jarvis-at-skic
464ea75730
Build and Deploy Documentation / build (push) Failing after 9s
Styling: - Tailwind CSS v4 (preflight:false) integrated via PostCSS plugin - shadcn/ui New York style with --radius:0 (sharp corners everywhere) - Zero border-radius on ALL elements: buttons, code blocks, cards, admonitions, badges - Navbar FORCE dark (#0a0a0a) in both light/dark modes via !important - ThemeSynchronizer: bridges Docusaurus data-theme → shadcn .dark class - Active sidebar item gets left border accent instead of rounded bg React UI components (src/components/ui/index.tsx): - Badge (6 variants: wip, stable, planned, success, warning, outline) - Card + CardHeader + CardTitle + CardContent - Callout (note/info/success/warning/danger) - sharp, left-border style - PropertiesTable - monospace typed API reference tables - StatusIndicator - operational/degraded/outage/maintenance Navigation restructure: - Introduction → Platform → Products → Engineering → Runbooks - Old 'applications/' removed, moved to products/stock-market-pro/ - New Platform section: infrastructure overview + CI/CD pipeline docs - New Engineering section: monorepo structure, Go stack, conventions - New Runbooks section: incident severity, common procedures Implementation guidelines (engineering/guidelines.md): - Monorepo layout with go.work workspace - Go as primary language (why + standard library table) - Python as sidecar for data/ML workloads - Code conventions: gofmt, error wrapping, logging, testing - Git conventions: branch naming, conventional commits - Makefile targets per app - How to add a new application (step-by-step)
1.2 KiB
1.2 KiB
sidebar_position
| sidebar_position |
|---|
| 1 |
Runbooks
Operational procedures for SKIC Playground services.
:::note Runbooks are living documents. Update them when procedures change or new issues are discovered. :::
Incident Severity
| Level | Definition | Response Time |
|---|---|---|
| P0 | Platform down / all services unreachable | Immediate |
| P1 | Single critical service down | < 15 min |
| P2 | Degraded performance / non-critical service down | < 1 hour |
| P3 | Minor issue / cosmetic | Next working session |
Common Procedures
Restart a container
# SSH to TrueNAS
ssh root@lego-cloud.eu
# List running containers
docker ps
# Restart
docker restart <container-name>
# Check logs
docker logs -f --tail 100 <container-name>
Check Gitea Actions runner status
# On the runner host
systemctl status gitea-runner
# Restart if needed
systemctl restart gitea-runner
Force re-deploy from latest image
docker pull <registry>/<image>:<tag>
docker stop <container>
docker rm <container>
docker run -d --name <container> ... <image>:<tag>
Application-specific runbooks
Individual application runbooks will be added here as applications go into production.