Event ingestion queue latency under increased throughput
warningResource ContentionUpdated Mar 8, 2026(via Exa)
How to detect:
Increased latency in producing and consuming from the ingestion event queue when event throughput exceeds queueing system capacity, resulting in degraded pipeline throughput and accumulating processing lag of 30-40 minutes. No data loss occurs during degraded performance.
Recommended action:
Monitor ingestion queue latency and lag metrics for increasing trends. Compare current event throughput against provisioned capacity. Upgrade compute infrastructure for the queueing system to handle increased event load. After throughput restoration, expect 3-4 hours recovery time to process backlog from 30-40 minute lag.