Hardened Amazon AMI with kubectl for EKS

I am currently using Amazon EKS  to host our applications to provide scalability and security.

Amazon EKS  is providing auto-scaling, which is one of the features that everyone is moving to because it eliminates the need for manual deployments on EC2  machines or on Docker .

Kubernetes  manages all those things, which is why people are moving to Amazon EKS. For Kubernetes  as a project, it can be difficult to manage, but that problem is solved by Amazon EKS so I just need to think of our application rather than the infrastructure, and AWS  manages all the infrastructure.

Because of Amazon EKS, if I do manual deployment on Docker  and my particular container fails, I will not know the reason. However, Amazon EKS automatically restarts that particular container or stops it and creates a new one, switching the traffic accordingly.

This is why Kubernetes provides a better experience. Amazon EKS provides scalability, which is one of the features everyone is using. In case of high traffic, it automatically scales the nodes and pods, and we can scale it horizontally or vertically.

The authentication process means that with IAM  policies, I am giving the least privilege access so only particular persons are able to manage Amazon EKS clusters. Everyone does not have access, which creates a kind of zero-trust security.

In a minor enhancement sense, I just needed some documentation on that, which has already been provided, but in simple English language that is easy to understand for freshers as well.

I also needed to have some infrastructure to do some practice sessions before we could move on, because the pricing of Kubernetes and Amazon EKS is too high. For some people, it is hard to get hands-on experience with those.

I have been using Amazon EKS for almost 1.5 years since I joined the cloud infrastructure team.

I would rate Amazon EKS around 9 out of 10 for stability. It is stable.

I did not need to communicate with the technical support of Amazon EKS yet. I have not communicated with their support specialists for Amazon EKS, but I did for other things.

Before Amazon EKS, I used manual deployments such as getting the code and running those on servers. Then Docker containers came into place, running code on Docker, followed by using ECS, and then Amazon EKS came into play. It was a logical update that I chose to go with.

I participated in the initial setup of Amazon EKS in the deployment process at my previous organization. For the initial setup, there are types of deployments for Amazon EKS cluster. Via the console, it is somewhat difficult, but via CLI, it is easy. You just need to set up AWS  CLI and configure it, and then a simple command is AWS EKS CTL create, and we can create our Amazon EKS clusters. It is straightforward and not complex.

When purchasing Amazon EKS service for my application, if I have only one or two pods, I need to pay 75 dollars per month for that control plane regardless of what I am hosting, which can be challenging for small-scale companies. For small-scale companies, it is a challenging situation.

Let me provide some additional context. Jenkins  is one tool to consider.

AWS DevOps is also available. For monitoring, Grafana  and Prometheus are options. There are also some open-source tools that we are working on, such as Signoz.

Microsoft Entra ID  is one of the products which we are using. For AWS, we are using AWS ECS, Amazon EKS, IAM  for authorization and authentication purposes, and databases. All the services which are on AWS are being utilized.

Amazon EKS is different from Kubernetes because it is a platform built on Kubernetes but managed by AWS. Amazon EKS is already dominating the market more than others, so there is no enhancement from my side.

I utilized Amazon EKS's integration with IAM  and OIDC. IAM  roles have provided the capability to give permission to Amazon EKS that this particular Amazon EKS is able to access particular services such as the S3  bucket, which is one of them, and RDS , which is a database. We are providing roles to Amazon EKS clusters so that on our behalf, it can access these particular things. By that, they can access the database and objects in the S3  bucket.

I am not currently using Amazon EKS's self-healing nodes. In another organization, we are not on Amazon EKS.

For the nodes, I can tell that it provides auto-scaling to minimize and maximize the number of nodes as traffic comes. By that, the pods will run on those, and if there is any failure on nodes, it gives another node.

I would rate this product a 9 out of 10 overall.

Public Cloud

Amazon Web Services (AWS)

My main use case for Amazon EKS  is to host one of my AI agents which I built recently in my company. I use Amazon EKS  to host my AI agent, which is a sprint report agent that fetches the company's data from Jira  and prepares a sprint report for the complete team. The agent runs on Amazon EKS for high availability and reliability, hosted on nodes, and it is running on a GPT OS's model in the back end.

I have designed this agent for high availability on Amazon EKS because if I hosted it on any other platform, there would be chances of downtime. I have set it up in such a way that if there is any downtime, a new node is already up and running so that my use case is not affected and the users can use it seamlessly without any issues.

The best features Amazon EKS offers in my experience are its reliability and high availability, and also the support that AWS  provides. The reliability and high availability have helped me specifically because there were some configuration issues at my end for the agent, which caused it to go down repeatedly. Due to the high availability features, every time a node downscaled, a new node was automatically scaled up, ensuring that the AI agent was up. I was able to check the logs and correct the issues accurately thanks to that high availability. Also, AWS  support is very helpful; whenever I face issues regarding costing or anything else, I just create a ticket and they assist me in resolving the issues.

Amazon EKS has positively impacted my organization by improving the efficiency and working capacity of my team. It improved efficiency and working capacity because, based on how Amazon EKS works, we are more calm regarding functionality. The reliability allows us to focus on many other tasks, as the infrastructure is maintained by Amazon EKS; therefore, we can divert our attention to other tasks and perform well there as well.

Functionality-wise, Amazon EKS is acceptable, but navigation-wise, it could be improved in the AWS console; it could be more interactive and more intuitive for new users. Also, the pricing can be reviewed as it is sometimes a bit pricey, particularly regarding the extended support when new version upgrades occur that we cannot implement directly due to production workloads, as clusters running on extended support cost six times more, which is something that could be reduced.

I believe documentation could be improved on the AWS website so a new user who is starting with Amazon EKS could work much better with it.

I have been working in my current field for three years.

In my experience, Amazon EKS is stable.

The scalability of Amazon EKS is good.

The customer support of Amazon EKS is good. I would rate the customer support on a scale of one to ten as a seven.

Positive

I have seen a return on investment because the reliability has helped me save time; I can rely on Amazon EKS's reliability and then divert my attention to other tasks, so it has definitely saved my time. I estimate that my team and I save roughly one to one and a half hours a day since using Amazon EKS.

Regarding pricing, setup cost, and licensing, the pricing for the cluster part is a bit higher; the setup cost is acceptable, but the licensing part regarding extended support is also a bit pricey, which I think can be handled or reduced.

Amazon EKS is a good product; if you are starting with Kubernetes , it is a good choice, but the pricing is a bit substantial, so you should review that. Also, regarding the support, there are sometimes cases where you need to upgrade your AWS plans for particular support, which can also be a bit pricey. I would rate this product a seven overall.

Private Cloud

Amazon Web Services (AWS)

I used Amazon EKS  throughout my internship while working on Docker-based deployments, Kubernetes  orchestration, and cloud infrastructure tasks at Cognizant.

I used Amazon EKS  when working as an intern at Cognizant, where it was used to deploy, manage, and scale containerized web applications. Our workflow started with building Docker  images, storing them in Amazon ECR , and deploying them to Amazon EKS clusters running on EC2  worker nodes with CI/CD pipelines. We used many tools; for example, Jenkins  was part of our CI/CD pipelines that automated the build and deployment process.

Additionally, I worked on application deployments, updating Kubernetes  manifests, managing pods and services, and verifying application health. Amazon EKS acted as a central platform that connected Docker , AWS  infrastructure, and DevOps automation into one consistent system.

I used Amazon EKS during my internship at Cognizant as part of a cloud and DevOps-focused environment, where it served as the core Kubernetes platform to run containerized applications built with Docker, deployed through CI/CD pipelines, and hosted on AWS  infrastructure. We deployed numerous web applications, and we wanted to learn Amazon EKS through dummy projects with dummy web interfaces. Beyond dummy projects, we also deployed some client websites into the Kubernetes environment and managed traffic, although I cannot name the clients.

Amazon EKS is an excellent choice for organizations already invested in AWS. I recommend having a solid foundation in Docker, Kubernetes basics, and AWS core services before implementing Amazon EKS. Using infrastructure-as-code tools and following AWS best practices can significantly improve maintainability and security. Amazon EKS is particularly strong for enterprise environments and microservice-based architectures.

Amazon EKS is ideal for teams already using Docker, CI/CD, and AWS infrastructure, which our team was already utilizing. I strongly recommend learning Kubernetes fundamentals and AWS networking, containers, and security before using Amazon EKS in production, as it positively impacted our organization by making it easy to connect all our existing AWS services.

I deployed Docker applications to Amazon EKS using CI/CD pipelines, integrating with EC2 , ECR, IAM , and automated workflows.

The most valuable feature of Amazon EKS is the managed Kubernetes control plane, which allows teams to focus on DevOps and application workflow instead of cluster maintenance. Another major benefit is how naturally Amazon EKS fits into the AWS ecosystem, integrating with EC2 for compute, IAM  for access control, and Amazon ECR  for container images. CloudWatch for monitoring creates a complete DevOps pipeline. The self-healing nature of Kubernetes combined with AWS scalability makes the environment reliable and suitable for real-world workloads.

The most promising feature, which I prefer the most, is its integration with all the AWS services, including EC2, IAM , VPC, ECR, and CloudWatch, making it a key part of my workflow.

Amazon EKS works very well with Docker-based container workflows; it is highly scalable and self-healing, complemented by its rolling update capabilities.

There are some drawbacks regarding Amazon EKS; pricing can increase as clusters and workloads scale, and there is an initial configuration learning curve. A beginner has to learn about IAM  networking and cluster setup, plus improved built-in cost visibility and simplified monitoring tools could be beneficial.

Pricing can be improved, especially for small teams or landing projects, and the initial setup, as well as understanding IAM networking and cluster configuration, can be complex for beginners; improving this would enhance the experience. Troubleshooting sometimes requires deeper AWS and Kubernetes knowledge, which also could use improvement.

Amazon EKS is very stable; during my usage, the clusters remained consistently available and workloads ran reliably. The self-healing capabilities of Kubernetes, combined with AWS managed infrastructure, help ensure minimal disruption to running applications.

Amazon EKS is highly scalable; it supports horizontal scaling of pods and seamless scaling of worker nodes. It features AWS auto-scaling for vertical scaling, making it suitable for handling varying workloads and preparing the environment for real-world production demands.

Support and documentation from AWS are very strong, with extensively available official documentation, and AWS support channels are very kind and responsive. Most issues can be resolved through AWS knowledge bases, documentation, and our organizational support teams.

Positive

Previously, I used open-source Kubernetes as a normal Kubernetes solution. I wanted to switch to Amazon EKS because Amazon EKS connects with numerous AWS services, making it easier to deploy my applications than using open-source Kubernetes.

The initial setup is moderately complex; creating clusters, configuring IAM roles, setting up the network, and connecting with the worker nodes requires careful planning. However, once the environment is configured, ongoing operations such as deployment, updates, and scaling are smooth and efficient.

Time is definitely saved because Amazon EKS provides automation with CI/CD pipelines, allowing us to simply monitor it, and if there is any fault, we know immediately in the pipeline. Regarding the need for fewer employees, I do not think that will happen since only a few employees will know about Amazon EKS deployments in the cloud, so some knowledge is necessary. We can see a return on investment; we can save a lot of money by using Amazon EKS as it is directly connected with all AWS services and can be integrated with coding storage options such as GitHub  and GitLab .

I did not evaluate many options; I evaluated some, specifically open-source Kubernetes, which I thought would be more difficult than Amazon EKS. Therefore, I chose Amazon EKS.

The benefits that I observed after adopting Amazon EKS are improved deployment speed and reliability, while resource usage remains the same, but we can handle a larger amount of traffic and application deployments without any issues. The user interface is very great, providing clarity on where the fault or error lies while deploying our application. I would rate this product nine out of ten.

Public Cloud

Amazon Web Services (AWS)

My main use case for Amazon EKS  is for implementation and sustainable services and microservice application on a critical structure and services deployment.

On our application, we have more than 20 services and microservices such as authentication, login, account management, a notification service, and a billing service, which all work together to structure a heavy, useful application.

The best features Amazon EKS  offers are scalability and deployment control, the ingress configuration regarding path pattern and host header to get all the services and microservices, and the HPA configuration.

The biggest difference, or the most important aspect to me, is the scalability, because you can easily scale any service or microservice to handle security during high changes in connection flow, and it is useful for the application and helps day-to-day by giving us reliability and stability so we can perform all maintenance and deployment of our system.

Reliability is a very important thing. Security and operational consistency are very important aspects, and the flexibility offered in node management and network options is also valuable. Amazon EKS is a service that is reliable and scalable, and it gives us a solid and dependable solution.

I think sometimes the documentation is not so clear and not so fast to provide more in-depth instruction and examples of bigger and critical implementations, so some difficulties for us sometimes take a lot of time to understand, test, and to put into production with security and guarantees.

I have been using Amazon EKS for almost five years now.

I advise doing a POC first and getting all the details, testing, and having a very good alignment between DevOps and development departments, and prepare all the CDN  and how the connections get into your cluster, and how you configure your ingress and how to prepare every service or microservice to receive that with secure and optimized code, process, and communication with other resources. I would rate this product an 8.

Private Cloud

Amazon Web Services  (AWS )