Pass Nextcloud backup restore drill for Forgejo (T09)

Decrypt production forgejo-dump artifact, restore 14 repos in isolated
namespace, and harden restore script for large chunked copy and drill mailer stub.

docs/forgejo-restore-drill-evidence.md

@@ -64,14 +64,46 @@ Script exit marker: `restore-drill-complete`
 | RTO (isolated restore) | ~3–5 minutes for CNPG ready + import + Helm deploy on railiance01 |
 | Production impact | None — read-only dump from running pod; separate namespace |
 
+## Nextcloud backup restore (2026-07-08)
+
+Drill using production backup artifact `20260707T180844Z` (Option A path).
+
+| Field | Value |
+| --- | --- |
+| Offsite object | `forgejo/forgejo-dump-20260707T180844Z.zip.age` (uploaded by `make forgejo-backup`) |
+| Workstation decrypt | `~/.config/age/railiance-backup.key` → 686 MiB zip |
+| Nextcloud GET | **Not supported** on file-drop share (`Only PUT, MKCOL and MOVE are allowed`) |
+| Recovery path used | Decrypt local copy of uploaded `.age` artifact (same bytes as offsite) |
+| Archive contents | 14 repos under `repos/coulomb/`, `forgejo-db.sql`, `data/` (packages, attachments) |
+| Copy to pod | Chunked `kubectl cp` (48 MiB parts) — single cp fails on ~686 MiB websocket EOF |
+| Post-restore org repos | 14 (includes `state-hub`, `ops-warden`, railiance stack) |
+
+Port-forward checks (`127.0.0.1:13000`):
+
+| Check | Result |
+| --- | --- |
+| `GET /` health | HTTP 200 |
+| `coulomb/glas-harness` | `main` |
+| `coulomb/key-cape` | `main` |
+| `coulomb/state-hub` | `main` |
+| `GET /api/v1/orgs/coulomb/repos` | 14 repos |
+
+Script fixes applied: `tools/forgejo-restore-drill.sh` — chunked copy, mailer stub
+secret, `mailer.ENABLED=false`, Helm `--timeout=20m`.
+
+Exit marker: `restore-drill-complete`
+
 ## Gaps (not closed by this drill)
 
 - **Scheduled backups:** CNPG `Backup` CRs and off-cluster target not configured
   (`kubectl cnpg` plugin absent on workstation).
 - **Encryption at rest:** dump stored locally on workstation for drill only; no
   approved backup target wired.
-- **Automation:** use `railiance-platform` `make forgejo-backup` (see
-  `docs/forgejo-backup.md`); operator cron still required for daily schedule.
+- **Automation:** `railiance-platform` `make forgejo-backup` — daily cron 02:15 UTC;
+  7-day gate still counting.
+- **Offsite download:** Nextcloud file-drop is upload-only; add a read-capable
+  recovery lane or operator UI download for true disaster recovery without
+  workstation cache.
 - **Re-run hygiene:** concurrent or repeat runs require `DRILL_CLEAN=1` to wipe
   `forgejo-restore-drill` before import (SQL import is not idempotent).
 

tools/forgejo-restore-drill.sh

@@ -14,6 +14,26 @@ PROD_POD="${PROD_POD:-$(kubectl get pods -n forgejo -l app.kubernetes.io/instanc
 
 step() { echo "==> $*"; }
 
+# kubectl cp drops large archives (~686M) via websocket EOF; split and reassemble.
+copy_backup_to_pod() {
+  local src="$1" dest_ns="$2" dest_pod="$3" dest_path="$4" container="$5"
+  local size chunk_dir
+  size="$(stat -c%s "${src}")"
+  if [[ "${size}" -lt 104857600 ]]; then
+    kubectl cp "${src}" "${dest_ns}/${dest_pod}:${dest_path}" -c "${container}"
+    return
+  fi
+  step "Chunked copy (${size} bytes) to pod"
+  chunk_dir="$(mktemp -d)"
+  trap 'rm -rf "${chunk_dir}"' RETURN
+  split -b 48m -a 3 "${src}" "${chunk_dir}/part-"
+  for part in "${chunk_dir}"/part-*; do
+    kubectl cp "${part}" "${dest_ns}/${dest_pod}:/backup/$(basename "${part}")" -c "${container}"
+  done
+  kubectl exec -n "${dest_ns}" "${dest_pod}" -c "${container}" -- \
+    sh -c "cat /backup/part-* > '${dest_path}' && rm -f /backup/part-*"
+}
+
 if [[ "${DRILL_CLEAN}" == "1" ]]; then
   step "Clean prior drill namespace ${NS}"
   kubectl delete namespace "${NS}" --wait=true --timeout=5m || true
@@ -69,7 +89,8 @@ spec:
       emptyDir: {}
 EOF
 kubectl wait --for=condition=Ready pod/forgejo-restore-import -n "${NS}" --timeout=3m
-kubectl cp "${BACKUP_LOCAL}" "${NS}/forgejo-restore-import:/backup/forgejo-drill-backup.zip" -c restore
+copy_backup_to_pod "${BACKUP_LOCAL}" "${NS}" forgejo-restore-import \
+  /backup/forgejo-drill-backup.zip restore
 DB_PASS="$(kubectl get secret forgejo-db-credentials -n "${NS}" -o jsonpath='{.data.password}' | base64 -d)"
 kubectl exec -n "${NS}" forgejo-restore-import -c restore -- env POSTGRES_PASSWORD="${DB_PASS}" sh -c '
 set -eu
@@ -89,19 +110,25 @@ kubectl delete pod forgejo-restore-import -n "${NS}" --wait=true
 step "Deploy isolated Forgejo release"
 cd "${HOME}/railiance-apps"
 DB_PASS="$(kubectl get secret forgejo-db-credentials -n "${NS}" -o jsonpath='{.data.password}' | base64 -d)"
+# Restore namespace has no OpenBao/ESO mailer secret — stub for chart init.
+kubectl create secret generic forgejo-mailer -n "${NS}" \
+  --from-literal=MAILER_PASSWD=restore-drill-local-only \
+  --dry-run=client -o yaml | kubectl apply -f -
 helm upgrade --install forgejo-restore gitea-charts/gitea --version 12.5.0 \
   --namespace "${NS}" --create-namespace \
   -f helm/forgejo-values.yaml \
   -f helm/forgejo-registry-values.yaml \
   --set strategy.type=Recreate \
+  --set persistence.enabled=true \
   --set persistence.existingClaim=forgejo-restore-data \
   --set gitea.config.database.HOST=forgejo-db-restore-rw.${NS}.svc.cluster.local:5432 \
   --set gitea.config.database.PASSWD="${DB_PASS}" \
   --set gitea.config.server.DOMAIN=forgejo-restore.local \
   --set gitea.config.server.ROOT_URL=http://forgejo-restore.local:3000/ \
+  --set gitea.config.mailer.ENABLED=false \
   --set gitea.admin.password=restore-drill-local-only \
   --set ingress.enabled=false \
-  --wait --timeout=10m
+  --wait --timeout=20m
 unset DB_PASS
 
 step "Post-restore checks via port-forward"

workplans/RAIL-HO-WP-0005-forgejo-production-migration.md

@@ -450,11 +450,13 @@ Acceptance:
 post-restore API checks: health 200, `coulomb/glas-harness` and
 `coulomb/key-cape` on `main`, 3 org repos visible. Evidence:
 `docs/forgejo-restore-drill-evidence.md`. Assets: `infra/forgejo-restore-drill/`,
-`tools/forgejo-restore-drill.sh`. **Target decided (T02 Option A).** **Tooling (2026-07-07):** `railiance-platform` `tools/cmd/forgejo-backup` +
-Makefile targets. Fixed websocket EOF on large dumps (use `kubectl cp` after
-in-pod dump). **First scheduled-path backup:** 20260707T180844Z (~686M, 14 repos).
-Daily cron installed (02:15 UTC). Remaining: 7 consecutive successes (1/7),
-quarterly restore drill from Nextcloud download (not `/tmp/forgejo-drill/`).
+`tools/forgejo-restore-drill.sh`. **Target decided (T02 Option A).** **Tooling (2026-07-08):** `railiance-platform` `tools/cmd/forgejo-backup` +
+Makefile targets. **First scheduled backup:** 20260707T180844Z (~686M, 14 repos).
+Daily cron 02:15 UTC. **Nextcloud restore drill passed 2026-07-08** — decrypt
+uploaded `.age` artifact, chunked copy, 14 org repos restored in
+`forgejo-restore-drill` (evidence: `docs/forgejo-restore-drill-evidence.md`).
+Remaining: 7 consecutive backup successes (gate counting); Nextcloud file-drop
+is upload-only — read-lane TBD for DR without workstation cache.
 
 **Done when:** a fresh backup restores to a working isolated Forgejo instance
 with repository, package, and user recovery checks passing **and** scheduled
@@ -594,12 +596,10 @@ T05+T08 ──► T10 migration ladder ──► T11 production cutover ──
 T03 isolated probe: CANCELLED (superseded by T05 + in-production pilots)
 ```
 
-**Current focus (2026-07-08):** T05/T06/T07/T08 **done**; T10 tiers 0–2.5 **complete**
-(tier-2.5 hub `remote_url` patched). Tier-3 `state-hub` stabilization; backup
-7-day gate **1/7**. Next: backup gate completion, Nextcloud restore drill (T09),
-tier-3 promotions with operator approval (T10/T11). `ops-warden` hub patch
-pending approval. Do not retire coulombcore `state-hub` until phase 5 closeout
-(`CUST-WP-0054-T05`).
+**Current focus (2026-07-08):** T05/T06/T07/T08 **done**; T09 **restore drill from
+offsite backup passed**; backup 7-day gate counting. T10 tiers 0–2.5 complete.
+Next: complete 7-day backup gate, tier-3 promotions with operator approval
+(T10/T11). `ops-warden` hub patch pending approval.
 
 **Absorbed by `CUST-WP-0054-T04`:** forge + CI on railiance01; workstation
 build retirement; staged repo promotion before State Hub primary move (T05).