My primary use case with PerfectScale is having visibility on our Kubernetes clusters. It helps reduce our costs by providing recommendations to define resources for services as needed and avoid overprovisioning. There is an autoscaler that can be defined on the cluster or workload level, which allows us to control our workloads and pre-define the resources for each use case.

In my opinion, it has improved our organization because it gives us more control over our Kubernetes cluster. It also provides more knowledge on how to handle high-cost scenarios. I was able to break down the costs and build a solid report, which allowed us to take action against high-cost situations.

The cluster and workload autoscaler gives us the ability to have control over all the workloads' resources instead of managing them one by one. We typically start from there, and it does the job very well. The recommendations are based on historical data that PerfectScale gathers, providing us with more information on how the workload should be defined based on timeframe and revisions.

I think they should focus more on Kubernetes features that allow on-the-fly resource allocation without the need to restart services. They should implement this in their autoscaler to make it more useful in scenarios that require immediate scaling up or down. They should also offer more options for visualizing graphs in different ways, such as tabular views.

I have been using the solution for approximately 1.5 years.

Their support team consists of really good people who assist us in understanding the product. They also share upcoming and recently released features with us. They genuinely take care of their customers. I would rate it an eight out of ten. 

I used PerfectScale to gain insight into the workload and then optimized the resource allocation through manual configuration or automated allocation. By doing so, I was able to achieve Kubernetes cost savings.

Automated resource optimization using different policies based on the environment enabled the organization to achieve infrastructure cost savings.

PerfectScale provides insights into workload resource consumption and enables dynamic resource re-allocation.

At the beginning, the support was not very impressive. There were some mistakes in the documentation, such as how to exclude a workload from automation.

I previously used Prometheus, but it is not as easy to use as PerfectScale.

There was not really a setup cost involved.

I did not really consider alternate solutions.

It is worth trying.