Upgrade to Pro — share decks privately, control downloads, hide ads and more …

A guide to observability for Java apps in Kuber...

A guide to observability for Java apps in Kubernetes

Matthias Haeussler

October 17, 2021
Tweet

More Decks by Matthias Haeussler

Other Decks in Technology

Transcript

  1. From k9s to OpenTelemetry: A guide to observe (Java apps)

    in Kubernetes Matthias Häußler Chief Technologist, Novatec
  2. 2

  3. Agenda 1. Intro 2. Kubernetes API 3. Prometheus & Grafana

    4. Service Mesh − Sidecar-based − eBPF-based 5. Application-based 5
  4. 26

  5. 27

  6. 41

  7. ▪ Least intrusive ▪ No change to the cluster or

    any apps required ▪ Helpful for understanding Kubernetes and getting an overview ▪ Little network insights ▪ No service connections visible Characteristics 44
  8. ▪ Extends Kubernetes for limitations in network traffic awareness and

    shaping capabilities ▪ Concept of injecting sidecar proxy into each pod to have control and insight into entire network flow ▪ No change to application or application container required ▪ No application-level metrics Characteristics 61
  9. 70

  10. ▪ Injection of proxy component on node instead of pod

    level ▪ Linux low-level functionality leveraged for Kubernetes observability ▪ Fast growing technology in CNCF landscape ▪ Application and application container untouched ▪ Cluster needs to be configured once Characteristics 72
  11. 93

  12. ▪ Provides application level metrics and enables root cause analysis

    ▪ Changes to application or container required ▪ Agents specific to programming language and frameworks Characteristics 104