Question 1

What's the difference between Amazon Kinesis and Amazon SQS?

Accepted Answer

Kinesis is built for high-throughput data streaming where many consumers may read the same stream and you often care about ordering within a shard and replaying recent data. SQS is a message queue designed to decouple systems with point-to-point delivery (messages are typically processed once by a consumer) and is not meant for many consumers replaying the same event history.

Question 2

When should I use Amazon Kinesis?

Accepted Answer

Use Kinesis when you need to ingest and process continuous event data in near real time—like clickstreams, app logs, IoT telemetry, fraud signals, or operational metrics—especially when volume is high and you want multiple downstream consumers (analytics, alerting, storage) to read the same stream.

Question 3

How much does Amazon Kinesis cost?

Accepted Answer

Cost depends on which Kinesis capability you use and how much data you ingest and process. For Kinesis Data Streams, pricing is driven by stream capacity (shards or on-demand throughput), PUT payload units, and optional enhanced fan-out. For Kinesis Data Firehose, pricing is typically based on data volume ingested and optional data transformation (for example, Lambda). For Kinesis Data Analytics for Apache Flink, pricing is based on the compute resources your application uses (and any additional AWS services it integrates with). Always estimate using the AWS Pricing Calculator with your expected records/sec, record size, retention, and number of consumers.

Kinesis

Definition

Real-World Example

Related Terms

Cloud Provider Equivalencies

Explore More Cloud Computing Terms