API p99 latency spike

SEV2resolved

Postgres connection pool exhaustion caused tail latencies to spike. Resolved by rolling pod restart + pool size increase.

ID: INC-2026-048Commander: op.jen

Opened

1w ago

Resolved

1w ago

Duration

42m

Severity

SEV2

Affected

Reverse-chronological updates

op.jen1w ago

Resolved — pool bumped from 20 → 60. Post-mortem scheduled.

op.jen1w ago

Identified pool exhaustion in checkout handler.

pagerduty1w ago

SEV2 opened — api-p99 > 500ms.