So I had an issue where ndb_restore was utterly failing and made it so it was impossible to restore ndbcluster tables from a backup. Trying to use it was flooding us with all sorts of sporadic errors…

[NdbApi] ERROR — Sending TCROLLBACKREQ with Bad flag
—————
theNoOfSentTransactions = 1 theListState = 0 theTransArrayIndex = 5
—————
Temporary error: 410: REDO log files overloaded (decrease TimeBetweenLocalCheckpoints or increase NoOfFragmentLogFiles)
—————
Temporary error: 266: Time-out in NDB, probably caused by deadlock
—————

Finally was able to work it out, but man, can I just say that ndb_restore really could use an overhaul to make it a little more friendly? Okay, go! 🙂