How to Remove Management Node From A Completely Down Host Using MySQL Cluster Manager (MCM)

(Doc ID 2115959.1)

Last updated on MAY 24, 2017

Applies to:

MySQL Cluster - Version 7.4 and later
Information in this document applies to any platform.

Symptoms

When a host containing the Management node (ndb_mgmd) is completely down, it is no longer possible to remove the missing process from the cluster.

Cluster can not be started or stopped as well.

mysql> show status -r mycluster;
+--------+----------+--------+---------+-----------+---------+
| NodeId | Process  | Host   | Status  | Nodegroup | Package |
+--------+----------+--------+---------+-----------+---------+
| 49     | ndb_mgmd | host01 | stopped |           | 7.4.6   |
| 50     | ndb_mgmd | host02 | unknown |           | 7.4.6   |
| 1      | ndbd     | host03 | added   | 0         | 7.4.6   |
| 2      | ndbd     | host04 | unknown | 0         | 7.4.6   |
| 51     | mysqld   | host01 | stopped |           | 7.4.6   |
| 52     | mysqld   | host02 | unknown |           | 7.4.6   |
+--------+----------+--------+---------+-----------+---------+
6 rows in set (1.20 sec)
mysql> stop cluster mycluster;
ERROR 1002 (00MGR): Agent on host host02:18620 is unavailable
mysql> remove process 50 mycluster;
ERROR 7027 (00MGR): Process ndb_mgmd 50 must be stopped before removal
mysql> stop process 50 mycluster;
ERROR 1002 (00MGR): Agent on host host02:18620 is unavailable

Cause

Sign In with your My Oracle Support account

Don't have a My Oracle Support account? Click to get started

My Oracle Support provides customers with access to over a
Million Knowledge Articles and hundreds of Community platforms