Mark/infrastructure
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
Meta repo for infrastructure tracking and sysadmin scripting
56 commits 1 branch 0 tags 184 KiB
Find a file
Exact
hermes 65950c7108
All checks were successful
CI / lint (push) Successful in 33s
Details
ci: re-run via runner pickup verification
2026-09-26 20:06:48 -07:00
.forgejo/workflows ci: fix markdown lint baseline failures 2026-09-12 01:17:26 -07:00
.githooks fix: shell script bugs found in bughunt 2026-08-11 20:23:03 -07:00
docs fix #80: supervise the user-level Forgejo runner alongside per-repo runners 2026-09-14 10:15:05 -07:00
issues ci: fix markdown lint baseline failures 2026-09-12 01:17:26 -07:00
scripts fix #79: add per-repo runner bootstrap wrapper and tests 2026-09-14 10:18:23 -07:00
tests fix #79: add per-repo runner bootstrap wrapper and tests 2026-09-14 10:18:23 -07:00
.gitignore Untrack .kilo/.gitignore; drop allowlist exception 2026-09-14 10:05:59 -07:00
.markdownlint.json ci: fix markdown lint baseline failures 2026-09-12 01:17:26 -07:00
package.json fix #58: add diagnostic tooling for CI run failures 2026-06-16 21:09:44 -07:00
pyproject.toml fix #58: add diagnostic tooling for CI run failures 2026-06-16 21:09:44 -07:00
README.md fix #81: supervise per-repo Forgejo runner containers 2026-09-14 10:01:56 -07:00

README.md

Infrastructure Meta Repo

Tracking issues, scripts, and documentation for the tailnet infrastructure.

Purpose

  • Issue tracking: System-level bugs (Hindsight duplicates, CUDA mismatches, provider rotation)
  • Sysadmin scripts: Health checks, cleanup utilities, backup helpers, environment setup, runner supervision
  • Topology docs: Tailnet service mappings, host assignments

Structure

infrastructure/
├── README.md          # This file
├── issues/            # Issue templates and tracking
├── scripts/           # Re-runnable sysadmin scripts
└── docs/              # Infrastructure topology and diagrams

Quick Links

Forgejo runner supervision

Per-repo runner containers created by scripts/forgejo-runner-per-repo.sh have no built-in supervisor. Use scripts/forgejo-runner-watchdog.sh to keep them alive:

bash scripts/forgejo-runner-watchdog.sh once           # bring every registered runner up
bash scripts/forgejo-runner-watchdog.sh list           # show registered runners + container state
bash scripts/forgejo-runner-watchdog.sh install-cron   # cron `once` every minute

For hosts with systemd user services, run python3 scripts/systemd/forgejo-runner-generator.py after registering or removing a runner so that one forgejo-runner@<repo>.service exists per registered repo; systemctl --user enable --now forgejo-runner@<repo>.service then keeps each container alive.