Everything you’ve ever wanted to know about the network — from data center to container to cloud: Kentik

The Silicon Review

Digital Experience Monitoring is a category of performance analysis technologies that provides visibility into the end-user experience, or what analyst firm Gartner calls the “omnichannel user journey.” That’s a fancy way of saying how the users behave as they interact with on-prem and cloud applications to do their job. Application developers are beginning to recognize the value that DEM can bring to the end-user experience. For this reason, many DevOps teams are building in APIs early in the software development life cycle as part of the testing process of new releases. These APIs allow DEM systems to properly measure the performance of the end-user experience and the underlying application components. These more modern DEM solutions are especially sought after by infrastructure and operations leaders because they want to use DEM to regain insight into the performance of applications as they migrate to the cloud.

Network engineers also recognize the value of DEM. The performance of the network directly impacts the digital experience of application users. For this branch of DEM, synthetic traffic is generated to proactively measure network performance for packet loss, latency and jitter. Since DEM can be used to measure the performance from the user’s point of view, the data collected can help uncover the impact that degraded performance can have on the business. Synthetic tests conducted by a DEM solution will ideally be performed from multiple geographic locations in order to identify problems with specific network links and understand performance metrics with respect to users in different regions. For example, Kentik has built out a global network of synthetic testing agents that are used by customers to verify performance levels of all major public or private cloud-based applications and SaaS applications. These agents have been located inside Amazon Web Services, Google Cloud Platform, Microsoft Azure, Alibaba Cloud, IBM Clouds and these vantage points keep growing.

Kentik demystifies Kubernetes networking

Kentik Kube gives cloud and infrastructure engineers detailed network traffic and performance visibility both inside and among their Kubernetes clusters — so they can quickly detect and solve network problems, and surface traffic costs from pods to external services.

This allows them to:

Discover which services and pods are experiencing network delays
Ensure pod, node and namespace communication patterns adhere to policy
Know exactly who was talking to which pod, and when
Identify service misconfigurations without the need to capture packets
Identify all clients and requesters consuming your Kubernetes services

“Enterprise applications have ridiculously complex and hybrid infrastructures,” says Avi Freedman, Co-founder and CEO of Kentik. “As network and platform engineers run critical applications and infrastructure, Kentik Kube shows performance and connectivity within the context of their entire network and internet infrastructure.”

Pods and services often experience network delays or errors that degrade the digital experience, and it’s difficult to identify them quickly. With the inherent complexity of microservices; network, cloud and infrastructure teams are left wondering if the network reality matches their design, who are the top requesters consuming Kubernetes services or which microservices are oversubscribed, and how the infrastructure is communicating both with itself and across the internet. Kentik Kube relies on data generated from a lightweight eBPF agent installed on Kubernetes clusters. It sends data back to the Kentik SaaS platform, allowing teams to query, graph, and alert on their infrastructure. With this new data, coupled with Kentik’s advanced analytics engine, these teams can move faster, reduce MTTI and MTTR and answer critical questions about the health and performance of their network.

How it works

Kentik Kube relies on data generated from a lightweight eBPF sidecar that is installed onto your Kubernetes cluster. It sends data back to the Kentik SaaS platform, allowing you to query, graph and alert on conditions in your data. Kentik Kube supports all Kubernetes clusters, whether cloud-managed (AKS, EKS and GKS) or on-prem, self-managed clusters using the most widely implemented network models.

Network performance and cost: Discover which services and pods are experiencing network delays, and surface traffic costs from pods to external services.

Top talkers: Identify clients/requesters consuming your Kubernetes services so you can track down problematic connections.

Policy validation: Ensure that your network reality matches your design. See which pods, namespaces and services are speaking with each other to ensure that your configured policy is working as expected.

Total infrastructure visualization: Know which pods were deployed on which nodes — even historically. See which pods and services are communicating with non-Kubernetes infrastructure or the internet.

Kentik Kube provides east-west and north-south traffic analytics inside and among Kubernetes clusters. Kentik will automatically detail your network map once you have deployed the eBPF sidecar. Kentik Kube can display details to see if your route tables, NACLs, etc. are all configured correctly. You can drill down into a cluster to see if there are latency or other issues. Kentik’s eBPF telemetry agent deployed into these clusters lets you see the traffic between the nodes and the latency.

Meet the leader behind the success of Kentik

Avi Freedman, Co-Founder and CEO has decades of experience as a leading technologist and executive in networking. He was with Akamai for over a decade, as VP Network Infrastructure and then Chief Network Scientist. Prior to that, Avi started Philadelphia’s first ISP (netaxs) in 1992, later running the network at AboveNet and serving as CTO for ServerCentral.