Files

codex c9764ba73c docs(CUST-WP-0054-T05): Phase 4 sweep checkout migration evidence

Record railiance01 clone tree, host_paths registration, and tooling for
bulk clone and path registration. Update T05 workplan with Phase 4 completion.

2026-07-06 19:24:28 +02:00

3.1 KiB

Raw Blame History

T05 Phase 4 Evidence — Sweep Checkout Migration (2026-07-06)

Workplan: CUST-WP-0054-T05
Parent runbook: docs/state-hub-railiance01-cutover-plan.md
no_secret_material_recorded: true

Checklist

Step	Action	Result	Evidence
4.1	Clone root `/home/tegwick/<slug>`	pass	74 host_paths registered; 73 new clones + 4 retried
4.2	Bulk clone from git remotes	pass	`railiance01-bulk-clone-repos.sh` on railiance01
4.3	Register `host_paths`	pass	host=`239.62.205.92.host.secureserver.net` × 74
4.4	Sweep on railiance01	pass	`POST /consistency/sweep/remote-all` exit_code=0
4.5	Writeback on railiance01	pass	`the-custodian` fix: `push: pushed` from `/home/tegwick/`

Phase 4 status: complete.

Infrastructure changes

state-hub image (`main-phase4-sweep`)

Imported to railiance01 k3s (local build; promote via Forgejo CI next):

Dockerfile: git, openssh-client
STATE_HUB_SWEEP_HOSTNAME env (Kubernetes pod hostname cannot contain dots)
api/classification.py: railiance01 path fallback for allowed-values YAML
scripts/consistency_check.py: _sweep_hostname() for host_paths resolution

Helm (`deploy/railiance/apps/helm/state-hub-values.yaml`)

sweep:
  enabled: true
  hostname: 239.62.205.92.host.secureserver.net
  hostPath: /home/tegwick
  sshHostPath: /home/tegwick/.ssh

Deployment mounts host clone tree + SSH config; postStart sets git config --global safe.directory '*' (root in pod, tegwick-owned repos).

railiance01 host

SSH: ~/.ssh/config with forgejo-remote + gitea-remote (workstation id_gitea copied)
Clone tree: /home/tegwick/<slug> (77 entries incl. tooling dirs)
Failed-then-fixed slugs: ihp-railiance-probe, llm-connect, markitect-project (from markitect_project), vergabe_teilnahme (from vergabe-teilnahme)

Tooling added (the-custodian)

tools/railiance01-bulk-clone-repos.sh
tools/railiance01-register-host-paths.sh

Verification

# Single-repo fix inside pod — railiance01 path, push succeeds
kubectl -n state-hub exec deploy/state-hub -- \
  python3 /app/scripts/consistency_check.py --repo the-custodian --fix \
  --api-base http://127.0.0.1:8000
# → repo_path: /home/tegwick/the-custodian, push: pushed

# Full sweep canary
curl -s -X POST http://127.0.0.1:8000/consistency/sweep/remote-all \
  -H "Content-Type: application/json" \
  -d '{"max_seconds":300,"source":"phase4-canary"}'
# → exit_code: 0, 8 repos processed (budget-limited first pass), skipped_missing: 0

.custodian-brief.md on railiance01 (17:08 UTC) is newer than workstation (15:27 UTC).

Workstation host_paths[bnt-lap001] retained for dev; production sweep resolves host_paths[239.62.205.92.host.secureserver.net].

Follow-ups (Phase 5)

Promote main-phase4-sweep image via Forgejo CI (replace hand-imported tag)
Allow longer first-pass sweep budget or multiple cron cycles to reconcile all 74 repos
activity-core scheduled sweeps should recover (were failing with exit_code=1 pre-Phase-4)

Next: Phase 5

48–72h stabilization before coulombcore hub teardown.

3.1 KiB Raw Blame History Unescape Escape