Configure Garage object storage on garage01 with S3 API, Vault secrets
for RPC secret and admin token, and Caddy reverse proxy for HTTPS access
at s3.home.2rjus.net via internal ACME CA. Includes flake entry, VM
definition, and Vault policy for the host.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add a systemd timer that triggers builds for all hosts every 2 hours
via NATS, keeping the binary cache warm.
- Add scheduler.nix with timer (every 2h) and oneshot service
- Add scheduler NATS user to DEPLOY account
- Add Vault secret and variable for scheduler NKey
- Increase nix-cache02 memory from 16GB to 20GB
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Removed:
- hosts/nix-cache01/ directory
- services/nix-cache/build-flakes.{nix,sh} (replaced by NATS builder)
- Vault secret and AppRole for nix-cache01
- Old signing key variable from terraform
- Old trusted public key from system/nix.nix
Updated:
- flake.nix: removed nixosConfiguration
- README.md: nix-cache01 -> nix-cache02
- Monitoring rules: removed build-flakes alerts, updated harmonia to nix-cache02
- Simplified proxy.nix (no longer needs hostname conditional)
nix-cache02 is now the sole binary cache host.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Parameterize harmonia.nix to use hostname-based Vault paths
- Add nix-cache services to nix-cache02
- Add Vault secret and variable for nix-cache02 signing key
- Add nix-cache02 public key to trusted-public-keys on all hosts
- Update plan doc to remove actions runner references
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The actions runner on nix-cache01 was never actively used.
Removing it before migrating to nix-cache02.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Configure builder to build nixos-servers and nixos (gunter) repos
- Add builder NKey to Vault secrets
- Update NATS permissions for builder, test-deployer, and admin-deployer
- Grant nix-cache02 access to shared homelab-deploy secrets
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
New build host to replace nix-cache01 with:
- 8 CPU cores, 16GB RAM, 200GB disk
- Static IP 10.69.13.25
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add vault secrets for Radarr and Sonarr API keys to enable
exportarr metrics collection on monitoring01.
- services/exportarr/radarr - Radarr API key
- services/exportarr/sonarr - Sonarr API key
- Grant monitoring01 access to services/exportarr/*
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Enable Kanidm users to authenticate to OpenBao via OIDC for Web UI access.
Members of the admins group get full read/write access to secrets.
Changes:
- Add OIDC auth backend in Terraform (oidc.tf)
- Add oidc-admin and oidc-default policies
- Add openbao OAuth2 client to Kanidm
- Enable legacy crypto (RS256) for OpenBao compatibility
- Allow imperative group membership management in Kanidm
Limitations:
- CLI login not supported (Kanidm requires HTTPS for confidential client redirects)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Create shared policy granting all hosts access to nixos-exporter nkey
- Add policy to both manual and generated host AppRoles
- Remove duplicate kanidm01/monitoring02 entries from hosts-generated.tf
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
When one host fetches the latest flake revision, it publishes to NATS
and all other hosts receive the update immediately. This reduces
redundant nix flake metadata calls across the fleet.
- Add nkeys to devshell for key generation
- Add nixos-exporter user to NATS HOMELAB account
- Add Vault secret for NKey storage
- Configure all hosts to use NATS for revision sharing
- Update nixos-exporter input to version with NATS support
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Deploy Grafana test instance on monitoring02 with:
- Kanidm OIDC authentication (admins -> Admin role, others -> Viewer)
- PKCE enabled for secure OAuth2 flow (required by Kanidm)
- Declarative datasources for Prometheus and Loki on monitoring01
- Local Caddy for TLS termination via internal ACME CA
- DNS CNAME grafana-test.home.2rjus.net
Terraform changes add OAuth2 client secret and AppRole policies for
kanidm01 and monitoring02.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
New test-tier host for monitoring stack expansion with:
- Static IP 10.69.13.24
- 4 CPU cores, 4GB RAM, 20GB disk
- Vault integration and NATS-based deployment enabled
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- New test-tier VM at 10.69.13.23 with role=auth
- Kanidm 1.8 server with HTTPS (443) and LDAPS (636)
- ACME certificate from internal CA (auth.home.2rjus.net)
- Provisioned groups: admins, users, ssh-users
- Provisioned user: torjus
- Daily backups at 22:00 (7 versions)
- Prometheus monitoring scrape target
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
ns1 needs access to shared/dns/* for zone transfer key and
shared/homelab-deploy/* for the NATS listener.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Old VM had incorrect hardware-configuration.nix with hardcoded UUIDs
that didn't match actual disk layout, causing boot failure (emergency mode).
Recreated using template2-based configuration for OpenTofu provisioning.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Remove hosts/template/ (legacy template1) and give each legacy host
its own hardware-configuration.nix copy
- Recreate ns2 using create-host with template2 base
- Add secondary DNS services (NSD + Unbound resolver)
- Configure Vault policy for shared DNS secrets
- Fix create-host IP uniqueness validator to check CIDR notation
(prevents false positives from DNS resolver entries)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Enable vault.enable and homelab.deploy.enable on vault01 so it can
receive NATS-based remote deployments. Vault fetches secrets from
itself using AppRole after auto-unseal.
Add systemd ordering to ensure vault-secret services wait for openbao
to be unsealed before attempting to fetch secrets.
Also adds vault01 AppRole entry to Terraform.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
VMs are now bootstrapped and running. Remove temporary flake_branch
and vault_wrapped_token settings so they use master going forward.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The homelab-deploy listener requires access to shared/homelab-deploy/*
secrets. Update hosts-generated.tf and the generator script to include
this policy automatically.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add vault.enable = true to testvm01, testvm02, testvm03
- Add homelab.deploy.enable = true for remote deployment via NATS
- Update create-host template to include these by default
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Three permanent test hosts for validating deployment and bootstrapping
workflow. Each host configured with:
- Static IP (10.69.13.20-22/24)
- Vault AppRole integration
- Bootstrap from deploy-test-hosts branch
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add homelab.deploy.enable option (requires vault.enable)
- Create shared homelab-deploy Vault policy for all hosts
- Enable homelab.deploy on all vault-enabled hosts
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add system/homelab-deploy.nix module that automatically enables the
listener on all hosts with vault.enable=true. Uses homelab.host.tier
and homelab.host.role for NATS subject subscriptions.
- Add homelab-deploy access to all host AppRole policies
- Remove manual listener config from vaulttest01 (now handled by system module)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add homelab-deploy flake input and NixOS module for message-based
deployments across the fleet. Configure DEPLOY account in NATS with
tiered access control (listener, test-deployer, admin-deployer).
Enable listener on vaulttest01 as initial test host.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Instead of creating a long-lived Vault token in Terraform (which gets
invalidated when Terraform recreates it), monitoring01 now uses its
existing AppRole credentials to fetch a fresh token for Prometheus.
Changes:
- Add prometheus-metrics policy to monitoring01's AppRole
- Remove vault_token.prometheus_metrics resource from Terraform
- Remove openbao-token KV secret from Terraform
- Add systemd service to fetch AppRole token on boot
- Add systemd timer to refresh token every 30 minutes
This ensures Prometheus always has a valid token without depending on
Terraform state or manual intervention.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add exporters and scrape targets for services lacking monitoring:
- PostgreSQL: postgres-exporter on pgdb1
- Authelia: native telemetry metrics on auth01
- Unbound: unbound-exporter with remote-control on ns1/ns2
- NATS: HTTP monitoring endpoint on nats1
- OpenBao: telemetry config and Prometheus scrape with token auth
- Systemd: systemd-exporter on all hosts for per-service metrics
Add alert rules for postgres, auth (authelia + lldap), jellyfin,
vault (openbao), plus extend existing nats and unbound rules.
Add Terraform config for Prometheus metrics policy and token. The
token is created via vault_token resource and stored in KV, so no
manual token creation is needed.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Remove backup_helper_secret variable and switch shared/backup/password
to auto_generate. New password will be added alongside existing restic
repository key.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Replace sops-nix secrets with OpenBao vault secrets across all hosts.
Hardcode root password hash, add extractKey option to vault-secrets
module, update Terraform with secrets/policies for all hosts, and
create AppRole provisioning playbook.
Hosts migrated: ha1, monitoring01, ns1, ns2, http-proxy, nix-cache01
Wave 1 hosts (nats1, jelly01, pgdb1) get AppRole policies only.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Fix error "500 can't upload to storage type 'zfspool'" by using "local"
storage pool for cloud-init disks instead of the VM's storage pool.
Cloud-init disks require storage that supports ISO/snippet content types,
which zfspool does not. The "local" storage pool (directory-based) supports
this content type.
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Fix OpenTofu error where static IP and DHCP branches had different object
structures in the subnets array. Move conditional to network_config level
so both branches return complete, consistent yamlencode() results.
Error was: "The true and false result expressions must have consistent types"
Solution: Make network_config itself conditional rather than the subnets
array, ensuring both branches return the same type (string from yamlencode).
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
