1. Execute a query against Redshift
2. Query immediately hits a Redshift queue
3. Cancel the query in DataGrip by pressing the 'Stop' button multiple times in quick succession. Console reports a java.io.EOFException
4. Reconnect the same console & get new PID via pg_backend_pid()
5. Check Redshift... the original query is still queued for execution
So, my question. What is happening if the Stop button is clicked a bunch of times. If clicked once, it seems to spin for a while and finish more 'gracefully' but kinda slow. If you click multiple times it seems to sever the connection rapidly, but pretty dirty. What is going on here - how come the 'dirty' shutdown doesn't kill the query on Redshift?