In a Linux-based NDB-cluster consisting of 7 API nodes, 1 Management node and 4 NDB nodes, I am trying to drop a table.

When I try to drop the table using MySQL Client, I receive the following error:

mysql> drop table table_name;
ERROR 1051 (42S02): Unknown table 'database_name.table_name'

I should be able to drop tables in an NDB Cluster by using the MySQL Client, but after a short search on Google I found I could also use the ndb_drop_table tool.

I then got the following results:

./ndb_drop_table --ndb-nodeid=13 --ndb-mgmd-host=xx.xx.xx.xx:1186 table_name -d database_name
Dropping table table_name...
4243: Index not found

NDBT_ProgramExit: 1 - Failed

Searching for this error, I got exactly one (unanswered) result: http://lists.mysql.com/cluster/1608

As a last resort I tried to delete the entire database, but this command internally drops all tables first, so I got the same results:

mysql> drop database database_name;
ERROR 1051 (42S02): Unknown table 'database_name.table_name'

Please advise on how to delete this table.


After a few days of searching, I found out the problem was caused by a MySQL version mismatch. We recently created tho new API nodes and installed the latest version of MySQL (5.6.11-ndb-7.3.2).

The results described in the original question occurred when trying to delete the table (stored on NDB nodes running MySQL 5.5.27-ndb-7.2.8) from one of these new servers.

The table could be deleted without trouble from an older host, also running MySQL 5.5.27-ndb-7.2.8. After downgrading the new API nodes to MySQL 5.5.31-ndb-7.2.13, these hosts no longer caused any trouble.

I hope this answer might help someone in the future.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.