We have a testing scenario where we see large number of active sessions in pg_stat_activity ( >100) and the wait_event and wait_event_type columns for these active sessions in pg_stat_activity are all null.
Is it safe to assume that if a session is active in pg_stat_activity and has NULL a wait_event and a NULL wait_event_type that the session is on the CPU?
I’m asking because in our testing we are observing a significant number of active sessions in pg_stat_activity with NULL wait_events and NULL wait_event_types and when we look at the corresponding OS process for these active sessions we observe that they are in a sleep state.
Is there anything we can do to understand what is happening when sessions are active with null wait_events in pg_stat_activity but the corresponding OS processes are sleeping?