Basically, we use it for our data lakes because users are not always making the best decisions money-wise. They mainly look at the performance of their environments, and they select very performant machines or storage classes even if they don't need them. 

And so the cost optimizer will tell us if that user is using a very expensive storage or if it's not using not using it. Then, of course, we will force the users to go to cheaper storage. That's one thing. 

An end user always or, in most cases, always selects the best because he thinks or believes he needs it.

 But as he is not a technical person, in many cases, he is not able to decide if he needs it or not. The cost controller is very interesting because it will show exactly what the end user is using and what he needs. So, that is a very, very good product. We make a lot of money out of it.

For our specific needs, Kubecost has everything we require. Performance monitoring across CPU, network, and storage is comprehensive.

I like the overall product because I can select what monitoring should be enabled and whatnot. 

In our case, we really focus on performance because it's clear that the price is related to most performance setups. So the more performance, the more expensive. 

So we look into the performance that the customer needs, and then based upon that feedback from the remote control, we change the parameters. And even the end user will not notice it is not using it, so we just make money without any impact on the end users.

In future releases, I would like to see faster monitoring because it could potentially improve overall stability in the production environment.  

I have been using it for a year. 

I would rate the stability an eight out of ten. While the monitoring itself has a slight performance impact on your environment, it's inherent to the system and unavoidable.

Faster monitoring would minimize its impact on production environments.

The scalability is okay. I would rate the scalability a ten out of ten. For cost optimization, not many people use it. The whole environment uses it, but only two or three specialized people actively analyze the data.

It's for specialized personnel, and with that team, we can monitor the entire environment effectively.

Understanding the applications, data flows, and user base is key before deployment.

You need a thorough understanding of your company's size, target audience, and software/data flow to define effective monitoring choices and identify improvement opportunities. This requires skilled personnel to analyze and implement.

We handled it internally thanks to our in-house expertise. However, I imagine most customers would benefit from external consultants or third-party services.

Deployment time is longer because we have to decide the right monitoring parameters. This has a significant impact on results, so expertise is crucial.

Third-party consultants or specialized personnel might be necessary for most customers as they may lack the in-house knowledge for optimal parameter selection.

For an AWS setup, simply moving to the cloud without optimization can be expensive. The real savings come from using Kubecost features like autoscaling and serverless functions to optimize your resource usage. If you treat it like a data center migration without fine-tuning, it might cost more.

The price is reasonable, considering the value it delivers. 

In all honesty, once you have your optimal design, you could just turn it off and then activate it maybe once every six months or once every year to make a check again. 

You only need it in the beginning. Once people are optimized and fine-tuned, you can stop monitoring. So, it's very well designed. It's a good product. It's it's running nicely. It gives us fair numbers of advice on how to improve costs.

There's a small licensing fee, but you don't need to run it 24/7. Initially, you might use it full-time for a few months to gather data, then switch to periodic monitoring like monthly or bi-monthly. Ultimately, the cost savings from optimization outweigh the licensing fee.

The price scales based on your monitoring parameters, such as the frequency and duration of data collection. The more data you collect, the more you pay for it.   

Overall, I would rate the solution a ten out of ten.  It's an excellent product. 

Make sure you have dedicated people with a good understanding of your environment to use it effectively.

I use Kubecost to optimize cost across 20 enterprises. 

So, cost optimization is my primary use case.

Kubecost has helped the organization optimize costs for various services, including Kubernetes, containers, Docker, and others. It also helps optimize node costs in Kubernetes.

I mostly like the dashboards. They are very informative and helpful. I also like the optimization functionality. It effectively reduces costs.

I like these features because they are effective in optimizing costs.

The integration with other solutions could be improved.

It's not difficult to integrate with terabyte images from Docker. I built the YAMOs and put them on Kubernetes. However, there is room for improvement in integrating with other solutions.

I've been using Kubecost for six months.

I would rate the stability a ten out of ten. It is very stable. 

I would rate the scalability a seven out of ten. The people have been using it. So, for an organization, there are six to ten people using it. 

I like the customer support. 

Positive

I would rate my experience with the initial setup a ten out of ten, where one is difficult and ten is easy.

It's on cloud. It's deployed on Azure, AWS, and Google Cloud.

I deployed it. One person is enough. 

We have seen a return on investment. For example, the company spent $500 on Kubecost, and they saved $220 for one enterprise and $330 for another enterprise.

The cost is cheap. Kubecost has an open-source core.

I advise to use Helm or YAML as a deployment method for Kubecost. I would recommend this solution to others. I like it.

Overall, I would rate it a ten out of ten. 

Kubecost is a good cost management tool for services that are accessed by users from the organization.

It analyzes your Kubernetes cluster, providing recommendations to optimize costs. Our project involved integrating three tools for cost optimization, remediation strategies, and security measures.

The standout feature is its ability to provide a comprehensive view of the environment's costs, which is particularly crucial as Kubernetes adoption becomes widespread in the industry. It offers a detailed examination of your cluster, including the types of instances utilized, allocated CPU and RAM, and resource distribution for specific applications. Another one of its strengths is the real-time nature of its updates, refreshing every hour to deliver precise insights and recommendations with a non-disruptive approach; the recommendations it provides won't compromise your system but can significantly positively impact your monthly expenditure. It is worth mentioning its ease of deployment. With minimal configurations required, it simplifies the process with a set of commands, making it straightforward to implement. The tool also offers a user-friendly dashboard, enhancing the overall experience by providing a visually intuitive representation of the data.

There is a significant potential for enhancing it through the incorporation of advanced technologies like AI and generative AI. By leveraging AI capabilities, the tool could evolve into a more automated solution, streamlining processes and decision-making. This automation could empower the tool to apply optimizations and make decisions based on predefined conditions, reducing the need for manual intervention.

It stands out as a highly stable product, validated by its recommendation from AWS for obtaining valuable cost feedback. The fact that AWS endorses it adds a layer of trustworthiness.

It serves as a singular tool capable of consolidating diverse environments. Whether you have multiple clusters for various purposes such as staging, testing, and production, or if these clusters are spread across different cloud providers like Azure, GCP, Oracle, and AWS, it excels at bringing them all together. Its scalability is a key strength, offering a centralized dashboard that provides visibility into all these environments. In total, there are two to three teams, amounting to around thirty to forty people, actively using its capabilities.

I would rate their support nine out of ten. Whenever we reached out to them, they promptly opened a ticket and assigned an agent to assist us over Slack. Throughout our engagement, they demonstrated a high level of cooperation and assisted in every possible way. 

Positive

It provides two deployment options: a straightforward deployment within the cluster, ideal for basic insights and quick metrics using installed agents, and a more detailed analysis option, allowing users to obtain real-time cost data directly from AWS by configuring S3 buckets and generating a Cost and Usage Report. Despite the intricate setup for the detailed configuration, clear instructions on the website simplify the process, taking about an hour or an hour and a half. Alternatively, users seeking a swift deployment and concise overview can opt for a simple deployment within a single cluster, achieved with just one or two commands. This basic deployment takes only five to ten minutes.

We opted for the open-source version as our team comprised Kubernetes specialists with certifications in the field. Although we didn't personally use the paid version, I acknowledge that it could be a worthwhile investment. The cost of the tool may seem nominal compared to the potential savings in infrastructure expenses. The decision to go for the paid version might depend on the scale of your environment. For smaller clusters, the open-source version can provide sufficient insights. However, for larger environments with hundreds of nodes and thousands of containers, exploring the paid version could be a beneficial consideration.

Our focus was primarily on Kubecost because we were exclusively working within the Kubernetes environment.

I highly recommend using it, especially if your environment is based on Kubernetes. By integrating it with your cluster and following the steps and configurations outlined on their official website, you can gain valuable insights into your cluster's performance and costs. It boasts a great community and excellent support, featuring a user-friendly interface that provides comprehensive insights into the cluster. It offers detailed information about resources, including pods, containers, services, and various Kubernetes objects. Additionally, it seamlessly integrates with major cloud providers such as AWS, GCP, and Azure. Overall, I would rate it eight out of ten.