I want to see what is happening on the Oracle server right now... or few hours before;
I need to find which SQL statements were responsible for the recent spike of server activity and response time degradation, see their SQL text, what they were waiting for, etc;
I want to know what this long running query is doing now;
I want to get a feeling for how the server behaves in response to your SQL changes;
all sql's ( during runtime ) taking >6 s are logged into v$session_longops
run statspack.snapshots are desired intervals using dbms_job or crond
you will get top sqls ( exec time / parse time ) in any given timeframe.
unfortunately, there's no quickfix solution to that.
Setting up statspack and scheduling snapshot during your peak workloads and interpreting report requires knowledge and exp.