Spinnaker Observability in Practice with Kubernetes.
summary
An open-source Armory Observability Plugin ships to reduce observability vendor costs by filtering Spinnaker metric cardinality and exporting to Prometheus and New Relic (does not add tracing or logging). Real deployments show surprisingly high cardinality — about 70,000 unique metrics from labels — so the plugin adds a /prometheus management endpoint, integrates with Prometheus operator and Grafana dashboards, and successfully centralizes useful Spinnaker metrics while warning about exposing management endpoints publicly.
AI-written summary · about
chapters (8)
- 0:00Start
- 2:16What can be done today...
- 6:54Metrics
- 8:29Logging
- 9:56Tracing
- 10:53What we get with Observability plugin
- 12:37DEMO TIME
- 13:02Enable the plugin
related talks
34mMonitoring Spinnaker with Prometheus Operator on GKE
32mUnderstanding Observability in Kubernetes
19mSpinnaker Deployments on Kubernetes from a Security Perspective
01h 04mOpening Keynote [Full]
17mIntroduction to Spinnaker on Kubernetes
32mTesting Kubernetes apps' observability end-to-end
10mSpinnaker @ Scale for Snap’s Multi Cloud Service Mesh
11mKubernetes Observability