After executing an upgrade exercise in a dataguard environment (19c), the broker was still showing the status: “ROLLING DATABASE MAINTENANCE IN PROGRESS”

DGMGRL> show configuration;

Configuration - my_dg_config

  Protection Mode: MaxAvailability
  Members:
  orcl_dga - Primary database

  Members Not Receiving Redo:
  orcl_dgb - Physical standby database 
  orcl_dgc - Physical standby database 

Fast-Start Failover:  Disabled

Configuration Status:
    ROLLING DATABASE MAINTENANCE IN PROGRESS

DGMGRL> exit

This is due to a bug: Bug 26921039 – INCORRECT DGMGRL SHOW CONFIGURATION STATUS AFTER POST ROLLING UPGRADE SWITCHOVER

Solution is simple:

Execute the following statement on PRIMARY database and bounce it:

DBMS_ROLLING.DESTROY_PLAN ();

SQL> exec dbms_rolling.destroy_plan

PL/SQL procedure successfully completed.

SQL> exit

SQL> startup force;
ORACLE instance started.

Total System Global Area 1610609200 bytes
Fixed Size                  8897072 bytes
Variable Size             436207616 bytes
Database Buffers         1157627904 bytes
Redo Buffers                7876608 bytes
Database mounted.
Database opened.
SQL> exit

Let’s check again the broker’s status:

[email protected] ~]$ dgmgrl /

Configuration - my_dg_config

  Protection Mode: MaxAvailability
  Members:
  orcl_dga - Primary database
    orcl_dgb - Physical standby database 
    orcl_dgc - Physical standby database 

Fast-Start Failover:  Disabled

Configuration Status:
SUCCESS   (status updated 0 seconds ago)

DGMGRL> 

Now it’s looking good 🙂