Hi,
I have an Oracle11gR2 (11.2.0.4) database for which parallel_degree_policy = limited.
I also have parallel_max_servers = 36, and parallel_degree_limit = 36.
At that stage, I would expect that with that parameter auto DOP be enabled and parallel queing being not enabled (that is what I can read from the Oracle documentation).
Through the Cloud Control, I have sessions that are logged as a specific user, and all use a parallel degree of 12. This parallel degree is enforced by a trigger for ddl/dm/query, and for that specific user. If I count the number of parallel processes that are displayed (this is 12 for each session) I get a number that is higher than 36 (the value of parallel_max_servers).
Can someone explain me why it is so, and not limited to 36?
In one session, logged to a different user, I have parallel_degree_policy = auto, set at session level.
the query: select * from v$io_calibration_status returns a status "NOT AVAILABLE" which means that I/O calibration has not be done for that database. This is a necessary condition for using auto DOP. Yet, I observe, through the Cloud Control that that session is queued for parallel process.
Can someone explain me why is it so? I would expect that the query for this session be not queued.
Thanks by advance for any tips and explanation.
Groet, Laurent