This website uses cookies to ensure you get the best experience.
Learn more.
A Kinesis Data Streams producer writes events for thousands of devices. Every record uses the partition key telemetry, and writes are throttled even though other shards have capacity. Events must remain ordered per device, not globally. Which change best addresses the bottleneck?
Published
- Keep the constant key and add consumer applications until writes stop throttling.
- Keep the constant key and increase only the stream retention period.
- Use a well-distributed device identifier as the partition key and size stream capacity for the workload.
- Assign a different random key to every event and rely on global ordering across shards.
Comments — 0
Sign in to comment