Partition pruning regression
Your events table is range-partitioned by day across ~400 daily partitions. A dashboard query that filters `WHERE event_time >= now() - interval '1 day'` used to scan one or two partitions and return in 150ms; since a 'small refactor' deploy it's gotten slow and the DB is reading far more than before. `EXPLAIN ANALYZE` now shows the query scanning ALL 400 partitions instead of pruning to the recent ones. The data, indexes, and stats are unchanged. The refactor wrapped the time filter in a helper that computes the cutoff. Triage and mitigate.
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.