hamburger icon close icon

Is Kubernetes creating blind spots in your end-to-end observability?

Can you map containerized applications to the physical underlying resources?

The market for Kubernetes observability and monitoring tools is unique.

On one hand, you have a variety of excellent tools that are built for application owners to maintain and to scale their applications, leaving the infrastructure teams to adapt to the applications’ needs.

On the other hand, many different tools can be used to manage and to monitor infrastructure without an understanding of the applications that are driving the resource consumption.

To the application owner, Kubernetes is their infrastructure. But what about the underlying resources that support the clusters, such as hosts, networks, and storage? This point is where I think that Kubernetes introduces a blind spot in end-to-end observability and where the market tends to fall a bit flat. And it’s also where NetApp® Cloud Insights can help.

Application development in an example organization

I’d like to elaborate by using a specific example of a large financial organization that came to the NetApp Cloud Insights team to ask for help. This organization uses Kubernetes at scale. They have thousands of nodes across multiple clusters that are built on various platforms and technologies, both on premises and in the cloud. They have some purpose-built clusters and some Kubernetes services, and much of their application development focuses on containers.

It’s also an organization that, just a few years ago, had successfully architected their infrastructure as their own private cloud, and things were clicking in terms of infrastructure as a service (IaaS). Application owners understood all the resources that were available, the service levels and costs, and exactly which levers they could pull to improve performance or to reduce cost. And it was all possible because of the end-to-end visibility of their entire infrastructure, and their ability to map their applications and needs to the supporting infrastructure.

The Kubernetes gap between applications and physical infrastructure

But Kubernetes changed all of that. The large-scale adoption of containerized applications certainly brought to this organization measurable benefits for both the application and the infrastructure teams. But they were missing the ability to connect the dots between applications and physical infrastructure, breaking the utility service model.

The application owners became blind to available resources beyond the already-built Kubernetes clusters, and when performance issues arose, they couldn’t identify underlying resource contention.

The infrastructure owners could easily identify resource contention, but they had no idea how it was affecting the applications’ reliability or performance.

And this example is just one organization. Many others are operating with the same challenges today, and many other organizations will face these challenges in the future. The as-a-service model in the cloud can absolutely help, but we’ve seen an overwhelming majority of organizations that continue to maintain their on-premises footprints.

Full visibility from Cloud Insights

We built NetApp Cloud Insights to bridge this gap—with full-stack visibility, from the namespace and containers down to the physical infrastructure that’s being consumed, including hosts and storage.


With the Cloud Insights Kubernetes cluster explorer, you can easily navigate the stack, identify risks and resource consumption, and understand who your top consumers are.



Cloud Insights dashboards highlight the most important metrics and key performance indicators (KPIs) from both sides. So, both your applications teams and your infrastructure teams benefit.


For metrics that are common to infrastructure teams—like IOPS, throughput, latency, and utilization— you can also easily see the attribution and a correlation to application performance metrics.


And so that you don’t miss SLAs, Cloud Insights full-stack alerting proactively alerts you to application and infrastructure issues.


And yet another feature is coming soon: advanced Kubernetes chargeback for roll-up reports of infrastructure consumption and utilization per namespace and the associated costs.



Reconnect your systems and your teams

All of these Cloud Insights features can help your organization more successfully run Kubernetes as a service. You can bridge silos and enable your applications and infrastructure teams to get back to the understanding of consumption, performance, and availability that they once had.

Learn more about how NetApp Cloud Insights can help bridge the gap between applications, Kubernetes, and underlying infrastructure. Check out my session, “BRK-1246-3, Monitor  Kubernetes Like a Boss,” at NetApp INSIGHT® 2021.

Solutions Architect