DEV Community

Cover image for Simplify your EKS Cluster Observability with Prometheus & Grafana

Simplify your EKS Cluster Observability with Prometheus & Grafana

Good day fellow Bloggers,

Its been a while since I posted a blog, I have been settling down in my new role and studying for a beast of an exam coming up.

Without further delay, Kuberenetes Observability Made Easy!

Image description

I would encourage everyone get used of the CloudFormation YAML Manifest construction, whilst we are seeing more vendor agnostic IaC platforms such as Terraform, it definitley helped me prepare for my AWS DevOps Professional Exam coming up, so I would encourage you to make sure you sharpen up your skillsets wherever you can. I have the link to the AWS CloudFormation Documentation so familiarize yourself with the whitepapers and dive right in!

https://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/Welcome.html

The below context will give you an example of my custom YAML CloudFormation Stackset:

Image description

Once compiled you can use API Calls with the AWS CLI installed either to your WSL (Windows Sub-System for Linux) or PowerShell Terminals. The CloudFormation stacks simply build and can be destroyed and re-constructed in minutes.

Image description

So utilizing the above toolsets I created a EKS Cluster and added a second stack with references to my existing stack.

Image description

Once deployed you can then add further toolsets such as the Prometheus stack set using the Helm Community Install and your almost at the end of your race.

I will recommend installing Kubernetes Lens as its much easier to install and manage your K8s cluster then directly managing it via CLI / However I am a CLI Jockey and still prefer the terminal.

Image description

Ensure that within your settings you configure your ./Kube/Config context destination so that lens can get straight to work as per the below:

Image description

Finally once the prometheus and helm stack completes the install you will be ready to start monitoring logs, alerting & Graphs / Dashboards using both Prometheus and Grafana.

Image description

Image description

For more tutorials feel free to visit: https://deadkithedeveloper.click/

Or

Connect with me on LinkedIn via
https://www.linkedin.com/in/devonpatrick-adkins-%E2%98%81-1068781a3/

Have a great week ahead!

Regards
DevonPatrick
AWS Community Builder & Cisco Champion

Top comments (2)

Collapse
 
indika_wimalasuriya profile image
Indika_Wimalasuriya

What do you think of levaring Amazon Opensearch instead of spending on ground up setting up on ELK?

Collapse
 
deadki0001 profile image
DevonPatrick Adkins

Amazon OpenSearch is always a more cost-effective and convenient option as its a fully managed AWS Service that does all the heavy lifting for you in terms of Logic - Scaling, Security, Monitoring Systems and even Managing backups. Opposed to the ELK Stack this requires a little more of a hands on approach whereby you need to do most of the hands on configuration yourself. My answer would be its really dependent on your businesses specific needs and don't really have the operational expenditure on skilled staff to do this on your behalf that performs the management and administration of the ELK stack on your behalf. Hope this helps answer your question.