Admin only - Server stats - duplicate SQL queries

Found while testing #1341 - but it needs a bit of thought

When you run a large query (eg export an analysis node as CSV) Postgres parallelises things, so you end up seeing the same query 4 times

I think we should instead perhaps group by distinct SQL query, and then make the pid an aggregated query (joined by string) and then kill them as a group if you click them?

I think if you kill 1 of a parallelized query it will kill them all, so probably doesn't matter

You could also just show 1 PID, and maybe (x4 threads) or something