ERROR 1005 (HY000) at line **: Can't create table '**' (errno: 136)

1.When you get this error immediately run "SHOW WARNINGS" to see the storage engine level error codes. (mysql>) 2.Display: Got error 904 'Out of fragment records (increase MaxNoOfOrderedIndexes)' from NDB 3.$ ndb_config --type ndbd -q Id,MaxNoOfOrderedIndexes,MaxNoOfUniqueHashIndexes -r '\n' 4.$ ndb_show_tables | grep -c OrderedIndex 5. ndb_mgm> ALL REPORT MEMORY This clearly shows that your settings in the config.ini are not being applied to the cluster. Restart ndb_mgmd with --initial then do a rolling restart of all ndbd nodes to ensure that they use the increased values. Every table will have at least 1 Hash index used by the primary key. If you do not define a primary key for a table, ndb will define a hidden one for you. so setting MaxNoOfUniqueHashIndexes=1250 much lower than MaxNoOfTables=4096 is likely to run you into this "No more room in index file" message as soon as or before you have 1250 tables. I had exact same problem and increasing MaxNoOfOrderedIndexes to 1024 did indeed do the trick