Optimizing Kubernetes workloads can be tough work. Not only do customers need to pull and analyze all the data for a workload and its usage patterns, they need to then determine what optimal looks like, validate that, and finally go and implement the optimizations.

Not anymore. What If analysis in Harness Cloud Cost Management now provides the capability to customize resource optimization recommendations in Kubernetes between cost and performance needs. Customers get the data validation they need, and can customize the recommendations based on context-relevant CPU and memory needs – now all they need to do is go and make the change. Easy.

Manual Validation Makes Saving Money Hard

Customers value the Kubernetes cost savings recommendations that Harness provides and wanted to see the math behind the magic to simplify their validation process. Previously, they had no way of knowing how recommendations were being made, and it required immense manual toil to pull the numbers and do the math on their end. Ultimately, recommendations don’t matter if the validation takes so much time that the cost savings opportunity is no longer relevant. Customers wanted to act on the potential savings.

A big part of that validation process was contextualizing the potential savings being surfaced by Harness. As much money as a recommendation might save, the people who need to implement the savings recommendations are the engineers and application teams who have the context necessary to say whether or not it’s actually feasible.

After all, there may be a good reason that a workload is provisioned the way it is, and that context isn’t necessarily available to a tool. In these validation scenarios, teams had to do their own math to figure out how best to optimize the workloads they owned, and match that up against what was being recommended.

The difficulty of doing this had consequences:

  1. It resulted in a lot of savings opportunities being invalidated, either because it didn’t have the right context and didn’t make sense to implement, or because validation took so long that the recommendation became stale.
  2. Teams became hesitant to trust and adopt the cloud cost management platform, creating friction between cost owners and application owners, and leading to frustration about not being able to realize the promised value from the platform.

Solution

At Harness, we’re committed to exceeding customer expectations. In addition to giving customers what they’ve asked for, we’ve gone a step further with What If analysis. With this new improvement to cost savings recommendations in Harness Cloud Cost Management, customers get the following:

  1. Simple data validation. Imagine spending minutes or hours validating cost savings recommendations instead of days or weeks. Customers can now see the data used to back up cost savings recommendations so that they can easily see why the recommendation was made, and minimize the toil of having to do manual validation. Less time spent validating = more realizing cost savings.
  2. Customize recommendations based on context. Customers can adjust the data that gets analyzed to make the cost savings recommendations. This way, they can modify the recommendation to reflect the needs for CPU and memory based on their specific context. In the end, they get a customized workload optimization recommendation that is tailored to what matters most, instead of blindly having to trust the tool. More relevant optimization recommendations = more realized cost savings.
  3. Make common decisions faster. Know that a certain workload will always need to perform at the highest level? What about dev workloads that can skimp on resources and be okay? Customers now get preset options to show recommendations that are cost- or performance-optimized. Less repetitive tasks = more time delivering cool features.
Cost-optimized workload recommendation

What are some of the scenarios in which customers might want to do this kind of What If analysis? Here are some examples of when cost vs. performance might make a difference:

Cost-Optimized WorkloadsPerformance-Optimized Workloads
Dev/TestProduction
Non-CriticalMission-Critical
InternalCustomer-Facing
R&D Cost CentersRevenue Drivers

How to Get Started

Getting started using What If analysis in Harness Cloud Cost Management is easy. All customers need to do is log in to the Harness platform and view their cost saving recommendations the same way as always. Users will be greeted by the new look and feel of the recommendations, including What If analysis, by default.

Learn more about how to use the feature and get started using What If analysis with the Harness docs.

Keep Reading