Kubernetes / Observability / Sponsored

4 Common Kubernetes-Monitoring Traps to Avoid

12 Nov 2018 10:07am, by

AppDynamics sponsored this post.

Eric Johanson
Eric is a software engineer at AppDynamics. Prior to AppDynamics, Eric held engineering and sales roles at AltX before it was acquired by Addepar.

Kubernetes (K8S) now seems to be the de facto standard for managing and deploying applications built on microservices and containers — and it’s easy to understand why. Kubernetes is the largest open source community, it’s backed by the Cloud Native Computing Foundation (CNCF), it’s DevOps-friendly, and it delivers the benefits of the hybrid cloud. What’s not to love?

But while 69 percent of respondents in a recent survey said Kubernetes is their top choice for container orchestration, deploying and managing Kubernetes isn’t necessarily a walk in the park. Despite Kubernetes’ flexibility, there are operational workflow complexities — many involving application performance management (APM) — that need to be managed effectively in order for Kubernetes to deliver on its promised benefits.

Rethinking Your Kubernetes Monitoring Strategy

A recent CNCF survey revealed that 38 percent of respondents identified monitoring as one of their biggest Kubernetes-adoption challenges — one that grows even larger to 46 percent as the size of the enterprise increases. So, how can the modern IT leader simplify Kubernetes monitoring to drive efficiencies while also optimizing performance?

Current Kubernetes monitoring approaches fall short for a few different reasons, from lack of end-to-end visibility to error-prone migrations. Here are four common challenges you might run into when monitoring Kubernetes — and tips on how to solve them.

Challenge 1: Lack of End-to-End Visibility

One of the most common challenges of traditional monitoring in Kubernetes is the lack of end-to-end visibility into customer touchpoints and distributed applications.

As a result, IT teams do not have any insights into how end-user experience and application performance impact business KPIs — leaving no way of knowing what to fix or improve.

To overcome this, it’s important to utilize a Kubernetes monitoring solution that baselines normal performance and — through the power of machine learning — can intelligently alert IT teams when problems occur.

Challenge 2: Alert Storms

While having visibility into all application issues seems like a good thing, it can quickly spiral out of control and become a hindrance when multiple issues arise at the same time. After all, do you really need an alert every time a job completes or a new container is spun up?

Without the ability to prioritize alerts, IT teams typically have to respond to and triage the root cause of every issue — resulting in poor user experience and lost revenue.

Robust Kubernetes monitoring solutions can help you identify and troubleshoot the exact underlying issue — drilling down to the line of code, individual device, Kubernetes service and even individual container level.

Challenge 3: Troubleshooting Woes

The cost of application downtime can be crippling — a critical application failure can cost up to $1 million per hour. Time is money and IT teams shouldn’t be wasting time when it comes to detecting the root cause of an issue.

The problem is that many of today’s monitoring tools lack the ability to perform an automated root cause analysis within the Kubernetes environment, making troubleshooting a time-consuming nightmare that often results in high MTTR and reduced downtime.

To avoid this, make sure your Kubernetes monitoring solution provides visibility into application dependencies and validation of migration success by comparing user experience pre- and post-migration.

Challenge 4: Error-Prone Migration to Kubernetes

Migrating traditional applications to Kubernetes can be error-prone and time-consuming. Companies migrating existing monolithic applications to microservices deployed on Kubernetes lack visibility into the Kubernetes environment and, as a result, are unable to see real-time interactions of every microservice or traditional application.

With a solution that provides one-step install and unified monitoring through a single pane of glass, IT teams can leverage their existing skill set, processes, and tooling.

Delivering Flawless Application Performance on Kubernetes

The use of Kubernetes to transform how companies deploy and run applications in distributed, multicloud environments has risen dramatically in popularity — and shows no signs of slowing down. But for organizations running traditional or microservice-based applications on Kubernetes, the shortcomings of traditional monitoring approaches are apparent.

As a result, organizations must rethink their approach to monitoring in Kubernetes in order to simplify complex enterprise workflows, drive cost efficiencies and boost productivity. With end-to-end, unified visibility into the entire Kubernetes stack and Kubernetes-orchestrated applications, IT teams can deliver flawless application experiences — and ensure their Kubernetes investment drives better business outcomes. (Download our eBook to learn more: How to Monitor Kubernetes with Confidence.)

Feature image via Pixabay.