Continuous Observability: Shedding Light on CI/CD Pipelines

[@sergiykulanov5837]

A big thank you to this podcast for shedding light on such practices. We've harnessed the power of CNCF's open-source tools to enhance our CI/CD pipeline's observability, explicitly using a "Tekton (CloudEvents) --> FluentBit --> OpenSearch" scheme. This setup lets us capture many metrics, shedding light on the development process. It has transformed our CI/CD analytics, helping us make better decisions and refine workflow processes. I highly recommend exploring this approach for anyone looking to boost their pipeline's transparency and efficiency using open-source CNCF tools.

[@horovits]

thanks for the feedback, glad you liked the episode! Which organization are you at, using this stack? BTW you mentioned using CNCF tools, but just for accuracy - from the stack you mentioned only FluentBit is on the CNCF. Anyway all are solide open source projects 🙂

[@sergiykulanov5837]

@@horovits At KubeRocketCI, we use Tekton as our workflow engine. Honestly, Tekton doesn’t provide out-of-the-box rich pipeline analytics. The only build-in metrics are those related to controllers and basic ones. Solutions like Prometheus and OpenTelemetry can be used to collect these metrics. But we’ve adopted cloud events. You can build multi-dimensional reports by adding Kubernetes labels or any other metadata. For example, what is the average build time for a Java application, the slowest pipeline step, etc.? Another CNCF project, External Secret Operator, has been widely adopted to make real GitOps with Argo CD.