We’re excited to bring customers a new level of visibility for containerized infrastructure running on public cloud with Cloudability Container Insights 2.0. Running this infrastructure comes with unique governance challenges, and over the years we’ve worked hard to help customers bring financial accountability and optimization insights to this spend. Going back to 2020, Cloudability was the first solution to deliver fully integrated container cost management, and we’ve continued to innovate since. Recent advancements include factoring in family-based weightings, resource-contribution cost visibility and idle cost distribution in core analytics. With Container Insights 2.0 you’ll notice a whole new user experience, with new insights and ways to interact with detailed container information.

Important Note: During this transition phase you can toggle between the new and previous experience using the button at the top of the page.

Personalize Your Insights with Customizable Dashboards

For the first time you can create your own dashboards within the Container Insights feature. Every user starts with their own default dashboard, providing some great example KPIs and charts. We encourage all of our users to create their own dashboards, choosing from different widget types and filtering based on specific needs.

Figure 1: Example personalized dashboard

All dashboards are supported by the main pivot table at the bottom, providing an entry point for deeper exploration and our new Treemap functionality!

Visibility into Workloads, Containers and Nodes via Treemap

Users can navigate to the new treemap functionality either contextually through the main pivot table or via the dedicated tab at the top. Within the treemap page you will see three separate visualizations: the treemap, a daily cost chart, and the pivot table. The best way to explore costs and other metrics is via clicking in the treemap itself, with the natural path being Cluster -> Namespace -> Workload -> Container -> Node.

Figure 2: Treemap view, drilled down to containers

Bringing visibility to underlying Workloads, Containers and Nodes is an important component of the Container Insights 2.0 launch, going beyond the typical allocation constructs of Namespaces and Labels. We expect this to help engineers and FinOps practitioners to better understand the makeup of their clusters and simplify optimization analysis.

A Powerful New Efficiency Metric

Throughout Container Insights 2.0 you’ll notice a brand-new metric that can be used at all layers (e.g., at Cluster, Namespace or Node level) to quickly indicate efficiency and guide optimization actions. You can think of the new “Efficiency Score” as resource allocation efficiency, as in what percentage of the underlying resources (CPU, memory etc) from a cost perspective have been allocated to workloads. This allocation is calculated based on a combination of resource requests and measured usage, and is represented in the platform as “Utilized Cost”. Utilized Cost divided by the Total Cost gives you the efficiency score. Low efficiency scores therefore indicate excessive headroom, with large amounts of provisioned resources not tied to underlying workloads. Common causes are an excessive number of nodes (VMs in this case) within a cluster or poor workload placement (leading to localized inefficiency).

Stay Tuned for Coming Releases

Over the next few weeks there will be additional updates to Container Insights that will deliver significant value to customers. These include:

·      Surfacing detailed information at the node level including the VM id (e.g., EC2 instance ID), VM instance type and commitment status.

·      A Utilization Efficiency metric, which will expand on the Efficiency Score by considering whether allocated resources are actually being used.

·      The ability to share dashboards with team members.

Please follow this post if you’d like to be notified as these releases land, and of course please leave a comment if you have feedback to share, we’d love to hear from you!