OSDC 2017: Automating Kubernetes Cluster Operations with Operators by Timo Derstappen
0 likes21 views
The document discusses automating Kubernetes cluster operations using operators, specifically through the 'Giantnetes' platform. It emphasizes the importance of managing clusters efficiently, deploying micro operators for specific tasks, and utilizing open-source tools while maintaining security and reproducibility. The talk outlines the motivations, architecture, and functionalities of the proposed automation approach, targeting both bare metal and cloud environments.
OSDC 2017: Automating Kubernetes Cluster Operations with Operators by Timo Derstappen
- 1. Automating Kubernetes Cluster Operations with Operators by Timo Derstappen
- 2. Running Kubernetes in Kubernetes
- 3. What this talk is about Provisioning Kubernetes Clusters Advanced usage of Kubernetes
- 4. What this talk is not about Intro to Kubernetes How to build cloud native applications
- 5. Overview 1. Background / Motivation 2. Architecture 3. Operators 4. Outlook
- 6. Background We manage Kubernetes clusters for customers 24/7. Both on bare metal and in the cloud.
- 7. Kubernetes as a Service (KaaS) Own DC in Frankfurt, Germany AWS (Azure and Google next) Full access to clusters
- 8. On-premises KaaS We also manage Kubernetes in the DCs or AWS accounts of the customers itself.
- 9. Open Source Our tooling is open-source and we are working with the community to improve kubernetes operations.
- 10. Multiple Clusters "Soft" multi-tenancy not enough in enterprise context Different service classifications Different environments: separation of dev, test, prod Test tooling on new K8s versions
- 11. People must come to things in their own time, in their own way, for their own reasons, or they never truly come at all. - Dee Hock
- 12. RBAC, NetworkPolicies FTW We encourage to use these features and help with integration. Small clusters make little sense Support while processes need to be adapted Kubernetes matures - trust over time
- 14. Motivation Rather obvious. If you provide others with Kubernetes you would like to use the same concepts within your product as well.
- 15. De nition G8s: Giantnetes K8s: Kubernetes == Guest cluster
- 17. Overview 1. Background / Motivation 2. Architecture 3. Operators 4. Outlook
- 21. Why are we doing this again? Fully automate a powerful container platform Take away complexity of distributed applications from the developers
- 22. Immutability CoreOS (Hosts and Guests) Continuous pipeline (Container build, registry, kubernetes)
- 23. Reproducable builds and deployments at any point in time
- 24. Every manual change is ephemeral
- 25. Mean time to recovery!
- 26. Networking Flannel/VPC network between guest cluster nodes Calico between containers Network policy controller
- 31. Certi cates Each cluster has its own root ca Using PKI backends in Vault Certificates are rotated every day
- 32. Certi cates #2 Kubernetes components use mutual TLS Client certificates are used for RBAC
- 33. Overview 1. Background / Motivation 2. Architecture 3. Operators 4. Outlook
- 34. Iterations of the platform 1. K8s clusters via systemd units over fleet 2. K8s manifests to create K8s clusters used as templates 3. Writing operators
- 35. Operators Codify all operational tasks Manage desired state
- 36. Third-Party Resources Easily extend the Kubernetes API
- 37. KVM operator Create k8s clusters on bare-metal via KVM based on a TPR
- 38. AWS operator Create k8s clusters on AWS based on a TPR
- 45. The Cluster TPR de nes Kubernetes configuration/version Ingress configuration/version Network configuration Master/Node configurations
- 46. The Cert TPR de nes TTL Common Name SANs Cluster / Component
- 47. Overview 1. Background / Motivation 2. Architecture 3. Operators 4. Outlook
- 48. Micro Operators Operators should have single responsibilities. Keep them simple and maintainable
- 49. Operatorkit Our services are based upon microkit. We would like to reduce boilerplate in our operators and collect them in a library called operatorkit.
- 50. Self-hosted Having Giantnetes and all guest clusters running self-hosted will further ease the lifecycle management of the clusters.