Files
computer-playbook/roles/svc-bkp-secrets-2-local/README.md
kevinveenbirkenbachandClaude Opus 5 4986e80dd4 feat(readme): ask the reader for a domain the deployment can resolve
The Production block never set one. inventory provision writes only
ansible_connection for a local host, so DOMAIN_PRIMARY stayed at the
group_vars default of infinito.localhost and the deployment served a zone its
own resolver knows nothing about. CoreDNS templates answers for
INFINITO_DOMAIN_RE and nothing else, so every lookup of the deployment's own
name came back NXDOMAIN.

Mastodon is where that surfaced. tootctl accounts create validates the address
through EmailMxValidator, which resolves MX and then, for the domain itself,
A and AAAA. All six queries returned NXDOMAIN in run 33574056480, the
validator recorded email :unreachable, and the admin account was never created.
Only the guide rows were affected: the matrix rows write
DOMAIN_PRIMARY: "{{ lookup('env', 'INFINITO_DOMAIN') }}" into their host_vars
and resolve fine.

The template now carries DOMAIN=<your-domain> in both the host and the
container form and threads it into --vars, with the container form passing it
through the docker run boundary as well. The two guide runners substitute it
the way they already substitute the SSH key, so the replay deploys under
INFINITO_DOMAIN like every other CI deploy. compose_deploy.sh moves its env
load ahead of the substitutions to have the value; host_deploy.sh had no env
load at all.

A production reader is asked for a domain now instead of silently inheriting
infinito.localhost, which no real deployment wants either. The shipped default
keeps its value: sys-svc-nss documents that *.localhost resolves through the
myhostname NSS module without any zone, and host-side tasks lean on it.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-09-02 18:30:25 +02:00

5.9 KiB

Backup Host Secrets

Description

A scheduled, deduplicating backup of the host's generated secret material to the local backup directory: the Infinito.Nexus secrets and tokens, the self-signed root CA, the Let's Encrypt account + certificates, the ACME DNS credentials, and the node identity (ssh host keys + machine-id). None of this lives in a docker volume, the NFS export or the operator inventory, so it is the gap the volume/nfs backup roles do not cover.

Overview

The role installs a systemd unit that snapshots each configured source directory into <backups>/<machine-hash>/backup-secrets-to-local/<generation>/files/<name>, hard-linked against the previous generation. Optional sources are skipped when absent (the CA in Let's Encrypt mode, ACME/certbot in self-signed mode). The generation joins the same backup tree the pull and device roles consume, so it flows to the backup host and the encrypted device automatically.

Cosmos

The diagram places Backup Host Secrets in the Infinito.Nexus cosmos: the components it deploys (capabilities), the central services it consumes (dependencies), and its outward reach (federation and bridged external networks).

flowchart LR
    subgraph deps [Dependencies]
        dep_sys_ctl_cln_faild_bkps["sys-ctl-cln-faild-bkps 💻 ⚙️"]
    end
    subgraph role [svc-bkp-secrets-2-local 💻]
        svc_secrets_2_local["secrets-2-local"]
    end
    subgraph dependents [Dependents]
        dpt_svc_bkp_volume_2_local["svc-bkp-volume-2-local 💻"]
    end
    dep_sys_ctl_cln_faild_bkps -- "1:1" --> svc_secrets_2_local
    svc_secrets_2_local -- "1:1" --> dpt_svc_bkp_volume_2_local

Solid 1:1 edges are fixed relationships; dashed 0..1 edges are conditional (enabled only in matching deployments); red 0..0 edges are turned off in this role. Node markers show the role's deploy modes (💻 host, 🐳 compose, 🐝 swarm); ❌ marks a service that is explicitly turned off, and ⚙️ an Ansible role dependency declared in meta/main.yml.

Schema

flowchart TD
    TIMER["systemd timer<br>SYS_SCHEDULE_BACKUP_SECRETS_TO_LOCAL"] --> UNIT
    PRELOAD["sys-service-loader preload<br>MODE_BACKUP, before the app pass<br>(force_flush_instant + state started)"] --> UNIT
    UNIT["svc-bkp-secrets-2-local.&lt;version&gt;.&lt;domain&gt;.service"] --> LOCK["ExecStartPre: sys-lock against the manipulation group"]
    LOCK --> SCRIPT["ExecStart: script &lt;backups_dir&gt; backup-secrets-to-local<br>secrets= ca= acme= certbot= sources"]
    SCRIPT --> SOURCES["per-source rsync snapshots<br>--link-dest previous generation (hard links)<br>optional sources skipped when absent"]
    SOURCES --> TREE["&lt;backups_dir&gt;/&lt;sha256(machine-id)&gt;/<br>backup-secrets-to-local/&lt;YYYYmmddHHMMSS&gt;/files/&lt;name&gt;"]
    UNIT -->|failure| ALARM["OnFailure: alarm"]
    TREE --> PULL["svc-bkp-remote-2-local via ssh<br>user-backup ssh-wrapper: whitelisted ls/rsync per type<br>pulls the newest generation"]

Features

  • Complete secret coverage: secrets + tokens, self-signed CA, Let's Encrypt tree, ACME DNS credentials, ssh host keys and machine-id in one differential generation.
  • Differential snapshots: rsync --link-dest per subtree against the previous generation deduplicates unchanged files.
  • Mode-aware: absent optional sources are skipped, not failed, so the same role works in both self-signed and Let's Encrypt setups.
  • Chain-native: the generation lands in the standard backup tree, so svc-bkp-remote-2-local pulls it and svc-bkp-local-2-device mirrors it onto the encrypted device without extra wiring.

Quick Setup

Development

Clone, set up the workstation, and deploy Backup Host Secrets onto the local stack:

git clone https://github.com/infinito-nexus/core.git
cd core
make onboard
make compose-deploy mode=reinstall apps=svc-bkp-secrets-2-local full_cycle=false

Production

Install Backup Host Secrets directly onto the target machine: clone the repository, install the OS prerequisites and the repository toolchain, then deploy against localhost over a local connection (no SSH, no container):

git clone https://github.com/infinito-nexus/core.git
cd core
bash scripts/install/package.sh
make install
source scripts/meta/env/load.sh

APP=svc-bkp-secrets-2-local
DOMAIN=<your-domain>
TLS_MODE=self_signed
SSH_PUBLIC_KEY="<your-ssh-public-key>"
INVENTORY=inventories/production
infinito administration inventory provision "$INVENTORY" \
  --inventory-file "$INVENTORY/devices.yml" \
  --host localhost \
  --include "$APP" \
  --vars "{\"TLS_MODE\": \"$TLS_MODE\", \"DOMAIN_PRIMARY\": \"$DOMAIN\", \"users\": {\"administrator\": {\"authorized_keys\": [\"$SSH_PUBLIC_KEY\"]}}}"
infinito administration deploy dedicated "$INVENTORY/devices.yml" \
  --password-file "$INVENTORY/.password" \
  --diff -vv

Recover

Run files/python/recover.py on the target host:

recover.py <backups>/<machine-hash>/backup-secrets-to-local/<generation>/files [--restore-node-identity]

The script first starts the role's deployed backup unit (a fresh differential generation of the live material), then mirrors each present subtree back to its fixed system path (rsync -a --delete): secrets into /var/lib/infinito/secrets, ca into /etc/<domain>/ca, acme into /etc/letsencrypt, certbot into /etc/certbot. --no-safety-backup skips the pre-recover unit run. The node subtree (ssh host keys + machine-id) is restored only with --restore-node-identity, because overwriting the running machine-id / ssh host keys changes the host's identity mid-flight — do it on a fresh host after total loss.

Credits

Implemented by Kevin Veen-Birkenbach. Part of the Infinito.Nexus Project and maintained by Kevin Veen-Birkenbach. Licensed under the Infinito.Nexus Community License (Non-Commercial).