The Join That Made 900 Million RowsWarehouse · easy · ~10 minThe revenue dashboard's main query used to run in 40 seconds. Since yesterday it takes 28 minutes and the BI tool times out. Nothing about the SQL changed — it joins orders to customer_dim and aggregates.You pull the query profile. Work out what actually happened, and what fixes it — the answer is in the row counts.Evidence

Query Profile

Profile Detail

Tables

Checkpoints

The join takes 2.1M + 1.84M rows in and produces 918M rows out. What does that tell you?

What made the previously-fine join start fanning out?

Which fix addresses the root cause?

Why would resizing the warehouse to XL have been the wrong call, even though it might reduce the runtime? Answer interview-style.