19.12 Linux - PGA MEMORY VERY LARGE DUE TO Q PROCESSES - BACKGROUND Streams EMON Regular Slave

@Mohamed Houri - I have a 19.12 CDB with 30 PDBs on Linux and Windows and I am "not" using Advanced Queueing, nor Streams and the issue with Q Processes Memory Leak still happens.

Someone else was experiencing the same issue in their production environment.

https://community.oracle.com/mosc/discussion/comment/16841951#Comment_16841951

here's the steps provided from Oracle:

Check if AQ_SRVNTFN_TABLE_Q_1 dequeue is enabled or disabled:

select owner, name, queue_table, ENQUEUE_ENABLED, DEQUEUE_ENABLED from dba_queues where name = 'AQ_SRVNTFN_TABLE_Q_1' and owner='SYS';

If the dequeue is disabled, re-enable it with the following:

EXECUTE DBMS_AQADM.START_QUEUE (queue_name => 'SYS.AQ_SRVNTFN_TABLE_Q_1');

check again to confirm the change. This stopped the Q processes from growing in their case.

-----------------------------------------------------------------------------------------------

I applied the workaround described above on my databases 19.12 Linux and 19.12 Windows but the Q processes memory leak still happens and the Q processes continue to take more and more memory and do not release the memory.  : (

Update. I opened the Oracle Support SR on Aug/03/2021, more than 30 days ago, and I had 2 Spin-Off SRs, each time I had to provide more and more logs, and at each Spin-Off SR I had to engage with another support engineer and explain the problem all over again, the first parent SR owner was not following up directly with the Spin-Off SR owners to coordinate and explain the issue, hence we ended up wasting a lot of time, only when the first parent SR owner aggreged to file the bug that another Oracle support engineer took ownership of the parent SR and with his help we finally found a workaround that worked. I wish that person had been involved in the SR from the beginning, we would probably had encountered the workaround a lot sooner.