Cloudutsuk Platform 10
A self-hosted control plane on an NVIDIA DGX Spark: one SSO-protected portal for every lab, cloud resource and its cost, with central OpenTelemetry monitoring, local CI/CD and GitOps, a RAG chatbot on a local LLM and an alert-driven SRE agent.
Diagrams
Every diagram of the project in one place, each linking to the section that explains it.
Components, where they run and how a request moves through them.
- System architecture (interactive)
Networks and subnets, protocols and ports, TLS points, DNS and the way in and out.
- Network layout (subnets, ports, TLS, DNS) (animated)
How data is collected, transformed, stored and served, end to end.
Not drawn yet; belongs in Data and Integration
One important request, call by call, including retries and failure paths.
Not drawn yet; belongs in Technical Architecture
How people and services authenticate, and where authorization is decided.
Not drawn yet; belongs in Security and Privacy
Trust boundaries, entry points and the threats considered at each one.
Not drawn yet; belongs in Security and Privacy
Entities, relationships and ownership (ER diagram).
- Platform data model (ER) and ownership (mermaid)
Environments, clusters and hosts, and what is deployed where.
Not drawn yet; belongs in Infrastructure and Environments
From commit to production, with the gates and the rollback path.
- Commit to deploy with security gates (animated)
How metrics, logs and traces flow, and how an alert reaches a person.
Not drawn yet; belongs in Operations and Observability
Detection, triage, mitigation and follow-up for a production incident.
Not drawn yet; belongs in Operations and Observability
The main path a user takes, screen by screen.
Not drawn yet; belongs in User Experience and Design
Discover and Define
Why the project exists, who it serves and how success is measured.
What the system must do, how well it must do it, and what is in the first release.
Roadmap, MVP, milestones, budget and the rules for changing the plan.
Design
How users move through the product and what they see on every screen.
- Information architecture
- User flows
- Wireframes and prototypes
- Visual design and design system
- Interaction patterns
- Responsive behavior
- Accessibility design
- Usability testing
- Content and messaging
Not written yet
Components, technology choices, integrations, hosting and the decisions behind them.
- System context and component architecture
- Technology and framework selection
- API and integration architecture
- Data architecture and data models
- Identity and access architecture
- Hosting and deployment model
- Scalability and performance strategy
- Resilience and disaster recovery
- Architectural decision records
Threats, controls, secrets, privacy obligations and how incidents are handled.
- Threat modeling
- Authentication and authorization
- Secrets and key management
- Encryption requirements
- Network security
- Input validation and secure coding
- Vulnerability and dependency management
- Privacy and data-protection requirements
- Audit logging
- Security testing and incident response
Not written yet
Who owns the data, how it moves, how it is kept, and how it is migrated.
- Data ownership and classification
- Data retention and deletion
- Database schema and migrations
- Data validation and quality
- Import and export requirements
- External systems and third-party integrations
- Event and messaging contracts
- Backup and restore strategy
- Data migration and reconciliation
Not written yet
Build and Test
Repository, standards, local setup and the workflow every change goes through.
- Repository and branching strategy
- Coding standards
- Project structure
- Local development setup
- Environment configuration
- Dependency management
- Feature flags
- Documentation standards
- Development workflow and code review process
Not written yet
How the work is sliced into iterations and built, reviewed and documented.
- Sprint or iteration planning
- Feature development
- API and UI implementation
- Database and infrastructure implementation
- Integration development
- Technical debt tracking
- Code reviews
- Developer documentation
Not written yet
The test strategy, from unit tests to user acceptance, and the data and environments it needs.
- Unit testing
- Integration testing
- Contract testing
- End-to-end testing
- Regression testing
- Performance and load testing
- Security testing
- Accessibility testing
- Compatibility testing
- User acceptance testing
- Test data and test environment management
Not written yet
Every environment from dev to production, built as code and kept in step.
Ship
The pipeline from commit to production, with gates, versioning and rollback.
Readiness, runbooks, cutover and the decision to launch.
Run and Support
Monitoring, logs, traces, alerts and the targets the service is held to.
The documents, guides and training that let other people run and use the system.
- Product documentation
- User guides
- API documentation
- Architecture documentation
- Deployment instructions
- Operations runbooks
- Troubleshooting guides
- Support procedures
- Onboarding documentation
- Training materials
Not written yet
Who supports it, to what standard, and how it is patched, upgraded and recovered.
- Support model and ownership
- Service-level objectives
- Incident response
- Bug triage
- Patch and upgrade management
- Dependency updates
- Security remediation
- Performance optimization
- Backup verification
- Disaster recovery testing
- Technical debt management
Not written yet
Govern and Improve
Who decides, what must be audited, and the evidence that proves it.
- Project governance
- Decision-making and approval authorities
- Audit requirements
- Access reviews
- License management
- Compliance reporting
- Data governance
- Vendor and third-party management
- Policy enforcement
- Evidence collection
Not written yet
Learning from users, incidents and costs, and planning the end of life.
- Product analytics
- User feedback
- Retrospectives
- Reliability reviews
- Cost optimization
- Security reviews
- Architecture reviews
- Roadmap refinement
- Decommissioning and end-of-life planning
Not written yet
Architecture Decisions
Every significant decision with its context, the options that lost and the consequences, synced from homelab-gitops/docs/adr.
47 accepted 1 proposed 1 superseded or deprecated
- ADR-0020 Platform SLOs and burn-rate alerting
Lessons Learned
Problems hit while building it, with the symptom, the cause, the fix and what to do next time. Newest first.
- L073 A model’s context ceiling and download size decide whether it start...
- L072 Hermes Agent’s image wants root under s6; on a restricted cluster i...
- L071 Upstream LGTM dashboards looked empty because they’re written for m...
- L070 Grafana showed no contact points although every alert reached Slack
- L069 gpt-oss loaded fine, then failed every request on a missing tokeniz...
- L068 DNS over UDP filled core’s load balancer, and took its HTTPS and AP...
Interview Prep
The project as you would discuss it in an interview, answered from architect, developer, DevOps and SRE perspectives, with links back to the playbook.
- Project Overview and Your Role (4)
- System Design and Architecture (5)
- Technology Choices (3)
- Frontend Architecture (4)
- Backend and API Design (4)
- Database and Data Modeling (4)
- Frontend, Backend and Database Communication (5)
- Integrations and Background Processing (3)
- Security and Identity (4)
- Performance and Scalability (3)
- Reliability and Resilience (3)
- Testing and Quality (3)
- Infrastructure and Environments (3)
- CI/CD and Deployment (4)
- Monitoring and Observability (3)
- Cost, Maintenance and Support (2)
- Leadership, Stakeholders and Governance (3)
- Lessons Learned and Trade-offs (4)
Not written yet