Question 1

What's the difference between a streaming service and a message queue?

Accepted Answer

A message queue is usually designed for task distribution: messages are processed once and then removed, often with a single main consumer per message. A streaming service is designed for continuous event streams where multiple consumers can read the same events, often with replay (reading past events) and ordering within partitions. Streaming services are commonly used for analytics, monitoring, and event-driven architectures where you want fan-out and the ability to reprocess data.

Question 2

When should I use a streaming service?

Accepted Answer

Use a streaming service when you have continuous data arriving (clickstream, IoT telemetry, logs, transactions) and you need near-real-time processing, multiple downstream consumers, or the ability to replay events for debugging and reprocessing. If you only need simple background job processing with one consumer per task, a queue may be simpler.

Question 3

How much does a streaming service cost?

Accepted Answer

Costs typically depend on throughput (ingress/egress), number of partitions/shards (reserved capacity), retention duration, and any enhanced features (e.g., extended retention, cross-region replication). You also pay for downstream processing (stream processing jobs), storage for long-term retention (data lake/warehouse), and network egress if consumers are in different regions or clouds.

Streaming Service

Definition

Real-World Example

Related Terms

Cloud Provider Equivalencies

Explore More Cloud Computing Terms