How we accidentally created a Cloud on our Cloud
summary
Team automates 10–20 workshops a month that historically required about 30 Kubernetes/OpenShift clusters each (peaks at 150), building a UI and Go backend that wraps IBM Cloud APIs, uses AWX (Ansible Tower) to run Ansible playbooks and a React frontend for self‑service. It supports multi‑cluster creation, round‑robin placement across datacenters (eg. create 300 clusters spread automatically), post‑provisioning installs, and plans scheduled create/delete to cut cost and remove a two‑person manual bottleneck.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 1:16Defining a cloud
- 3:04Manual Kubernetes problems
- 5:51Unsustainable cluster setup
- 8:38Bash scripts solution
- 10:04Custom UI for clusters
- 13:20Limitations of custom UI
- 15:06Ideal automated solution
- 17:55Impact and cost savings
- 19:20Building a cloud on cloud
- 21:23When to build your own
- 23:09Reinventing workflows
- 25:18Custom cloud considerations
- 27:03Starting small
- 27:44Code walkthrough
- 32:04Future steps
related talks
29mWe accidentally created a Cloud on our IBM Cloud
13mSponsored Keynote: Multi What???? Multi-cloud, Multi-cluster, Multi-tenanted K8s
15mJourney to Hybrid Multi-Cloud
06mTalking Open Source & Cloud Native
36mResilient Multi-Cloud Strategies: Harnessing Kubernetes, Cluster API, and Cell-Based Architecture
32mBeyond CAS: Why the world needs another Kubernetes Cluster Autoscaler
18mKeynote: Mind the Gap: Bridging Cloud Native Innovation with Real-World Use
31mHow we evolved our cloud infrastructure