Partition pruning regression
Postgres with a `metrics` table range-partitioned by `event_date` into daily partitions (2 years of data). After a deploy that changed an ORM, a dashboard query that filters `WHERE event_date >= now() - interval '7 days'` went from 80 ms to 12 s. `EXPLAIN` shows it now scanning ALL ~730 partitions instead of the last 7. The query text looks logically the same. CPU is fine; it's pure work amplification. Triage, fix now, and prevent.
What a strong answer looks like
Stop the bleeding first (mitigate), then form hypotheses from real signals. Separate root cause from symptom, communicate status as you go, and close with what prevents a repeat.
0:00 of about 35 min
Which questions mattered is sealed until you submit. Telling you now would just be handing over the edge cases.
Run or narrate your approach, then ask the coach.