Amazon’s Application Load Balancer is a popular infrastructure component found in many application architectures in AWS.
Because load balancers route traffic to application services, they make an ideal place to set SLOs. You might think to set SLOs on the health of the load balancer itself, especially if there is a known failure mode related to the load balancer – such as, say, traffic that spikes faster than an ALB can adaptively scale.
SLOs on ALB health are certainly a valid use case. However, a more typical use of SLOs on an ALB is to set SLOs on the behavior of the services behind it, using the ALB as a convenient pulse point to measure the performance and stability of the services themselves. That’s what we’ll do in this module.