RAT: Initialize Replay for 19c Database Fails with Error: ORA-20223: Found ongoing replay in another container (Doc ID 2847277.1)

Last updated on FEBRUARY 11, 2022

Applies to:

Symptoms

Oracle Database 19.12.0.0.0, Real Application Testing

Initialize Replay fails with Error " ORA-20223: Found ongoing replay in another container"

SQL> exec dbms_workload_replay.initialize_replay(replay_name => 'Replay1', replay_dir => 'RAT_DIR');
BEGIN dbms_workload_replay.initialize_replay(replay_name => 'Replay1', replay_dir => 'RAT_DIR'); END;

*
ERROR at line 1:
ORA-20223: Found ongoing replay in another container
ORA-06512: at "SYS.DBMS_WORKLOAD_REPLAY_I", line 3687
ORA-06512: at "SYS.DBMS_WORKLOAD_REPLAY_I", line 3708
ORA-06512: at "SYS.DBMS_WORKLOAD_REPLAY", line 42
ORA-06512: at line 1

Changes

The sequence of steps that causes the error.

1) Database Replay fails on a PDB.
2) The DB is flashbacked and hence wrr$_replays table is truncated and dba_workload_replays shows no replay to be running.
3) When trying another reply in a different container it fails as in step #2 the replay stage is still stuck in memory.

ERROR at line 1:
ORA-20223: Found ongoing replay in another container

4) Ideally a DB Replay should be canceled after failure. This will clean up the replay in memory as well. If DB is flashbacked we can’t cancel_replay since the record is no longer there and a new replay cannot be started as the previous replay is still active in the memory.

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.