Art of Codeartofcode
Lesson 1 of 18

22. Message Queues

1
The picture to keep in your head
2
22.1 What are producers and consumers?
3
22.2 Why go async?
4
22.3 What are the delivery guarantees?
5
22.4 How do acks and visibility timeouts work?
6
22.5 What are retries and dead-letter queues?
7
22.6 What is an idempotent consumer?
8
22.7 How does ordering work?
9
22.8 What is backpressure?
10
22.9 How do you scale consumers?
11
22.10 Log (Kafka) vs queue (RabbitMQ, SQS): what's the difference?
12
22.11 How do delayed and scheduled jobs work?
13
Memory card and where the analogy breaks
14
Practice: five levels
15
Interview questions
16
Solutions
17
Sources
18
Code lab
22. Message Queues

The picture to keep in your head

3 min read+10 XP

Every question below is answered twice: first at the bank, so you can picture it, then in the system, so you know exactly how it works. Each answer ends with a memory code to lock it in.

The picture to keep in your head

Walk into a busy bank. You don't stand at a counter waiting for a clerk to be free. You press a button, take a token, number 47, and go sit down. Your token goes on the board. When a clerk at any counter is free, the board calls the next number. When your work is done, the clerk clears your token. If the bank gets busy, the manager opens more counters. And if one customer's problem can't be solved at the counter, after a few tries they're sent to the manager's desk, so the queue keeps moving.

You didn't have to wait at the counter. The clerks didn't have to rush. Nobody was forgotten, even when the bank was packed.

That's a message queue: one part of a system leaves a job, another part picks it up when it's free, and the queue holds everything in between.

Our running example: the grocery app's checkout. When a customer places an order, three slower jobs need to happen: send the SMS confirmation, print the packing slip at the store, and add loyalty points. At a Big Sale, orders peak at around 1,200 a second (an assumption for practice).

At the bankIn the system
Taking a token and sitting downThe producer sends a message and moves on
The token boardThe queue (or log)
Clerks at the countersConsumers (workers)
"Done, next!" and the token is clearedAcknowledging (acking) a message
A clerk who walks off mid-job: the token is called againVisibility timeout and redelivery
The manager's desk for the stuck casesThe dead-letter queue (DLQ)
A token called twice: don't pay out twiceIdempotent consumers
Opening more counters at rush hourScaling consumers
"Come back at 3 pm"Delayed and scheduled jobs

the bank's token system mapped to producers, queues, consumers, acks and the dead-letter queue.

Figure: take a token and sit down; the work gets done, just not while you stand at the counter.