We are excited to announce deeper integration between the Wavefront cloud monitoring platform and the Pivotal Container Service (PKS) starting with the PKS release 1.1. Invoking monitoring is now embedded into PKS, empowering SREs, DevOps engineers and developers to start sending Kubernetes metrics to the Wavefront cloud-native metrics monitoring and analytics platform. This turn-key integration delivers deep visibility into the health of Kubernetes and containers in production.
Starting Wavefront from PKS is Easy
Kubernetes, a major component of PKS, is built to support dynamic scaling of clusters, nodes, pods, and containers to meet the demands of containerized applications. While Kubernetes does a fine job orchestrating the many underlying components, SREs need to deal with rising complexity that can impair visibility into the different Kubernetes layers.
Now with this deep integration, we simplify Kubernetes observability and enforce it from the start, a standard for ensuring proper scaling. When you start adding new Kubernetes clusters, PKS 1.1 sends performance and health metrics to the SaaS-based Wavefront platform on-demand. To start seeing important KPIs in charts and pre-packaged dashboards, you just need to point to your Wavefront instance or Wavefront free trial by adding the Wavefront URL, the Wavefront token and then specifying the alert target per the illustration below.
This deeper integration enables SREs and DevOps teams running Kubernetes to create pre-defined Wavefront alerts for important KPIs. You can easily modify or enforce desired alerts across all Kubernetes clusters, helping teams to simplify resource management or to quickly resolve any infrastructure bottlenecks.
An example of Wavefront’s analytics-driven alerts that can be easily customized includes alerting on the health status of the node, pod, or container such as CPU load, memory usage or (a file system’s) node usage. Proactive alerting simplifies management of Kubernetes clusters and containerized applications. This helps Kubernetes operators to be on top of their environment at all times no matter how dynamic the rate of change.
Enhanced Wavefront PKS Integration with New Kube-state Metrics and Dashboards for Dynamic Kubernetes Insight
In addition, we have enhanced Wavefront’s PKS integration with pre-packaged dashboards built on top of kube-state metrics in addition to existing cAdvisor and Heapster powered metrics and dashboards, which have already been providing detailed resource utilization insight essential for Kubernetes admins.
Kube-state is a simple service that listens to the Kubernetes API server and generates metrics. Because of real-time efficiency of Wavefront, any changes in kube-state metrics are immediately reflected in Wavefront, a monitoring platform designed to handle massive scale like tens of thousands running containers (see the Mixpanel customer case study in our Container Monitoring eBook) A sample Wavefront dashboard for PKS is included below:
Check out the documentation on Github for the full list of kube-state metrics that can be collected with our integration. With kube-state metrics analyzed by the Wavefront platform, you can get comprehensive insight into the health of all Kubernetes components. The Wavefront PKS integration features over 60 pre-built charts for deep views into the health status and resource loads for:
- Nodes, namespaces, pods, pod containers
- Services and replica sets
Many Wavefront customers start with our pre-packaged dashboards, like the ones above. But Wavefront analytics can show so much more. Once PKS Kubernetes metrics are in Wavefront, they can be correlated with containerized applications and microservices performance metrics or other infrastructure metrics. Wavefront also delivers an easy way to instrument your applications using open-source – statsD, Micrometer, DropWizard, etc. – to help SREs troubleshoot and alleviate any containerized application availability or performance issues.
Wavefront by VMware is a one-stop shop for your cloud-native application and Kubernetes environment. To learn more, check these additional resources:
- Docker visibility
- Deep PKS and Kubernetes insights
- Custom Cloud Applications visibility – using StatsD, DropWizard or Micrometer
- Multi-cloud infrastructure insight for your Kubernetes hosts, no matter where your hosts are running – AWS, Google Cloud or Azure.
About the AuthorFollow on Twitter Follow on Linkedin More Content by Stela Udovicic