Discrepancy between mysqlsh and performance_schema with 8.0.11

[Grant Gavares]

I have a 3 node cluster running in docker containers using mysql 8.0.11. I can successfully bootstrap the cluster and get all three nodes to join the master after running:

dba.creatCluster('my_cluster', {'adoptFromGR': True})

All the nodes join and, after a brief period syncing with the master, the cluster becomes stable.

Next, I stop the master node gracefully. Once I've done this, any attempt to query the cluster status using mysqlsh from any of the remaining nodes gives error:

Dba.getCluster: This function is not available through a session to an instance belonging to an unmanaged replication group (RuntimeError)

If I query 'performance_schema.replication_group_members', it appears to indicate that the cluster is stable and that the primary has migrated to a new cluster node:



mysql> select member_host, member_state, member_role, member_version from performance_schema.replication_group_members;
+---------------+--------------+-------------+----------------+
| member_host | member_state | member_role | member_version |
+---------------+--------------+-------------+----------------+
| mysql-4.local | ONLINE | PRIMARY | 8.0.11 |
| mysql-6.local | ONLINE | SECONDARY | 8.0.11 |
| mysql-2.local | ONLINE | SECONDARY | 8.0.11 |
| mysql-3.local | ONLINE | SECONDARY | 8.0.11 |
+---------------+--------------+-------------+----------------+

In the example above, I've killed mysql-1.local which was the initial master node.



Edited 1 time(s). Last edit at 05/14/2018 10:48PM by Grant Gavares.