Add homelab recovery plan command

This commit is contained in:
juvdiaz 2026-06-29 17:20:40 -06:00
parent 168d61a820
commit 00f8801fd7
5 changed files with 117 additions and 1 deletions

View File

@ -271,6 +271,7 @@ Run a read-only health snapshot from the Debian server with:
```bash
./jeannie status
./jeannie capacity
./jeannie recover-plan
```
It reports host memory/disk, systemd services, Docker Compose stacks,
@ -295,6 +296,10 @@ summarizes Debian memory/disk/Docker usage, Kubernetes node and PVC usage,
Pimox VM/storage allocation, RPi Docker/disk state, and ends with placement
guidance for deciding what can safely run next.
`recover-plan` prints the disaster recovery order and lightweight prerequisite
checks, from Debian and Gitea through DNS, Pimox, Kubernetes, GitOps apps, and
edge verification.
Focused doctor commands run narrower read-only checks and print the most likely
next step:

View File

@ -271,6 +271,7 @@ Run a read-only health snapshot from the Debian server with:
```bash
./{{ main_script }} status
./{{ main_script }} capacity
./{{ main_script }} recover-plan
```
It reports host memory/disk, systemd services, Docker Compose stacks,
@ -295,6 +296,10 @@ summarizes Debian memory/disk/Docker usage, Kubernetes node and PVC usage,
Pimox VM/storage allocation, RPi Docker/disk state, and ends with placement
guidance for deciding what can safely run next.
`recover-plan` prints the disaster recovery order and lightweight prerequisite
checks, from Debian and Gitea through DNS, Pimox, Kubernetes, GitOps apps, and
edge verification.
Focused doctor commands run narrower read-only checks and print the most likely
next step:

View File

@ -57,6 +57,11 @@ deployment readiness, pod restarts, node pressure, disk pressure, Traefik
Kubernetes node and PVC usage, Pimox VM/storage allocation, RPi Docker/disk
state, and practical placement guidance.
`recover-plan`
: Print the ordered disaster recovery checklist and lightweight prerequisite
checks for Debian, Gitea, RPi DNS, Pimox, Kubernetes, GitOps apps, and edge
verification.
Prometheus alerts
: High-signal alert rules are managed as the `apps/homelab-alerts` GitOps app.
They cover node readiness, restart storms, unavailable deployments, storage

10
jeannie
View File

@ -5967,6 +5967,11 @@ capacity_report() {
"${REPO_ROOT}/scripts/capacity-report"
}
recover_plan() {
require_debian_server "recover-plan"
"${REPO_ROOT}/scripts/recover-plan"
}
case "${1:-}" in
up)
up
@ -5989,6 +5994,9 @@ case "${1:-}" in
capacity)
capacity_report
;;
recover-plan)
recover_plan
;;
validate)
validate_homelab
;;
@ -6133,7 +6141,7 @@ case "${1:-}" in
echo "Log: ${JEANNIE_LOG_FILE}"
;;
*)
echo "Usage: $0 {up|plan [all|provisioning|cluster|platform|apps|edge]|rebuild-cluster|stop-cluster|start-cluster|status|capacity|validate|access-audit|kubeconfig-readonly|apps|website-translation-model|website-ollama-listen|ollama-setup|deploy-gitea|rpi-services|bootstrap-gitea-repo|backup-gitea|drill-restore|drill-gitea-restore|drill-pihole-restore|install-gitea-runner|move-prometheus-stack-workers|doctor-versions|doctor-edge|doctor-gitea|doctor-rpi|doctor-cluster|preflight|doctor-preapply|inventory-check|state-backup|fix-debian-docker-root|secrets-init|secrets-check|tailnet-policy-check|ai-index|ai-check|security-scan|security-prepare|security-zap|security-k8s|security-host|security-trivy|security-secrets|security-nuclei|security-web|security-logs|security-runtime|security-attack-path|openwrt|nuke}"
echo "Usage: $0 {up|plan [all|provisioning|cluster|platform|apps|edge]|rebuild-cluster|stop-cluster|start-cluster|status|capacity|recover-plan|validate|access-audit|kubeconfig-readonly|apps|website-translation-model|website-ollama-listen|ollama-setup|deploy-gitea|rpi-services|bootstrap-gitea-repo|backup-gitea|drill-restore|drill-gitea-restore|drill-pihole-restore|install-gitea-runner|move-prometheus-stack-workers|doctor-versions|doctor-edge|doctor-gitea|doctor-rpi|doctor-cluster|preflight|doctor-preapply|inventory-check|state-backup|fix-debian-docker-root|secrets-init|secrets-check|tailnet-policy-check|ai-index|ai-check|security-scan|security-prepare|security-zap|security-k8s|security-host|security-trivy|security-secrets|security-nuclei|security-web|security-logs|security-runtime|security-attack-path|openwrt|nuke}"
exit 1
;;
esac

93
scripts/recover-plan Executable file
View File

@ -0,0 +1,93 @@
#!/usr/bin/env bash
set -euo pipefail
REPO_ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
KUBECONFIG="${KUBECONFIG:-${KUBECONFIG_PATH:-${HOME}/.kube/config}}"
check() {
local label="$1"
shift
printf '%-36s ' "${label}"
if "$@" >/tmp/homelab-recover-check.$$ 2>&1; then
printf 'ok\n'
else
printf 'check\n'
sed -n '1,3p' /tmp/homelab-recover-check.$$ | sed 's/^/ /'
fi
rm -f /tmp/homelab-recover-check.$$
}
has_gitea_backup() {
find "${LAB_GITEA_BACKUP_DIR:-/home/jv/backups/gitea}" -maxdepth 1 -type f -name 'gitea-*.zip' -print -quit | grep -q .
}
has_state_backup() {
find "${HOMELAB_TOFU_STATE_BACKUP_DIR:-${XDG_DATA_HOME:-${HOME}/.local/share}/homelab/tofu-state-backups}" -maxdepth 1 -type f -name 'tofu-state-*.tgz' -print -quit | grep -q .
}
main() {
cat <<'EOF'
== Homelab Recovery Order ==
1. Restore Debian control host baseline.
Command: ./jeannie validate
Check: Docker, containerd, Tailscale, kubeconfig path, repo checkout.
2. Restore or verify Gitea.
Commands:
./jeannie deploy-gitea
./jeannie drill-gitea-restore
./jeannie bootstrap-gitea-repo
3. Restore RPi DNS services.
Commands:
./jeannie rpi-services
./jeannie doctor-rpi
./jeannie drill-pihole-restore
4. Verify Pimox template and worker storage.
Commands:
./jeannie preflight
./jeannie doctor-preapply
5. Rebuild or start Kubernetes.
If state exists and nodes are stopped: ./jeannie start-cluster
If cluster must be rebuilt: ./jeannie rebuild-cluster
6. Restore platform and apps from GitOps.
Commands:
./jeannie apps
./jeannie gitops-status
7. Verify public edge.
Commands:
./jeannie doctor-edge
./jeannie cert-check
./jeannie status
8. Capture final recovery evidence.
Commands:
./jeannie release-snapshot
./jeannie backup-gitea
./jeannie state-backup
EOF
echo "== Current Recovery Inputs =="
check "repo validation" "${REPO_ROOT}/scripts/validate-homelab"
check "Gitea backup archive" has_gitea_backup
check "OpenTofu state backup" has_state_backup
check "Pi-hole restore inputs" "${REPO_ROOT}/scripts/restore-drill" pihole
check "Kubernetes API" kubectl --kubeconfig "${KUBECONFIG}" get --raw=/readyz
cat <<'EOF'
== Recovery Rule ==
Prefer restoring source-of-truth first: repo, Gitea, DNS, state backups, then
rebuild compute. Treat Pimox worker disks as rebuildable unless a service has an
explicit backup and restore path.
EOF
}
main "$@"