We are thrilled to announce the expansion of the Kubernetes observability experience in Wavefront with the next generation release of the Wavefront Collector for Kubernetes. This release has a significant number of changes that give Kubernetes operators, SREs and developers, richer visibility into the health, state, and performance of their Kubernetes environment. In addition, the auto-discovery capability in the Wavefront Collector for Kubernetes can identify and monitor your Kubernetes workloads. To get more details on Wavefront’s new Kubernetes capabilities check out this demo.
Expanded Deployment Options
The Wavefront Collector can now be installed in either the daemonset mode or the simple mode. The daemonset mode gives you resiliency and scalability in a Kubernetes environment. The collector includes leader election for enhanced fault-tolerance against pod failures while monitoring cluster level resources.
In talking to a number of customers, we’ve recognized the need to have a solution for test/dev environments and have retained the simple mode in which you can deploy one Wavefront Collector per Kubernetes cluster.
This release of the Wavefront Collector expands support for different data sources. We now collect the core Kubernetes metrics, host level metrics, and systemd metrics, and we also scrape Prometheus endpoints to collect data from the API Server, NGINX and etcd.
Expanded Auto-Discovery Capabilities
Auto-discovery in Wavefront is enabled via annotations or discovery rules. When enabled, the Wavefront Collector can auto-discover pods and services that expose metrics and can start collecting metrics from these sources.
While the annotations-based auto discovery primarily supports Prometheus scrape targets, the rules-based discovery supports a lot more. For starters the rules-based discovery can auto discover Kubernetes resources based on container images, namespaces, Prometheus endpoints and metrics via the Telegraf plugin. In addition, it can also decorate or transform the metrics via prefixes tags and filters.
With built in Telegraf plugin support for auto-discovery, the Wavefront Collector can dynamically identify any Telegraf supported workload and automatically start collecting metrics from it. Customers no longer have to worry about configuring Wavefront monitoring for new workloads.
Tools for At-Scale Deployments
Wavefront is designed to run at scale, and scale takes on a whole new meaning in the world of Kubernetes and containers.
To ensure that customers have the right levers to run at scale, we support a broad set of filters in the Wavefront Collector. These filters allow you to specify options such as metric names, tags and glob patterns. We also have black and white lists to ensure that customers have the maximum flexibility when identifying the data that is most useful to them.
In addition, the Wavefront Collector emits a set of comprehensive metrics to help you understand the health and performance of the collector itself. These include the number of discovery rules, leader election errors, runtime memory stats, scrape latency and more. The collector also shows detailed metrics that show the amount of data collected and filtered from each source.
The New Wavefront Kubernetes Dashboard
These new enhancements make Kubernetes monitoring a breeze. Even better, we’ve expanded and updated the Wavefront dashboards for Kubernetes monitoring so the Kubernetes experience in Wavefront provides instant value to you from day 1.
Try out our new Kubernetes capabilities! If you don’t have an account sign up for a free trial, go to Integrations, find Kubernetes and follow the instructions. You can also visit our open source GitHub page for the Wavefront Collector for Kubernetes so you can see what’s going on under the covers.Get Started with Wavefront Follow @_karthiknarayan Follow @WavefrontHQ
About the AuthorFollow on Twitter More Content by Karthik Narayan