At this year’s KubeCon, there was a lot of talk about Kubernetes cost optimization and allocation; and many of these attendees were looking for alternative options to Kubecost. This blog is not intended to provide pros and cons of Kubecost, but instead offer some reasons as to why organizations need a Kubernetes cost allocation and optimization solution and provide some information on Fairwinds Insights, a platform that provides workload cost allocation, optimization and on top of that offers Kubernetes security and guardrail capabilities.
Cloud native tooling offers many advantages such as autoscaling, but also leaves DevOps and devs open to configuring the way they want. That means in many cases, Kubernetes limits and requests are not set, CPU and memory are not set and cloud cost can skyrocket. The result is surprise cloud bills with many unknowns. As organizations look to reduce spend, this is an area that is demanding attention.
Unfortunately, the Kubernetes platform can become a blackhole of cloud spend. That’s why vendors like Kubecost and Fairwinds offer solutions. While there are absolutely some great organizations that track cloud spend a level up in cloud providers like AWS, Azure, GCP, GKE, AKS or EKS, i.e. companies like CloudHealth, most don’t provide Kubernetes-level detail. As consumption of containers increases, this is a required solution.
Kubernetes cost allocation is hard. The dynamic nature of Kubernetes scheduling means where containerized workloads are run can always change. All this change means it is really difficult to break down costs in Kubernetes by workloads. Proper configuration is needed which calls for solutions that can monitor cost, make recommendations for optimization and provide workload cost allocation, and provide cost showback.
Fairwinds Insights offers cost management across Kubernetes and containers to make recommendations to right-size application resources. It provides a centralized, consistent and optimized view into Kubernetes costs. FinOps and DevOps teams alike use Insights to gain a view across many people and workloads to better align on opportunities to reduce spend or to provide evidence based on actual usage as to why increased cloud consumption may be required.
Core features available of Fairwinds Insights Kubernetes cost capabilities includes:
Workload Cost Allocation - Use actual cloud spend and workload usage to understand historical costs (up to 13 months) incurred across multiple clusters, aggregations, and custom time periods.
Kubernetes Cost Optimization - Evaluate individual applications and find opportunities to reduce costs without impacting application performance.
Rightsizing Advice - Maximize the efficiency of compute and memory utilization for your Kubernetes workloads with monitoring and advice on resource requests and limits.
Kubernetes Cost Showback - Report Kubernetes usage costs to finance teams, allocate to developers and track savings over time.
Multi-Cluster Cost and Usage - Get a breakdown of cluster capacity and usage across cloud resources. Understand how much is spent on idle capacity, shared vs. app-specific resources, and effective node scaling.
Cloud Billing Integration - Use actual AWS cloud bills to calculate Kubernetes costs by workload, namespace or label. Provides accurate, usage-based cost data across multiple business dimensions.
Quality of Service Controls - Collaborate on application right-sizing, giving purpose-built recommendations that eliminate guesswork and drive better workload efficiency and performance.
Fairwinds Insights can be tried for free just like Kubecost. Users simply need to register with our team, add the agent and see recommendations in the Insights dashboard.
Note: If you are a KubeCost or OpenCost, a CNCF sandbox open source project, user and want to try Fairwinds Insights, you can bring your existing Prometheus installation, or choose to install a pre-configured Prometheus package from Fairwinds Insights. Get in touch.
One unsurprising take away from KubeCon is that Kubernetes users are looking to consolidate vendors as organizations look to reduce spend. With the looming recession, and cloud billing increasing, it’s no surprise that cost is becoming a focus.
At the core of Kubernetes cost optimization is configuration. That’s why this functionality is part of the Fairwinds Insights platform. Insights scans Kubernetes clusters to identify misconfigurations for security, reliability and cost efficiency. It also implements Kubernetes guardrails (or policy) so that organizations can set a “paved road” for developers to build applications that are secure, cost effective and compliant.
Fairwinds Insights is available to use for free. You can sign up here.
In one platform users get the full gamut of what’s needed to get Kubernetes aligned with business goals (i.e. ship applications faster, scale reliably, gain cloud cost reports, etc) and security. Developers can work smarter with safety nets without having to be concerned with all the security, compliance or cost configurations Kubernetes requires.
Insights provides a more all encompassing solution to the challenge of Kubernetes cost by addressing not just one pain point (like Kubecost), but three. DevOps teams can stop being a Kubernetes help desk or spending time attempting to decipher cloud bills and pricing for the finance team. It frees up time for DevOps teams to do the innovative platform work they want to do - helping companies retain the Kubernetes talent they need.
For those organizations that might not be ready for Insights, an open source tool, Goldilocks, is available to try. It helps organizations right size workloads based on cluster metrics. You can check it out on GitHub or on the documentation.