted for scrutiny by the metadata change monitor thread are
  checked for mismatch and synchronized by the binlog thread.
  The objects detected by the change monitor are stored in a
  queue and are synchronized in the order of their arrival
- The synchronization is done at the rate of 1 object per epoch
- Usage of this automatic synchronization is driven by the detection
  logic which is in turn controlled by ndb_metadata_check and
  ndb_metadata_check_interval variables
- The Ndb_metadata_synched_count variable can be used to track
  the number of objects synchronized. Further information about the
  synchronization can be found through log messages written by the
  binlog thread
- If synchronization of an object should fail with a non-temporary
  error, the object is then placed in a blacklist. Subsequent
  cycles of detection by the change monitor thread will then
  ignore all such blacklisted objects. This ensures that the binlog
  thread doesn't waste resources in trying to sync the same object
  where it shall run into the same error repeatedly. Such objects
  have to be synchronized manually by the user through table
  discovery or in extreme cases by restarting the MySQL server to
  trigger full schema synchronization. The entries in the blacklist
  are checked by the change monitor thread at the beginning of every
  detection cycle. Should the mismatch between NDB Dictionary
  and the DD not exist anymore, the object is removed from the
  blacklist and is once again eligible for automatic synchronization
  should the need arise
- The Ndb_metadata_blacklist_size variable can be used to track
  the number of objects in the blacklist. Further information about
  objects in the blacklist can be found through log messages written
  by the binlog and change monitor threads

Change-Id: Ib2dc48cde67888af17a969a75556cb6b2aa3e154