Question
00:05 on the 1st of the month: all INSERTs into your time-partitioned `events` table (Postgres, declarative range partitions by month) start failing with "no partition of relation events found for row." Ingestion is down — every write for a timestamp in the new month is rejected, and the upstream queue is backing up fast. Separately, over the past few weeks a few read queries that filter on `event_time` have quietly gotten slower. Context: partitions were always created by hand / a one-off script when the table was set up a year ago; nobody automated future partition creation, and the cron that was supposed to pre-create next month's partition was removed in a 'cleanup' last sprint. Triage and act.
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.