At Banzai Cloud we are passionate about observability, and we expend a great amount of effort to make sure we always know what’s happening inside our Kubernetes clusters. All clusters provisioned with Pipeline - our multi- and hybrid-cloud container management platform - are provided with, and rely upon, each of the three pillars of observability: federated monitoring, centralized log collection and traces. In order to automate log collection on Kubernetes, we opensourced a logging-operator built on the Fluent ecosystem.

Whenever we talk about Kubernetes monitoring and the topic of Prometheus comes up, so does Grafana; the two pretty much go hand in hand. Prometheus query language This article will not serve as an introduction to the powerful Prometheus query language. If you’re not comfortable with Prometheus, now would be a good time to familiarize yourself with the basics of querying and monitoring best practices with Prometheus on Kubernetes. We will only use simple queries with filters and labels in this post.

Two core features of the Pipeline platform are advanced security and observability for all Kubernetes deployments. We make sure that all secrets are securely stored, transported, images scanned, deployments monitored, and logs centrally collected. As you might expect from us, we use the industry standard in security components, such as Vault, Prometheus, Grafana, Fluent and more. This post is about monitoring Vault with Prometheus (on Kubernetes) and displaying metrics on Grafana.