Implementing Prometheus Collector on Amazon EKS

Amazon was excited to present a new feature today: Prometheus measurements may now be automatically and agentlessly discovered and collected from Amazon Elastic Kubernetes Service (Amazon EKS) with the help of Amazon Managed Service for Prometheus collector. A scraper that finds and gathers metrics from Amazon EKS apps and infrastructure is part of the Amazon Managed Service for Prometheus collector; it eliminates the need to operate any collectors in-cluster.

With Amazon Managed Service for Prometheus, this new feature offers completely managed monitoring and alerting that is compatible with Prometheus. The collector’s ability to be completely managed, automatically scaled, and sized to match your use case is one of its main advantages. This implies that in order for collectors to gather the accessible metrics, no computation is required on your part. This aids in the cost-effective optimization of metric gathering for monitoring your EKS-based applications and infrastructure.

With this release, AWS managed collection, a completely managed and agentless collector, and customer managed collection are now the two main Prometheus metrics collection methods supported by Amazon Managed Service for Prometheus.

Using Amazon Managed Service for Prometheus Collector: An Overview

Now that Amazon have this additional option, let’s see how to use AWS managed collectors to import metrics into an Amazon Managed Service for Prometheus workspace. After that, Amazon will assess the gathered metrics using Amazon Managed Grafana.

By choosing Send Prometheus metrics to Amazon Managed Service for Prometheus, you may now use AWS managed collector when you build a new EKS cluster using the Amazon EKS console. You have the option to select your current Amazon Managed Service for Prometheus workspace or create a new one in the Destination section. The getting started guide will teach you more about setting up a workspace.

You may then use the editor to define your scraper settings or upload an existing configuration if you want. How you want the scraper to find and gather metrics is controlled by the scraper configuration. Please visit the Prometheus Configuration page to view the settings that you can configure.

The list of scrapers that are currently operating in your EKS cluster may be viewed by selecting the Observability tab on your cluster page when the formation process is complete.

Employing AWS APIs and CLI

In addition to the Amazon EKS dashboard, you can install an AWS managed collection via the AWS Command Line Interface (AWS CLI) or APIs. This method is helpful if you wish to change the configuration of the current collector or add an AWS managed collector to an already-existing EKS cluster.

The majority of the parameter values, including your Amazon Managed Service for Prometheus workspace ARN and your EKS cluster ARN, are available in the corresponding AWS console. In addition, the scraper configuration, described as configurationBlob, must be defined.

Before passing the API call, you must encode the configuration file into base64 encoding after defining the scraper configuration. To encode sample-configuration.yml into base64 and copy it to the clipboard on my Linux development workstation, you can use the following command.

$ base64 sample-configuration.yml

Currently accessible

All AWS customers in all AWS Regions where Amazon Managed Service for Prometheus is supported can now access the collector capabilities of Amazon Managed Service for Prometheus.