Queues with SQS · Question 12 of 12
An Auto Scaling group of EC2 instances processes messages from an SQS queue. Which scaling approaches make sense?
2 answers are correct. Hard Created with AI
Show answer
Answer
- A Step scaling based on ApproximateNumberOfMessagesVisible
- C Target tracking on a custom "backlog per instance" metric
Explanation
What matters is how many messages are waiting. AWS recommends dividing the backlog by the number of instances and setting the target to the messages one instance can handle within the acceptable latency. A load balancer is not involved in this flow at all.