Kenji Kaneda

Chief Architect, CloudNatix

Today’s DevOps teams are always busy. Search DevOps memes and you’ll see many of the reasons why. Something always needs to be done – immediately. Whether talking about cluster availability, security incidents, or application upgrades, there’s never time to evaluate your cloud and k8 environments holistically. Unless of course, your VP of finance catches wind of an increasingly expensive cloud bill. Now it’s a top down priority, and you have to find the time. 

What must be done to reduce cost? 

You’ll have to get a true picture of your environment, analyze it, and then optimize it. These are very challenging tasks. Your cloud provider probably provides a breakdown, and with enough time and some attention to detail, you could probably run some parallel reports to make inferences on where you could potentially cut costs. Or worse, the changes are suboptimal. Despite their best efforts, companies tell us that this optimization results in 7-9% cost reduction.

Suppose you have tens of K8s clusters. Utilization is low. These clusters use some instance types that were chosen at some point in time. But how do you choose where to optimize your workloads? In theory you know there is a better instance type, but don’t know what it will impact overall. For example, do you know when your pods will start pending, will a decision actually increase your costs rather than decrease? And then once you make the change, what happens to the stability of your workloads?

What’s lacking in this scenario? 

A solution that automatically analyzes and optimizes the infrastructure is what’s lacking here. Devops teams are always busy with many high-priority tasks, you need a solution that can be easily applied. Also given the complexity of cloud and K8s, you need an automatic system that analyzes and optimizes the environment without relying on manual analysis. 

What if you could see where and how to optimize easily? 

CloudNatix customers report cluster-rightsizing as a key feature in keeping their environments optimized. By collecting historical usage of clusters, running various simulations to find optimal cluster configuration, CloudNatix cluster-rightsizing guarantees a recommendation you can apply safely without breaking your production environment. Also the installation is simple. You just need to install CloudNatix agent to your cluster to get the benefit. 

When your VP of Finance asks where you can cut costs, your answer is simple: “Here’s what we’ve saved in the last 90 days using CloudNatix.” Cost cutting and analysis is part of your day-to-day DevOps culture. 

Schedule a demo and learn how to make cost cutting part of your DevOps culture.