Hello,
1/ access scheduler_task_executions
by index with device_id = 97
seems ok
2/
I don't understand why
joining
scheduler_task_executions.id=scheduler_operation_executions.task_execution_id
is done using a parallel hash join
when a nested loop would be better (regarding the number of rows involved)
maybe because index on scheduler_operation_executions.task_execution_id
"index_task_execution_id_desc" btree (task_execution_id DESC NULLS LAST)
is not usable or bloated or because of DESC NULLS LAST ?
3/ join with results.operation_execution_id
by index
seems OK
Regards
PAscal
--
Sent from: https://www.postgresql-archive.org/PostgreSQL-performance-f2050081.html