Messaging on AWS in twenty-nine slides: the course project's upload pipeline, synchronous and asynchronous calls, a queue absorbing a spike and workers scaling on the backlog; SQS queues, the consumer loop, the visibility timeout, long polling, dead-letter queues, standard and FIFO queues; SNS topics, filter policies, fan-out and the envelope a queue receives; EventBridge buses, rules, event patterns and EventBridge Scheduler; Kinesis Data Streams beside SQS, Amazon MQ for lift-and-shift, Step Functions for orchestration, and how to choose among them. Every section ends with five exam questions.
The slides
The tool is twenty-nine slides: an introduction, then five sections, listed on its left. Every section ends with five
exam questions, each with its answer and the tempting wrong answer.
- 0. Introduction
- Messaging between the parts of an application: the project's upload pipeline, built in the lab.
- Synchronous and asynchronous calls: a web server waiting for a service, and one leaving work in a queue.
- 1. Decoupling
- A queue absorbs a spike: the backlog with two workers, and with workers that scale.
- Scaling the consumers on the backlog: messages visible divided by instances in service, by metric math.
- 2. Amazon SQS
- Amazon SQS queues and an SQS consumer: send, receive, process, delete.
- The visibility timeout: a worker that finishes, and one that crashes.
- Short and long polling, dead-letter queues, standard and FIFO queues and SQS figures.
- 3. Amazon SNS and fan-out
- Amazon SNS topics and subscription filter policies.
- The fan-out pattern, and the message SQS receives from SNS.
- 4. Amazon EventBridge
- Amazon EventBridge: buses, rules and targets.
- An event and the rule that matches it, EventBridge Scheduler, and EventBridge beside SNS.
- 5. Streams, brokers and workflows
- Amazon Kinesis Data Streams and Kinesis Data Streams beside SQS.
- Amazon MQ, AWS Step Functions, and choosing a messaging service.
The buttons on the left and right edges move between slides. So do the arrow keys and a swipe.
The lab
Messaging Lab: Uploads, a Topic and a Worker Fleet practices it. The older
article
About Data Streams introduces streams of data in general.
The SQS figures (1 MiB messages, 14 days of retention, a 12-hour visibility timeout, 20 seconds of long polling,
FIFO's 300 calls a second) are the SQS Developer Guide's quotas, read on 2026-11-16, as are Kinesis's per-shard limits
and 10 MiB records. The backlog drawing is a simulation: 10 messages a minute per worker. The account ID and the
bucket are placeholders.
Sources