As a relatively new AWS cloud customer embarking on our DevOps journey, we're eager to embrace Continuous Improvement practices. One of our recent endeavors involved enabling CloudFormation, along with enhanced drift detection, to ensure the reliability and availability of our services. Across our infrastructure, we maintain three distinct environments, each equipped with dedicated monitoring stacks.
Our goal is to introduce Thanos Sidecar into these three existing Prometheus/Grafana systems to create a centralized dashboard. This approach aims to minimize latency in service operations by reducing processing overhead associated with third-party package managers like Helm. Additionally, it enhances our ability to swiftly diagnose and resolve issues when they arise.
While we seek to implement these improvements, we're keen on maintaining the integrity of our existing infrastructure configuration and application services. To that end, we are in search of guidance and documentation from AWS that can help us navigate this transition effectively.
Here are some specific questions and topics we would like to explore:
1. Deploying Thanos Without Helm on Prometheus/Grafana in EKS Clusters: We aim to understand the steps and best practices for deploying Thanos on our existing Prometheus/Grafana setups running on EKS clusters without relying on Helm.
2. Add-Ons in "eks-monitoring" and Their Use of Helm: We've noticed that the configuration of the "eks-monitoring" add-ons contains Helm settings. We'd like to clarify if these add-ons are indeed deployed using Helm and, if so, how this integrates into our monitoring setup.
3. VM Sizing Recommendations for Development, UAT, and Production Environments: We are interested in AWS's recommendations regarding the ideal virtual machine (VM) sizes for different stages of our environments, namely development, user acceptance testing (UAT), and production. We aim to optimize resource allocation based on best practices and cost-efficiency.
We genuinely appreciate any insights, documentation, or recommendations that AWS or the community can provide to assist us in making these crucial improvements while ensuring the stability and performance of our AWS infrastructure. Thank you.