View a markdown version of this page

Data freshness higher than configured - Amazon Managed Streaming for Apache Kafka

Data freshness higher than configured

  • Symptom: DataFreshness is significantly higher than the configured interval.

  • Causes: A high number of source or Iceberg table partitions; large table metadata or many accumulated snapshots; (less commonly) transient service issues or low source throughput.

  • Resolution: Verify your partitioning is appropriate for the data and avoid excessive partition cardinality. Check whether the table has grown large metadata or many snapshots and, if so, enable S3 Tables maintenance (compaction and snapshot expiration) to keep the table performant. For low-throughput topics, increasing the configured data freshness gives the Channel more time to accumulate data for efficient delivery.