This is an arbitrary choice for demo purposes, but you could use any type of machine and metrics. We will intentionally use a very small machine for the demo, and we will scale the Pool based on CPU load. The Load Balancer will scale accordingly, distributing traffic to all underlying instances in the Instance Pool.Īs a first step, we need a running Instance Pool to run our web application. In this article, we will demonstrate how those building blocks, such as the Network Load Balancer and Instance Pools, can be used together to achieve an automatically scaling and highly available deployment.įor the purposes of this example, each member of the Instance Pool will run a webserver to serve web content and a Prometheus node exporter to gather metrics.Ī central Prometheus instance will collect the metrics, pass them to Grafana, and use a simple utility to scale the Instance Pool on demand based on the selected metrics. While each of the building blocks on its own is not enough to solve all problems you may be facing, they can and should be used together to build up your own infrastructure efficiently. In the last 12 months, we released several fundamental services that greatly simplify the deployment of larger architectures.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |