Hi
We have Exadata Full RAC which has 8 nodes --
DB Version is 12.1.0.2.0 and below are the parallel parameters.
SQL> sho parameter parallel
PARAMETER_NAME TYPE VALUE
------------------------------------------------------------ ----------- -------------------------------------------------------------
fast_start_parallel_rollback string LOW
parallel_adaptive_multi_user boolean FALSE
parallel_automatic_tuning boolean FALSE
parallel_degree_level integer 100
parallel_degree_limit string 16
parallel_degree_policy string MANUAL
parallel_execution_message_size integer 16384
parallel_force_local boolean FALSE
parallel_instance_group string
parallel_io_cap_enabled boolean FALSE
parallel_max_servers integer 204
parallel_min_percent integer 0
parallel_min_servers integer 96
parallel_min_time_threshold string AUTO
parallel_server boolean TRUE
parallel_server_instances integer 8
parallel_servers_target integer 204
parallel_threads_per_cpu integer 1
recovery_parallelism integer 0
SQL> sho parameter cpu
PARAMETER_NAME TYPE VALUE
------------------------------------------------------------ ----------- ---------------
cpu_count integer 48
parallel_threads_per_cpu integer 1
resource_manager_cpu_allocation integer 48
I know the formulae
parallel_threads_per_cpu*cpu_count*concurrent_users*5 and concurrent_users is 4 because we have ASMM and PGA_AGGREGATE_LIMIT IS SET
so the above formulae equals to 960 - can we increase it to 960 ? we are also working on DBRM project that limits the DOP based on consumer group's and we are limiting the number of nodes each application uses from 8 to 3 we might run out of slaves with the current parallel_max_servers value with the 3 nodes and we want to see how much we can increase from 204 value-