fix(wordpress): route the outpost OAuth path so the gate is a gate, not a dead end

Verified against the deployed cluster: GET /wp-admin/ correctly 302s to
https://www.renner.uno/outpost.goauthentik.io/start?rd=... — and that URL
returned WordPress's 404 page, because /outpost.goauthentik.io fell through to
the public '/' prefix router.

The effect was that /wp-admin was unreachable rather than protected. An
attacker gains nothing, but neither does the editor the gate exists for.

paperless needs the same route and solves it with a separate Ingress on
auth.renner.uno (apps/authentik/paperless-gate-ingress.yaml). Here the whole
flow stays on www.renner.uno, so the path belongs in this rule set.

These endpoints are the login flow itself and are served by the outpost, so
routing them to it is not an authentication hole.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014giKN9N6e5J2DF8QZmVh98

ansible/plays/install-blackbox-exporter.yml

@@ -1,19 +0,0 @@
----
-# Deploy Blackbox Exporter to Prometheus server for endpoint probing
-# Usage: ansible-playbook -i inventory/ plays/install-blackbox-exporter.yml
-
-- name: Install Blackbox Exporter
-  hosts: prometheus.srv.renner.uno
-  become: yes
-  roles:
-    - blackbox-exporter
-
-  post_tasks:
-    - name: Verify blackbox_exporter is running
-      uri:
-        url: "http://127.0.0.1:9115/metrics"
-        status_code: 200
-      register: exporter_health
-      retries: 3
-      delay: 5
-      until: exporter_health.status == 200

ansible/plays/install-log-aggregation.yml

@@ -1,86 +0,0 @@
----
-# Setup Log Aggregation Infrastructure
-# This playbook configures the complete logging stack:
-# 1. Loki server for log storage and querying
-# 2. Alloy agents on all VMs/CTs for log collection
-# 3. Syslog receiver for remote devices (Synology NAS, network devices)
-#
-# Usage: ansible-playbook -i inventory/ plays/setup-log-aggregation.yml
-#
-# To limit to specific hosts:
-#   --limit loki.srv.renner.uno          # Only configure Loki server
-#   --limit proxmox_all_running          # Only configure log agents
-
-- name: Setup Loki Log Aggregation Server
-  hosts: proxmox_logging
-  become: yes
-  gather_facts: yes
-
-  roles:
-    - common
-    - loki-server
-
-  post_tasks:
-    - name: Verify Loki is running
-      uri:
-        url: "http://127.0.0.1:3100/ready"
-        status_code: 200
-      register: loki_health
-      retries: 5
-      delay: 10
-      until: loki_health.status == 200
-
-    - name: Display Loki status
-      debug:
-        msg: "Loki is running and healthy at http://{{ inventory_hostname }}:3100"
-
-- name: Deploy Monitoring Agent to Loki Server (with syslog receiver)
-  hosts: proxmox_logging
-  become: yes
-  gather_facts: yes
-
-  vars:
-    loki_endpoint: "http://loki.srv.renner.uno:3100"
-    prometheus_remote_write_endpoint: "http://prometheus.srv.renner.uno:9090/api/v1/write"
-    # Syslog receiver settings are defined in host_vars/loki.srv.renner.uno/vars.yml:
-    # - alloy_syslog_receiver_enabled: true
-    # - alloy_syslog_receiver_port: 1514
-    # - alloy_syslog_receiver_protocol: "udp"
-    # - alloy_syslog_receiver_format: "raw"
-    # - alloy_stability_level: "experimental"
-
-  roles:
-    - monitoring-agent
-
-  post_tasks:
-    - name: Display syslog receiver info
-      debug:
-        msg: |
-          Syslog receiver is listening on UDP port {{ alloy_syslog_receiver_port | default(1514) }}
-          Configure remote devices to send syslog to: {{ inventory_hostname }}:{{ alloy_syslog_receiver_port | default(1514) }}
-      when: alloy_syslog_receiver_enabled | default(false)
-
-- name: Deploy Monitoring Agent to All Running VMs/CTs
-  hosts: proxmox_all_running:!proxmox_logging:!proxmox_noansible
-  become: yes
-  gather_facts: yes
-  serial: 10  # Deploy to 10 hosts at a time to avoid overwhelming the network
-
-  vars:
-    loki_endpoint: "http://loki.srv.renner.uno:3100"
-    prometheus_remote_write_endpoint: "http://prometheus.srv.renner.uno:9090/api/v1/write"
-
-  roles:
-    - monitoring-agent
-
-  post_tasks:
-    - name: Verify Alloy is running
-      systemd:
-        name: alloy
-        state: started
-      register: alloy_service
-
-    - name: Display agent status
-      debug:
-        msg: "Monitoring agent deployed successfully on {{ inventory_hostname }}"
-      when: alloy_service is defined

ansible/plays/install-loki.yml

@@ -1,26 +0,0 @@
----
-# Install Loki Log Aggregation Server
-# Usage: ansible-playbook -i inventory/ plays/install-loki.yml
-
-- name: Install Loki Log Aggregation Server
-  hosts: loki.srv.renner.uno
-  become: yes
-  gather_facts: yes
-
-  roles:
-    - common
-    - loki-server
-
-  post_tasks:
-    - name: Verify Loki is running
-      uri:
-        url: "http://127.0.0.1:3100/ready"
-        status_code: 200
-      register: loki_health
-      retries: 5
-      delay: 10
-      until: loki_health.status == 200
-
-    - name: Display Loki status
-      debug:
-        msg: "Loki is running and healthy at http://{{ inventory_hostname }}:3100"

ansible/plays/install-prometheus.yml

@@ -1,43 +0,0 @@
----
-# Install Prometheus Metrics Server
-# Usage: ansible-playbook -i inventory/ plays/install-prometheus.yml
-
-- name: Install Prometheus Metrics Server
-  hosts: prometheus.srv.renner.uno
-  become: yes
-  gather_facts: yes
-
-  vars:
-    # Synology NAS SNMP targets
-    snmp_synology_targets:
-      - name: nas-1
-        target: "10.2.40.50"
-      - name: nas-2
-        target: "10.3.40.57"
-
-    # SNMPv3 credentials from vault
-    snmp_use_v3: true
-    snmp_v3_username: "{{ vault_snmp_v3_username }}"
-    snmp_v3_auth_password: "{{ vault_snmp_v3_auth_password }}"
-    snmp_v3_priv_password: "{{ vault_snmp_v3_priv_password }}"
-
-  roles:
-    - common
-    - prometheus-server
-    - snmp-exporter
-    - pve-exporter
-    - unifi-poller
-
-  post_tasks:
-    - name: Verify Prometheus is running
-      uri:
-        url: "http://127.0.0.1:9090/-/ready"
-        status_code: 200
-      register: prometheus_health
-      retries: 5
-      delay: 10
-      until: prometheus_health.status == 200
-
-    - name: Display Prometheus status
-      debug:
-        msg: "Prometheus is running and healthy at http://{{ inventory_hostname }}:9090"

ansible/plays/update.yml

@@ -16,7 +16,6 @@
 - import_playbook: install-grist.yml
 - import_playbook: install-music-assistant.yml
 - import_playbook: update-docker.yml
-- import_playbook: install-log-aggregation.yml
 - import_playbook: update-pve-rolling.yml
 
 - name: Update Summary

ansible/roles/blackbox-exporter/defaults/main.yml

@@ -1,26 +0,0 @@
----
-blackbox_exporter_version: "0.25.0"
-blackbox_exporter_port: 9115
-blackbox_exporter_user: "blackbox_exporter"
-blackbox_exporter_group: "blackbox_exporter"
-
-# Endpoints to probe
-blackbox_http_targets:
-  - name: "gitea"
-    url: "https://gitea.renner.uno"
-  - name: "immich"
-    url: "https://immich.renner.uno"
-  - name: "authentik"
-    url: "https://auth.renner.uno"
-  - name: "home-assistant"
-    url: "https://ha.renner.uno"
-  - name: "grafana"
-    url: "https://grafana.renner.uno"
-  - name: "prometheus"
-    url: "https://prometheus.renner.uno"
-
-blackbox_dns_targets: []
-
-blackbox_tcp_targets:
-  - name: "postgres-ha"
-    target: "10.2.40.135:5432"

ansible/roles/blackbox-exporter/handlers/main.yml

@@ -1,6 +0,0 @@
----
-- name: Restart blackbox_exporter
-  systemd:
-    name: blackbox_exporter
-    state: restarted
-    daemon_reload: yes

ansible/roles/blackbox-exporter/tasks/main.yml

@@ -1,80 +0,0 @@
----
-- name: Create blackbox_exporter group
-  group:
-    name: "{{ blackbox_exporter_group }}"
-    system: yes
-    state: present
-
-- name: Create blackbox_exporter user
-  user:
-    name: "{{ blackbox_exporter_user }}"
-    group: "{{ blackbox_exporter_group }}"
-    system: yes
-    shell: /usr/sbin/nologin
-    createhome: no
-    state: present
-
-- name: Create blackbox_exporter directories
-  file:
-    path: "{{ item }}"
-    state: directory
-    owner: "{{ blackbox_exporter_user }}"
-    group: "{{ blackbox_exporter_group }}"
-    mode: '0755'
-  loop:
-    - /etc/blackbox_exporter
-    - /var/lib/blackbox_exporter
-
-- name: Download blackbox_exporter
-  get_url:
-    url: "https://github.com/prometheus/blackbox_exporter/releases/download/v{{ blackbox_exporter_version }}/blackbox_exporter-{{ blackbox_exporter_version }}.linux-amd64.tar.gz"
-    dest: "/tmp/blackbox_exporter-{{ blackbox_exporter_version }}.tar.gz"
-    mode: '0644'
-
-- name: Extract blackbox_exporter
-  unarchive:
-    src: "/tmp/blackbox_exporter-{{ blackbox_exporter_version }}.tar.gz"
-    dest: /tmp
-    remote_src: yes
-    creates: "/tmp/blackbox_exporter-{{ blackbox_exporter_version }}.linux-amd64"
-
-- name: Install blackbox_exporter binary
-  copy:
-    src: "/tmp/blackbox_exporter-{{ blackbox_exporter_version }}.linux-amd64/blackbox_exporter"
-    dest: /usr/local/bin/blackbox_exporter
-    owner: root
-    group: root
-    mode: '0755'
-    remote_src: yes
-  notify: Restart blackbox_exporter
-
-- name: Deploy blackbox_exporter configuration
-  template:
-    src: blackbox.yml.j2
-    dest: /etc/blackbox_exporter/blackbox.yml
-    owner: "{{ blackbox_exporter_user }}"
-    group: "{{ blackbox_exporter_group }}"
-    mode: '0644'
-  notify: Restart blackbox_exporter
-
-- name: Deploy blackbox_exporter systemd service
-  template:
-    src: blackbox_exporter.service.j2
-    dest: /etc/systemd/system/blackbox_exporter.service
-    mode: '0644'
-  notify: Restart blackbox_exporter
-
-- name: Enable and start blackbox_exporter
-  systemd:
-    name: blackbox_exporter
-    state: started
-    enabled: yes
-    daemon_reload: yes
-
-- name: Clean up downloaded files
-  file:
-    path: "{{ item }}"
-    state: absent
-  loop:
-    - "/tmp/blackbox_exporter-{{ blackbox_exporter_version }}.tar.gz"
-    - "/tmp/blackbox_exporter-{{ blackbox_exporter_version }}.linux-amd64"

ansible/roles/blackbox-exporter/templates/blackbox.yml.j2

@@ -1,60 +0,0 @@
-modules:
-  http_2xx:
-    prober: http
-    timeout: 5s
-    http:
-      valid_http_versions: ["HTTP/1.1", "HTTP/2.0"]
-      valid_status_codes: []  # Defaults to 2xx
-      method: GET
-      follow_redirects: true
-      fail_if_ssl: false
-      fail_if_not_ssl: false
-      tls_config:
-        insecure_skip_verify: false
-
-  http_post_2xx:
-    prober: http
-    timeout: 5s
-    http:
-      method: POST
-      valid_status_codes: []
-
-  tcp_connect:
-    prober: tcp
-    timeout: 5s
-
-  # DNS module for external domain resolution (tests upstream/cache)
-  dns_external:
-    prober: dns
-    timeout: 5s
-    dns:
-      query_name: "google.com"
-      query_type: "A"
-      valid_rcodes:
-        - NOERROR
-
-  # DNS module for internal domain resolution (tests local resolution)
-  dns_internal:
-    prober: dns
-    timeout: 5s
-    dns:
-      query_name: "k8s-1.srv.renner.uno"
-      query_type: "A"
-      valid_rcodes:
-        - NOERROR
-
-  # DNS module for common cached domain
-  dns_cached:
-    prober: dns
-    timeout: 5s
-    dns:
-      query_name: "cloudflare.com"
-      query_type: "A"
-      valid_rcodes:
-        - NOERROR
-
-  icmp:
-    prober: icmp
-    timeout: 5s
-    icmp:
-      preferred_ip_protocol: "ip4"

ansible/roles/blackbox-exporter/templates/blackbox_exporter.service.j2

@@ -1,22 +0,0 @@
-[Unit]
-Description=Prometheus Blackbox Exporter
-Documentation=https://github.com/prometheus/blackbox_exporter
-Wants=network-online.target
-After=network-online.target
-
-[Service]
-Type=simple
-User={{ blackbox_exporter_user }}
-Group={{ blackbox_exporter_group }}
-ExecStart=/usr/local/bin/blackbox_exporter \
-    --config.file=/etc/blackbox_exporter/blackbox.yml \
-    --web.listen-address=:{{ blackbox_exporter_port }}
-Restart=on-failure
-RestartSec=5s
-
-# Allow ICMP probing (requires CAP_NET_RAW)
-AmbientCapabilities=CAP_NET_RAW
-CapabilityBoundingSet=CAP_NET_RAW
-
-[Install]
-WantedBy=multi-user.target

ansible/roles/loki-server/defaults/main.yml

@@ -1,44 +0,0 @@
----
-# Loki Server Configuration
-
-# Version and download
-loki_version: "3.4.2"
-loki_arch: "linux-amd64"
-loki_download_url: "https://github.com/grafana/loki/releases/download/v{{ loki_version }}/loki-{{ loki_arch }}.zip"
-loki_bin_path: "/usr/local/bin/loki"
-
-# Directories
-loki_data_dir: "/var/lib/loki"
-loki_config_dir: "/etc/loki"
-loki_config_file: "{{ loki_config_dir }}/loki-config.yaml"
-
-# Server configuration
-loki_http_listen_port: 3100
-loki_grpc_listen_port: 9096
-loki_http_listen_address: "0.0.0.0"
-
-# Retention settings (90 days)
-loki_retention_period: "2160h"  # 90 days
-loki_retention_delete_delay: "2h"
-
-# Storage configuration
-loki_storage_type: "filesystem"
-loki_chunks_directory: "{{ loki_data_dir }}/chunks"
-loki_rules_directory: "{{ loki_data_dir }}/rules"
-
-# Index configuration
-loki_index_prefix: "loki_index_"
-loki_index_period: "24h"
-
-# Limits
-loki_ingestion_rate_mb: 10
-loki_ingestion_burst_size_mb: 20
-loki_max_streams_per_user: 10000
-loki_max_entries_limit_per_query: 5000
-
-# Query settings
-loki_max_query_length: "720h"  # 30 days max query span
-
-# Systemd service
-loki_user: "loki"
-loki_group: "loki"

ansible/roles/loki-server/handlers/main.yml

@@ -1,10 +0,0 @@
----
-- name: Restart loki
-  systemd:
-    name: loki
-    state: restarted
-    daemon_reload: yes
-
-- name: Reload systemd
-  systemd:
-    daemon_reload: yes

ansible/roles/loki-server/tasks/main.yml

@@ -1,95 +0,0 @@
----
-# Loki Server Installation and Configuration
-
-- name: Install dependencies
-  apt:
-    name:
-      - unzip
-      - curl
-    state: present
-    update_cache: yes
-
-- name: Create loki user
-  user:
-    name: "{{ loki_user }}"
-    system: yes
-    shell: /usr/sbin/nologin
-    create_home: no
-
-- name: Create directories
-  file:
-    path: "{{ item }}"
-    state: directory
-    owner: "{{ loki_user }}"
-    group: "{{ loki_group }}"
-    mode: '0755'
-  loop:
-    - "{{ loki_config_dir }}"
-    - "{{ loki_data_dir }}"
-    - "{{ loki_chunks_directory }}"
-    - "{{ loki_rules_directory }}"
-    - "{{ loki_data_dir }}/index"
-    - "{{ loki_data_dir }}/index_cache"
-    - "{{ loki_data_dir }}/compactor"
-
-- name: Download Loki
-  get_url:
-    url: "{{ loki_download_url }}"
-    dest: "/tmp/loki.zip"
-    mode: '0644'
-
-- name: Unzip Loki
-  unarchive:
-    src: "/tmp/loki.zip"
-    dest: "/tmp"
-    remote_src: yes
-
-- name: Install Loki binary
-  copy:
-    src: "/tmp/loki-{{ loki_arch }}"
-    dest: "{{ loki_bin_path }}"
-    mode: '0755'
-    remote_src: yes
-  notify: Restart loki
-
-- name: Create Loki configuration
-  template:
-    src: loki-config.yaml.j2
-    dest: "{{ loki_config_file }}"
-    owner: "{{ loki_user }}"
-    group: "{{ loki_group }}"
-    mode: '0644'
-  notify: Restart loki
-
-- name: Create Loki systemd service
-  template:
-    src: loki.service.j2
-    dest: /etc/systemd/system/loki.service
-    mode: '0644'
-  notify:
-    - Reload systemd
-    - Restart loki
-
-- name: Enable and start Loki service
-  systemd:
-    name: loki
-    state: started
-    enabled: yes
-    daemon_reload: yes
-
-- name: Wait for Loki to be ready
-  uri:
-    url: "http://127.0.0.1:{{ loki_http_listen_port }}/ready"
-    status_code: 200
-  register: result
-  until: result.status == 200
-  retries: 30
-  delay: 5
-
-- name: Clean up downloaded files
-  file:
-    path: "{{ item }}"
-    state: absent
-  loop:
-    - "/tmp/loki.zip"
-    - "/tmp/loki-{{ loki_arch }}"

ansible/roles/loki-server/templates/loki-config.yaml.j2

@@ -1,72 +0,0 @@
-# Loki Configuration
-# Generated by Ansible
-
-auth_enabled: false
-
-server:
-  http_listen_address: {{ loki_http_listen_address }}
-  http_listen_port: {{ loki_http_listen_port }}
-  grpc_listen_port: {{ loki_grpc_listen_port }}
-  log_level: info
-
-common:
-  instance_addr: 127.0.0.1
-  path_prefix: {{ loki_data_dir }}
-  storage:
-    filesystem:
-      chunks_directory: {{ loki_chunks_directory }}
-      rules_directory: {{ loki_rules_directory }}
-  replication_factor: 1
-  ring:
-    kvstore:
-      store: inmemory
-
-query_range:
-  results_cache:
-    cache:
-      embedded_cache:
-        enabled: true
-        max_size_mb: 100
-
-schema_config:
-  configs:
-    - from: 2024-01-01
-      store: tsdb
-      object_store: filesystem
-      schema: v13
-      index:
-        prefix: {{ loki_index_prefix }}
-        period: {{ loki_index_period }}
-
-storage_config:
-  tsdb_shipper:
-    active_index_directory: {{ loki_data_dir }}/index
-    cache_location: {{ loki_data_dir }}/index_cache
-
-compactor:
-  working_directory: {{ loki_data_dir }}/compactor
-  compaction_interval: 10m
-  retention_enabled: true
-  retention_delete_delay: {{ loki_retention_delete_delay }}
-  retention_delete_worker_count: 150
-  delete_request_store: filesystem
-
-limits_config:
-  retention_period: {{ loki_retention_period }}
-  ingestion_rate_mb: {{ loki_ingestion_rate_mb }}
-  ingestion_burst_size_mb: {{ loki_ingestion_burst_size_mb }}
-  max_streams_per_user: {{ loki_max_streams_per_user }}
-  max_entries_limit_per_query: {{ loki_max_entries_limit_per_query }}
-  max_query_length: {{ loki_max_query_length }}
-  reject_old_samples: true
-  reject_old_samples_max_age: 168h  # 7 days
-
-ruler:
-  alertmanager_url: ""
-  storage:
-    type: local
-    local:
-      directory: {{ loki_rules_directory }}
-
-analytics:
-  reporting_enabled: false

ansible/roles/loki-server/templates/loki.service.j2

@@ -1,25 +0,0 @@
-[Unit]
-Description=Loki Log Aggregation System
-Documentation=https://grafana.com/docs/loki/latest/
-After=network-online.target
-Wants=network-online.target
-
-[Service]
-Type=simple
-User={{ loki_user }}
-Group={{ loki_group }}
-ExecStart={{ loki_bin_path }} -config.file={{ loki_config_file }}
-ExecReload=/bin/kill -HUP $MAINPID
-Restart=on-failure
-RestartSec=10s
-LimitNOFILE=65536
-
-# Security hardening
-NoNewPrivileges=yes
-ProtectSystem=strict
-ProtectHome=yes
-ReadWritePaths={{ loki_data_dir }}
-PrivateTmp=yes
-
-[Install]
-WantedBy=multi-user.target

ansible/roles/prometheus-server/defaults/main.yml

@@ -1,48 +0,0 @@
----
-# Prometheus Server Configuration
-
-# Version and download
-prometheus_version: "3.2.1"
-prometheus_arch: "linux-amd64"
-prometheus_download_url: "https://github.com/prometheus/prometheus/releases/download/v{{ prometheus_version }}/prometheus-{{ prometheus_version }}.{{ prometheus_arch }}.tar.gz"
-prometheus_bin_path: "/usr/local/bin/prometheus"
-prometheus_promtool_path: "/usr/local/bin/promtool"
-
-# Directories
-prometheus_data_dir: "/var/lib/prometheus"
-prometheus_config_dir: "/etc/prometheus"
-prometheus_config_file: "{{ prometheus_config_dir }}/prometheus.yml"
-prometheus_rules_dir: "{{ prometheus_config_dir }}/rules"
-prometheus_targets_dir: "{{ prometheus_config_dir }}/targets"
-
-# Server configuration
-prometheus_listen_address: "0.0.0.0"
-prometheus_listen_port: 9090
-prometheus_web_external_url: ""
-
-# Retention settings (90 days)
-prometheus_retention_time: "90d"
-prometheus_retention_size: "30GB"
-
-# Scrape settings
-prometheus_scrape_interval: "30s"
-prometheus_evaluation_interval: "30s"
-prometheus_scrape_timeout: "10s"
-
-# Remote write (for Alloy agents)
-prometheus_remote_write_enabled: true
-
-# Systemd service
-prometheus_user: "prometheus"
-prometheus_group: "prometheus"
-
-# Alert notification endpoints
-prometheus_alertmanager_url: ""
-
-# Email alerting (via Grafana)
-alerting_email_enabled: true
-alerting_email_to: ""
-
-# Home Assistant webhook for alerts
-alerting_ha_webhook_enabled: true
-alerting_ha_webhook_url: ""

ansible/roles/prometheus-server/handlers/main.yml

@@ -1,15 +0,0 @@
----
-- name: Restart prometheus
-  systemd:
-    name: prometheus
-    state: restarted
-    daemon_reload: yes
-
-- name: Reload prometheus
-  systemd:
-    name: prometheus
-    state: reloaded
-
-- name: Reload systemd
-  systemd:
-    daemon_reload: yes

ansible/roles/prometheus-server/tasks/main.yml

@@ -1,129 +0,0 @@
----
-# Prometheus Server Installation and Configuration
-
-- name: Install dependencies
-  apt:
-    name:
-      - curl
-    state: present
-    update_cache: yes
-
-- name: Create prometheus user
-  user:
-    name: "{{ prometheus_user }}"
-    system: yes
-    shell: /usr/sbin/nologin
-    create_home: no
-
-- name: Create directories
-  file:
-    path: "{{ item }}"
-    state: directory
-    owner: "{{ prometheus_user }}"
-    group: "{{ prometheus_group }}"
-    mode: '0755'
-  loop:
-    - "{{ prometheus_config_dir }}"
-    - "{{ prometheus_data_dir }}"
-    - "{{ prometheus_rules_dir }}"
-    - "{{ prometheus_targets_dir }}"
-
-- name: Download Prometheus
-  get_url:
-    url: "{{ prometheus_download_url }}"
-    dest: "/tmp/prometheus.tar.gz"
-    mode: '0644'
-
-- name: Extract Prometheus
-  unarchive:
-    src: "/tmp/prometheus.tar.gz"
-    dest: "/tmp"
-    remote_src: yes
-
-- name: Install Prometheus binaries
-  copy:
-    src: "/tmp/prometheus-{{ prometheus_version }}.{{ prometheus_arch }}/{{ item.src }}"
-    dest: "{{ item.dest }}"
-    mode: '0755'
-    remote_src: yes
-  loop:
-    - { src: "prometheus", dest: "{{ prometheus_bin_path }}" }
-    - { src: "promtool", dest: "{{ prometheus_promtool_path }}" }
-  notify: Restart prometheus
-
-- name: Check for console templates
-  stat:
-    path: "/tmp/prometheus-{{ prometheus_version }}.{{ prometheus_arch }}/consoles"
-  register: consoles_dir
-
-- name: Copy console templates
-  copy:
-    src: "/tmp/prometheus-{{ prometheus_version }}.{{ prometheus_arch }}/{{ item }}/"
-    dest: "{{ prometheus_config_dir }}/{{ item }}/"
-    remote_src: yes
-    owner: "{{ prometheus_user }}"
-    group: "{{ prometheus_group }}"
-  loop:
-    - consoles
-    - console_libraries
-  when: consoles_dir.stat.exists
-
-- name: Create Prometheus configuration
-  template:
-    src: prometheus.yml.j2
-    dest: "{{ prometheus_config_file }}"
-    owner: "{{ prometheus_user }}"
-    group: "{{ prometheus_group }}"
-    mode: '0644'
-    validate: "{{ prometheus_promtool_path }} check config %s"
-  notify: Restart prometheus
-
-- name: Create alert rules
-  template:
-    src: alert-rules.yml.j2
-    dest: "{{ prometheus_rules_dir }}/alert-rules.yml"
-    owner: "{{ prometheus_user }}"
-    group: "{{ prometheus_group }}"
-    mode: '0644'
-  notify: Reload prometheus
-
-- name: Create node exporter targets file
-  template:
-    src: nodes.yml.j2
-    dest: "{{ prometheus_targets_dir }}/nodes.yml"
-    owner: "{{ prometheus_user }}"
-    group: "{{ prometheus_group }}"
-    mode: '0644'
-
-- name: Create Prometheus systemd service
-  template:
-    src: prometheus.service.j2
-    dest: /etc/systemd/system/prometheus.service
-    mode: '0644'
-  notify:
-    - Reload systemd
-    - Restart prometheus
-
-- name: Enable and start Prometheus service
-  systemd:
-    name: prometheus
-    state: started
-    enabled: yes
-    daemon_reload: yes
-
-- name: Wait for Prometheus to be ready
-  uri:
-    url: "http://127.0.0.1:{{ prometheus_listen_port }}/-/ready"
-    status_code: 200
-  register: result
-  until: result.status == 200
-  retries: 30
-  delay: 5
-
-- name: Clean up downloaded files
-  file:
-    path: "{{ item }}"
-    state: absent
-  loop:
-    - "/tmp/prometheus.tar.gz"
-    - "/tmp/prometheus-{{ prometheus_version }}.{{ prometheus_arch }}"

ansible/roles/prometheus-server/templates/alert-rules.yml.j2

@@ -1,100 +0,0 @@
-# Prometheus Alert Rules
-# Generated by Ansible
-
-groups:
-  - name: infrastructure
-    rules:
-      # Disk usage alerts
-      - alert: HighDiskUsage
-        expr: (node_filesystem_avail_bytes{fstype!~"tmpfs|overlay"} / node_filesystem_size_bytes{fstype!~"tmpfs|overlay"}) * 100 < 15
-        for: 5m
-        labels:
-          severity: warning
-        annotations:
-          summary: "High disk usage on {{ '{{' }} $labels.instance {{ '}}' }}"
-          description: "Disk usage is above 85% on {{ '{{' }} $labels.mountpoint {{ '}}' }} ({{ '{{' }} $value | printf \"%.1f\" {{ '}}' }}% free)"
-
-      - alert: CriticalDiskUsage
-        expr: (node_filesystem_avail_bytes{fstype!~"tmpfs|overlay"} / node_filesystem_size_bytes{fstype!~"tmpfs|overlay"}) * 100 < 5
-        for: 2m
-        labels:
-          severity: critical
-        annotations:
-          summary: "Critical disk usage on {{ '{{' }} $labels.instance {{ '}}' }}"
-          description: "Disk usage is above 95% on {{ '{{' }} $labels.mountpoint {{ '}}' }} ({{ '{{' }} $value | printf \"%.1f\" {{ '}}' }}% free)"
-
-      # Memory alerts
-      - alert: HighMemoryUsage
-        expr: (1 - (node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes)) * 100 > 90
-        for: 5m
-        labels:
-          severity: warning
-        annotations:
-          summary: "High memory usage on {{ '{{' }} $labels.instance {{ '}}' }}"
-          description: "Memory usage is above 90% ({{ '{{' }} $value | printf \"%.1f\" {{ '}}' }}%)"
-
-      # CPU alerts
-      - alert: HighCPUUsage
-        expr: 100 - (avg by(instance) (rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100) > 85
-        for: 10m
-        labels:
-          severity: warning
-        annotations:
-          summary: "High CPU usage on {{ '{{' }} $labels.instance {{ '}}' }}"
-          description: "CPU usage is above 85% for 10 minutes ({{ '{{' }} $value | printf \"%.1f\" {{ '}}' }}%)"
-
-  - name: services
-    rules:
-      # Service down alerts
-      - alert: ServiceDown
-        expr: up == 0
-        for: 2m
-        labels:
-          severity: critical
-        annotations:
-          summary: "Service {{ '{{' }} $labels.job {{ '}}' }} down on {{ '{{' }} $labels.instance {{ '}}' }}"
-          description: "Service has been unreachable for more than 2 minutes"
-
-      # DNS VIP health
-      - alert: DNSVIPDown
-        expr: absent(up{job="dns-servers"} == 1) or count(up{job="dns-servers"} == 1) < 2
-        for: 1m
-        labels:
-          severity: critical
-        annotations:
-          summary: "DNS cluster degraded"
-          description: "Less than 2 DNS servers are healthy"
-
-  - name: security
-    rules:
-      # High auth failure rate (placeholder - needs log-based metrics from Loki)
-      - alert: HighAuthFailureRate
-        expr: rate(authentik_login_failed_total[5m]) > 5
-        for: 2m
-        labels:
-          severity: warning
-        annotations:
-          summary: "High authentication failure rate"
-          description: "More than 5 auth failures per minute detected"
-
-  - name: loki
-    rules:
-      # Loki health
-      - alert: LokiDown
-        expr: up{job="loki"} == 0
-        for: 2m
-        labels:
-          severity: critical
-        annotations:
-          summary: "Loki log aggregation is down"
-          description: "Loki has been unreachable for more than 2 minutes"
-
-      # Loki ingestion rate
-      - alert: LokiHighIngestionRate
-        expr: rate(loki_distributor_bytes_received_total[5m]) > 10485760
-        for: 5m
-        labels:
-          severity: warning
-        annotations:
-          summary: "High log ingestion rate in Loki"
-          description: "Log ingestion rate is above 10MB/s"

ansible/roles/prometheus-server/templates/nodes.yml.j2

@@ -1,26 +0,0 @@
-# Node Exporter targets - auto-generated by Ansible
-# Updates require re-running the prometheus playbook
-#
-# NOTE: Most hosts have Alloy agents that push metrics via remote_write.
-# This file only includes hosts that need direct scraping (standalone node_exporter).
-# Alloy-managed hosts are excluded to avoid duplicate "down" targets.
-
-# PVE cluster nodes - need direct scraping (no Alloy agent)
-- targets:
-    - 'jupiter.renner.uno:9100'
-  labels:
-    job: 'node'
-    group: 'pve_cluster'
-    hostname: 'jupiter'
-- targets:
-    - 'neptune.renner.uno:9100'
-  labels:
-    job: 'node'
-    group: 'pve_cluster'
-    hostname: 'neptune'
-- targets:
-    - 'saturn.renner.uno:9100'
-  labels:
-    job: 'node'
-    group: 'pve_cluster'
-    hostname: 'saturn'

ansible/roles/prometheus-server/templates/prometheus.service.j2

@@ -1,38 +0,0 @@
-[Unit]
-Description=Prometheus Monitoring System
-Documentation=https://prometheus.io/docs/
-After=network-online.target
-Wants=network-online.target
-
-[Service]
-Type=simple
-User={{ prometheus_user }}
-Group={{ prometheus_group }}
-ExecStart={{ prometheus_bin_path }} \
-  --config.file={{ prometheus_config_file }} \
-  --storage.tsdb.path={{ prometheus_data_dir }} \
-  --storage.tsdb.retention.time={{ prometheus_retention_time }} \
-  --storage.tsdb.retention.size={{ prometheus_retention_size }} \
-  --web.listen-address={{ prometheus_listen_address }}:{{ prometheus_listen_port }} \
-{% if prometheus_remote_write_enabled %}
-  --web.enable-remote-write-receiver \
-{% endif %}
-{% if prometheus_web_external_url %}
-  --web.external-url={{ prometheus_web_external_url }} \
-{% endif %}
-  --web.enable-lifecycle
-
-ExecReload=/bin/kill -HUP $MAINPID
-Restart=on-failure
-RestartSec=10s
-LimitNOFILE=65536
-
-# Security hardening
-NoNewPrivileges=yes
-ProtectSystem=strict
-ProtectHome=yes
-ReadWritePaths={{ prometheus_data_dir }}
-PrivateTmp=yes
-
-[Install]
-WantedBy=multi-user.target

ansible/roles/prometheus-server/templates/prometheus.yml.j2

@@ -1,304 +0,0 @@
-# Prometheus Configuration
-# Generated by Ansible
-
-global:
-  scrape_interval: {{ prometheus_scrape_interval }}
-  evaluation_interval: {{ prometheus_evaluation_interval }}
-  scrape_timeout: {{ prometheus_scrape_timeout }}
-  external_labels:
-    cluster: homelab
-    datacenter: renner
-
-rule_files:
-  - "{{ prometheus_rules_dir }}/*.yml"
-
-{% if prometheus_alertmanager_url %}
-alerting:
-  alertmanagers:
-    - static_configs:
-        - targets:
-            - {{ prometheus_alertmanager_url }}
-{% endif %}
-
-scrape_configs:
-  # Prometheus self-monitoring
-  - job_name: 'prometheus'
-    static_configs:
-      - targets: ['localhost:{{ prometheus_listen_port }}']
-
-  # Loki metrics
-  - job_name: 'loki'
-    static_configs:
-      - targets: ['loki.srv.renner.uno:3100']
-
-  # File-based service discovery for dynamic targets
-  - job_name: 'alloy-nodes'
-    file_sd_configs:
-      - files:
-          - '{{ prometheus_targets_dir }}/*.yml'
-          - '{{ prometheus_targets_dir }}/*.json'
-        refresh_interval: 30s
-
-{% if prometheus_remote_write_enabled %}
-# Remote write receiver for Alloy agents
-# Agents push metrics to /api/v1/write
-{% endif %}
-
-# Static targets for infrastructure components
-  # Proxmox VE API metrics (via pve-exporter)
-  - job_name: 'pve'
-    scrape_interval: 60s
-    scrape_timeout: 55s
-    static_configs:
-      - targets:
-          - 'jupiter.renner.uno'
-          - 'neptune.renner.uno'
-          - 'saturn.renner.uno'
-    metrics_path: /pve
-    params:
-      module: [default]
-      cluster: ['1']
-      node: ['1']
-    relabel_configs:
-      - source_labels: [__address__]
-        target_label: __param_target
-      - source_labels: [__param_target]
-        target_label: instance
-      - target_label: __address__
-        replacement: localhost:9221  # pve-exporter address
-      - target_label: group
-        replacement: proxmox
-      - target_label: type
-        replacement: hypervisor
-
-  # RustFS (S3-compatible storage)
-  # NOTE: RustFS uses OpenTelemetry push-model via --obs-endpoint, not Prometheus pull.
-  # Unlike MinIO, it doesn't expose /minio/v2/metrics/cluster endpoint.
-  # Metrics collection would require setting up RUSTFS_OBS_ENDPOINT to push to an OTEL collector.
-  # For now, node-level metrics are available via Alloy agents on the RustFS hosts.
-
-  # Home Assistant metrics
-  - job_name: 'homeassistant'
-    metrics_path: /api/prometheus
-    bearer_token: '{{ homeassistant_prometheus_token }}'
-    scheme: http
-    static_configs:
-      - targets: ['homeassistant.iot.renner.uno:8123']
-        labels:
-          group: 'smarthome'
-          type: 'homeassistant'
-
-  # UniFi Poller metrics
-  - job_name: 'unifi'
-    static_configs:
-      - targets: ['localhost:9130']
-        labels:
-          group: 'network'
-          type: 'unifi'
-
-  # Caddy reverse proxy metrics (HA cluster)
-  - job_name: 'caddy'
-    static_configs:
-      - targets:
-          - 'caddy.dmz.renner.uno:2019'
-          - 'caddy2.dmz.renner.uno:2019'
-        labels:
-          group: 'ingress'
-          type: 'reverse-proxy'
-
-  # PostgreSQL metrics (via postgres_exporter)
-  - job_name: 'postgresql'
-    static_configs:
-      - targets: ['localhost:9187']
-        labels:
-          group: 'database'
-          type: 'postgresql'
-
-  # Authentik identity provider metrics
-  - job_name: 'authentik'
-    static_configs:
-      - targets:
-          - 'authentik-app-1.srv.renner.uno:9300'
-          - 'authentik-app-2.srv.renner.uno:9300'
-        labels:
-          group: 'identity'
-          type: 'authentik'
-
-  # Node Exporter - system metrics from all hosts
-  - job_name: 'node'
-    file_sd_configs:
-      - files:
-          - '{{ prometheus_targets_dir }}/nodes.yml'
-        refresh_interval: 60s
-
-  # Chrony Exporter - NTP metrics from PVE cluster
-  # One static_config per host so each carries its own hostname label. The
-  # previous form listed all three targets together and recovered the hostname
-  # in relabel_configs by regex-matching __address__ against the node's IP --
-  # which silently coupled the labels to the addressing. Changing a target
-  # would have left the regex unmatched and the hostname label simply absent,
-  # with the scrape still succeeding.
-  - job_name: 'chrony'
-    static_configs:
-      - targets: ['jupiter.renner.uno:9123']
-        labels:
-          group: 'ntp'
-          type: 'time-server'
-          hostname: 'jupiter'
-      - targets: ['neptune.renner.uno:9123']
-        labels:
-          group: 'ntp'
-          type: 'time-server'
-          hostname: 'neptune'
-      - targets: ['saturn.renner.uno:9123']
-        labels:
-          group: 'ntp'
-          type: 'time-server'
-          hostname: 'saturn'
-
-  # Immich photo management metrics
-  - job_name: 'immich'
-    static_configs:
-      - targets: ['immich.srv.renner.uno:8081']
-        labels:
-          group: 'media'
-          type: 'photo-management'
-
-  # Blackbox Exporter - HTTP endpoint probing
-  - job_name: 'blackbox-http'
-    metrics_path: /probe
-    params:
-      module: [http_2xx]
-    static_configs:
-      - targets:
-          - https://gitea.renner.uno
-          - https://immich.renner.uno
-          - https://auth.renner.uno
-          - https://ha.renner.uno
-          - https://grafana.renner.uno
-          - https://prometheus.renner.uno
-        labels:
-          group: 'probe'
-          type: 'http'
-    relabel_configs:
-      - source_labels: [__address__]
-        target_label: __param_target
-      - source_labels: [__param_target]
-        target_label: instance
-      - target_label: __address__
-        replacement: localhost:9115
-
-  # Blackbox Exporter - TCP probing
-  - job_name: 'blackbox-tcp'
-    metrics_path: /probe
-    params:
-      module: [tcp_connect]
-    static_configs:
-      - targets:
-          - 10.2.40.135:5432  # PostgreSQL HA VIP
-        labels:
-          group: 'probe'
-          type: 'tcp'
-    relabel_configs:
-      - source_labels: [__address__]
-        target_label: __param_target
-      - source_labels: [__param_target]
-        target_label: instance
-      - target_label: __address__
-        replacement: localhost:9115
-
-  # Blackbox Exporter - DNS probing (external domains like google.com)
-  - job_name: 'blackbox-dns-external'
-    metrics_path: /probe
-    params:
-      module: [dns_external]
-    static_configs:
-      - targets:
-          - 10.2.40.53  # Proxmox DNS VIP
-          - 10.2.40.1   # UDM Gateway DNS
-        labels:
-          group: 'probe'
-          type: 'dns'
-          query_type: 'external'
-    relabel_configs:
-      - source_labels: [__address__]
-        target_label: __param_target
-      - source_labels: [__param_target]
-        target_label: instance
-      - target_label: __address__
-        replacement: localhost:9115
-
-  # Blackbox Exporter - DNS probing (internal domains like k8s-1.srv.renner.uno)
-  - job_name: 'blackbox-dns-internal'
-    metrics_path: /probe
-    params:
-      module: [dns_internal]
-    static_configs:
-      - targets:
-          - 10.2.40.53  # Proxmox DNS VIP
-          - 10.2.40.1   # UDM Gateway DNS
-        labels:
-          group: 'probe'
-          type: 'dns'
-          query_type: 'internal'
-    relabel_configs:
-      - source_labels: [__address__]
-        target_label: __param_target
-      - source_labels: [__param_target]
-        target_label: instance
-      - target_label: __address__
-        replacement: localhost:9115
-
-  # Blackbox Exporter - DNS probing (cached domains)
-  - job_name: 'blackbox-dns-cached'
-    metrics_path: /probe
-    params:
-      module: [dns_cached]
-    static_configs:
-      - targets:
-          - 10.2.40.53  # Proxmox DNS VIP
-          - 10.2.40.1   # UDM Gateway DNS
-        labels:
-          group: 'probe'
-          type: 'dns'
-          query_type: 'cached'
-    relabel_configs:
-      - source_labels: [__address__]
-        target_label: __param_target
-      - source_labels: [__param_target]
-        target_label: instance
-      - target_label: __address__
-        replacement: localhost:9115
-
-{% if snmp_synology_targets is defined and snmp_synology_targets | length > 0 %}
-  # Synology NAS SNMP metrics
-  # snmp_exporter v0.27+ requires both synology and ucd_la_table modules
-  - job_name: 'synology'
-    scrape_interval: 60s
-    scrape_timeout: 55s
-    metrics_path: /snmp
-    params:
-      module: [synology, ucd_la_table, ucd_memory, if_mib]
-{% if snmp_use_v3 | default(false) %}
-      auth: [synology_v3]
-{% else %}
-      auth: [synology_v2]
-{% endif %}
-    static_configs:
-      - targets:
-{% for target in snmp_synology_targets %}
-          - '{{ target.target }}'  # {{ target.name }}
-{% endfor %}
-        labels:
-          group: 'storage'
-          type: 'nas'
-    relabel_configs:
-      - source_labels: [__address__]
-        target_label: __param_target
-      - source_labels: [__address__]
-        regex: '([^:]+)(:\d+)?'
-        replacement: '${1}'
-        target_label: instance
-      - target_label: __address__
-        replacement: localhost:9116  # snmp_exporter address
-{% endif %}

ansible/roles/promtail/defaults/main.yml

@@ -1,36 +0,0 @@
----
-# Promtail Configuration
-
-# Version and download
-promtail_version: "3.4.2"
-promtail_arch: "linux-amd64"
-promtail_download_url: "https://github.com/grafana/loki/releases/download/v{{ promtail_version }}/promtail-{{ promtail_arch }}.zip"
-promtail_bin_path: "/usr/local/bin/promtail"
-
-# Directories
-promtail_config_dir: "/etc/promtail"
-promtail_config_file: "{{ promtail_config_dir }}/promtail-config.yaml"
-promtail_positions_file: "/var/lib/promtail/positions.yaml"
-
-# Server configuration
-promtail_http_listen_port: 9080
-promtail_http_listen_address: "0.0.0.0"
-
-# Loki server
-promtail_loki_url: "http://loki.srv.renner.uno:3100/loki/api/v1/push"
-
-# Log scrape configurations
-# Override in playbook or host_vars for specific services
-promtail_scrape_configs:
-  - job_name: system
-    static_configs:
-      - targets:
-          - localhost
-        labels:
-          job: syslog
-          host: "{{ inventory_hostname }}"
-          __path__: /var/log/syslog
-
-# Systemd service
-promtail_user: "promtail"
-promtail_group: "promtail"

ansible/roles/promtail/handlers/main.yml

@@ -1,6 +0,0 @@
----
-- name: Restart promtail
-  systemd:
-    name: promtail
-    state: restarted
-    daemon_reload: yes

ansible/roles/promtail/tasks/main.yml

@@ -1,112 +0,0 @@
----
-- name: Install dependencies
-  apt:
-    name:
-      - unzip
-      - curl
-    state: present
-    update_cache: yes
-
-- name: Create promtail group
-  group:
-    name: "{{ promtail_group }}"
-    system: yes
-    state: present
-
-- name: Create promtail user
-  user:
-    name: "{{ promtail_user }}"
-    group: "{{ promtail_group }}"
-    system: yes
-    shell: /usr/sbin/nologin
-    home: /var/lib/promtail
-    create_home: yes
-    state: present
-
-- name: Add promtail user to systemd-journal group
-  user:
-    name: "{{ promtail_user }}"
-    groups: systemd-journal
-    append: yes
-
-- name: Create config directory
-  file:
-    path: "{{ promtail_config_dir }}"
-    state: directory
-    owner: "{{ promtail_user }}"
-    group: "{{ promtail_group }}"
-    mode: '0755'
-
-- name: Create positions directory
-  file:
-    path: "{{ promtail_positions_file | dirname }}"
-    state: directory
-    owner: "{{ promtail_user }}"
-    group: "{{ promtail_group }}"
-    mode: '0755'
-
-- name: Check if promtail binary exists
-  stat:
-    path: "{{ promtail_bin_path }}"
-  register: promtail_binary
-
-- name: Get installed promtail version
-  command: "{{ promtail_bin_path }} --version"
-  register: installed_version
-  changed_when: false
-  failed_when: false
-  when: promtail_binary.stat.exists
-
-- name: Download promtail
-  get_url:
-    url: "{{ promtail_download_url }}"
-    dest: /tmp/promtail.zip
-    mode: '0644'
-  when: not promtail_binary.stat.exists or (installed_version.stdout is defined and promtail_version not in installed_version.stdout)
-
-- name: Extract promtail binary
-  unarchive:
-    src: /tmp/promtail.zip
-    dest: /tmp/
-    remote_src: yes
-  when: not promtail_binary.stat.exists or (installed_version.stdout is defined and promtail_version not in installed_version.stdout)
-
-- name: Install promtail binary
-  copy:
-    src: "/tmp/promtail-{{ promtail_arch }}"
-    dest: "{{ promtail_bin_path }}"
-    remote_src: yes
-    mode: '0755'
-  when: not promtail_binary.stat.exists or (installed_version.stdout is defined and promtail_version not in installed_version.stdout)
-  notify: Restart promtail
-
-- name: Clean up downloaded files
-  file:
-    path: "{{ item }}"
-    state: absent
-  loop:
-    - /tmp/promtail.zip
-    - "/tmp/promtail-{{ promtail_arch }}"
-
-- name: Configure promtail
-  template:
-    src: promtail-config.yaml.j2
-    dest: "{{ promtail_config_file }}"
-    owner: "{{ promtail_user }}"
-    group: "{{ promtail_group }}"
-    mode: '0644'
-  notify: Restart promtail
-
-- name: Create systemd service
-  template:
-    src: promtail.service.j2
-    dest: /etc/systemd/system/promtail.service
-    mode: '0644'
-  notify: Restart promtail
-
-- name: Enable and start promtail
-  systemd:
-    name: promtail
-    state: started
-    enabled: yes
-    daemon_reload: yes

ansible/roles/promtail/templates/promtail-config.yaml.j2

@@ -1,79 +0,0 @@
-server:
-  http_listen_port: {{ promtail_http_listen_port }}
-  http_listen_address: {{ promtail_http_listen_address }}
-  grpc_listen_port: 0
-
-positions:
-  filename: {{ promtail_positions_file }}
-
-clients:
-  - url: {{ promtail_loki_url }}
-    tenant_id: default
-
-scrape_configs:
-{% for config in promtail_scrape_configs %}
-  - job_name: {{ config.job_name }}
-{% if config.journal is defined %}
-    journal:
-      max_age: {{ config.journal.max_age | default('12h') }}
-      labels:
-        job: {{ config.job_name }}
-        host: {{ inventory_hostname }}
-{% if config.journal.matches is defined %}
-      matches: {{ config.journal.matches }}
-{% endif %}
-{% if config.relabel_configs is defined %}
-    relabel_configs:
-{% for relabel in config.relabel_configs %}
-      - source_labels: {{ relabel.source_labels | to_json }}
-        target_label: {{ relabel.target_label }}
-{% if relabel.regex is defined %}
-        regex: '{{ relabel.regex }}'
-{% endif %}
-{% if relabel.replacement is defined %}
-        replacement: '{{ relabel.replacement }}'
-{% endif %}
-{% endfor %}
-{% endif %}
-{% endif %}
-{% if config.static_configs is defined %}
-    static_configs:
-{% for static in config.static_configs %}
-      - targets:
-{% for target in static.targets %}
-          - {{ target }}
-{% endfor %}
-        labels:
-{% for key, value in static.labels.items() %}
-          {{ key }}: {{ value }}
-{% endfor %}
-{% endfor %}
-{% if config.pipeline_stages is defined %}
-    pipeline_stages:
-{% for stage in config.pipeline_stages %}
-{% if stage.multiline is defined %}
-      - multiline:
-          firstline: '{{ stage.multiline.firstline }}'
-{% if stage.multiline.max_wait_time is defined %}
-          max_wait_time: {{ stage.multiline.max_wait_time }}
-{% endif %}
-{% endif %}
-{% if stage.regex is defined %}
-      - regex:
-          expression: '{{ stage.regex.expression }}'
-{% endif %}
-{% if stage.labels is defined %}
-      - labels:
-{% for label in stage.labels %}
-          {{ label }}:
-{% endfor %}
-{% endif %}
-{% if stage.timestamp is defined %}
-      - timestamp:
-          source: {{ stage.timestamp.source }}
-          format: '{{ stage.timestamp.format }}'
-{% endif %}
-{% endfor %}
-{% endif %}
-{% endif %}
-{% endfor %}

ansible/roles/promtail/templates/promtail.service.j2

@@ -1,24 +0,0 @@
-[Unit]
-Description=Promtail Log Collector
-Documentation=https://grafana.com/docs/loki/latest/clients/promtail/
-After=network-online.target
-Wants=network-online.target
-
-[Service]
-Type=simple
-User={{ promtail_user }}
-Group={{ promtail_group }}
-ExecStart={{ promtail_bin_path }} -config.file={{ promtail_config_file }}
-Restart=always
-RestartSec=5
-TimeoutStopSec=30
-
-# Security hardening
-NoNewPrivileges=yes
-ProtectSystem=strict
-ProtectHome=yes
-ReadWritePaths=/var/lib/promtail
-CapabilityBoundingSet=CAP_DAC_READ_SEARCH
-
-[Install]
-WantedBy=multi-user.target

ansible/roles/pve-exporter/defaults/main.yml

@@ -1,19 +0,0 @@
----
-# PVE Exporter defaults
-
-# Installation
-pve_exporter_version: "3.4.5"
-pve_exporter_install_dir: /opt/pve-exporter
-pve_exporter_config_dir: /etc/pve-exporter
-pve_exporter_venv_path: "{{ pve_exporter_install_dir }}/venv"
-
-# Service
-pve_exporter_user: pve-exporter
-pve_exporter_group: pve-exporter
-pve_exporter_port: 9221
-
-# Proxmox connection (override with vault variables)
-pve_exporter_pve_user: "{{ pve_user | default('root@pam') }}"
-pve_exporter_pve_token_name: "{{ pve_token_id | default('ansible') }}"
-pve_exporter_pve_token_value: "{{ pve_token_secret }}"
-pve_exporter_pve_verify_ssl: false

ansible/roles/pve-exporter/handlers/main.yml

@@ -1,6 +0,0 @@
----
-- name: Restart pve-exporter
-  systemd:
-    name: pve-exporter
-    state: restarted
-    daemon_reload: yes

ansible/roles/pve-exporter/tasks/main.yml

@@ -1,78 +0,0 @@
----
-# PVE Exporter installation and configuration
-
-- name: Install Python dependencies
-  apt:
-    name:
-      - python3
-      - python3-pip
-      - python3-venv
-    state: present
-    update_cache: yes
-
-- name: Create pve-exporter system user
-  user:
-    name: "{{ pve_exporter_user }}"
-    system: yes
-    shell: /usr/sbin/nologin
-    create_home: no
-
-- name: Create installation directory
-  file:
-    path: "{{ pve_exporter_install_dir }}"
-    state: directory
-    owner: "{{ pve_exporter_user }}"
-    group: "{{ pve_exporter_group }}"
-    mode: '0755'
-
-- name: Create configuration directory
-  file:
-    path: "{{ pve_exporter_config_dir }}"
-    state: directory
-    owner: "{{ pve_exporter_user }}"
-    group: "{{ pve_exporter_group }}"
-    mode: '0750'
-
-- name: Create Python virtual environment
-  command: python3 -m venv {{ pve_exporter_venv_path }}
-  args:
-    creates: "{{ pve_exporter_venv_path }}/bin/python"
-
-- name: Set venv ownership
-  file:
-    path: "{{ pve_exporter_venv_path }}"
-    state: directory
-    owner: "{{ pve_exporter_user }}"
-    group: "{{ pve_exporter_group }}"
-    recurse: yes
-
-- name: Install prometheus-pve-exporter
-  pip:
-    name: prometheus-pve-exporter
-    virtualenv: "{{ pve_exporter_venv_path }}"
-    state: present
-
-- name: Create configuration file
-  template:
-    src: pve.yml.j2
-    dest: "{{ pve_exporter_config_dir }}/pve.yml"
-    owner: "{{ pve_exporter_user }}"
-    group: "{{ pve_exporter_group }}"
-    mode: '0640'
-  notify: Restart pve-exporter
-
-- name: Create systemd service
-  template:
-    src: pve-exporter.service.j2
-    dest: /etc/systemd/system/pve-exporter.service
-    owner: root
-    group: root
-    mode: '0644'
-  notify: Restart pve-exporter
-
-- name: Enable and start pve-exporter service
-  systemd:
-    name: pve-exporter
-    state: started
-    enabled: yes
-    daemon_reload: yes

ansible/roles/pve-exporter/templates/pve-exporter.service.j2

@@ -1,22 +0,0 @@
-[Unit]
-Description=Prometheus PVE Exporter
-Documentation=https://github.com/prometheus-pve/prometheus-pve-exporter
-After=network-online.target
-Wants=network-online.target
-
-[Service]
-Type=simple
-User={{ pve_exporter_user }}
-Group={{ pve_exporter_group }}
-ExecStart={{ pve_exporter_venv_path }}/bin/pve_exporter \
-  --config.file {{ pve_exporter_config_dir }}/pve.yml \
-  --web.listen-address 0.0.0.0:{{ pve_exporter_port }}
-Restart=on-failure
-RestartSec=10s
-
-# Security hardening
-NoNewPrivileges=yes
-PrivateTmp=yes
-
-[Install]
-WantedBy=multi-user.target

ansible/roles/pve-exporter/templates/pve.yml.j2

@@ -1,8 +0,0 @@
-# PVE Exporter Configuration
-# Generated by Ansible
-
-default:
-  user: "{{ pve_exporter_pve_user }}"
-  token_name: "{{ pve_exporter_pve_token_name }}"
-  token_value: "{{ pve_exporter_pve_token_value }}"
-  verify_ssl: {{ pve_exporter_pve_verify_ssl | lower }}

ansible/roles/snmp-exporter/defaults/main.yml

@@ -1,33 +0,0 @@
----
-# SNMP Exporter defaults
-
-snmp_exporter_version: "0.30.1"
-snmp_exporter_listen_port: 9116
-snmp_exporter_user: "snmp_exporter"
-snmp_exporter_group: "snmp_exporter"
-
-snmp_exporter_install_dir: "/opt/snmp_exporter"
-snmp_exporter_config_dir: "/etc/snmp_exporter"
-
-# Synology NAS SNMP targets
-# Configure in host_vars or group_vars
-snmp_synology_targets: []
-# Example:
-# snmp_synology_targets:
-#   - name: nas-1
-#     target: 10.2.40.50
-#   - name: nas-2
-#     target: 10.2.40.51
-
-# SNMPv3 authentication (recommended for security)
-# These should be set in vault
-snmp_v3_username: "prometheus"
-snmp_v3_auth_protocol: "SHA"  # MD5 or SHA
-snmp_v3_auth_password: ""     # Set in vault
-snmp_v3_priv_protocol: "AES"  # DES or AES
-snmp_v3_priv_password: ""     # Set in vault
-snmp_v3_security_level: "authPriv"  # noAuthNoPriv, authNoPriv, authPriv
-
-# SNMPv2c fallback (less secure, but simpler)
-snmp_v2_community: "public"
-snmp_use_v3: true

ansible/roles/snmp-exporter/handlers/main.yml

@@ -1,6 +0,0 @@
----
-- name: Restart snmp_exporter
-  systemd:
-    name: snmp_exporter
-    state: restarted
-    daemon_reload: yes

ansible/roles/snmp-exporter/tasks/main.yml

@@ -1,114 +0,0 @@
----
-# SNMP Exporter installation and configuration
-
-- name: Create snmp_exporter group
-  group:
-    name: "{{ snmp_exporter_group }}"
-    system: yes
-    state: present
-
-- name: Create snmp_exporter user
-  user:
-    name: "{{ snmp_exporter_user }}"
-    group: "{{ snmp_exporter_group }}"
-    system: yes
-    shell: /usr/sbin/nologin
-    home: "{{ snmp_exporter_install_dir }}"
-    create_home: no
-
-- name: Create directories
-  file:
-    path: "{{ item }}"
-    state: directory
-    owner: "{{ snmp_exporter_user }}"
-    group: "{{ snmp_exporter_group }}"
-    mode: '0755'
-  loop:
-    - "{{ snmp_exporter_install_dir }}"
-    - "{{ snmp_exporter_config_dir }}"
-
-- name: Check if snmp_exporter is already installed
-  stat:
-    path: "{{ snmp_exporter_install_dir }}/snmp_exporter"
-  register: snmp_exporter_binary
-
-- name: Get installed snmp_exporter version
-  command: "{{ snmp_exporter_install_dir }}/snmp_exporter --version"
-  register: installed_version
-  changed_when: false
-  failed_when: false
-  when: snmp_exporter_binary.stat.exists
-
-- name: Download snmp_exporter
-  get_url:
-    url: "https://github.com/prometheus/snmp_exporter/releases/download/v{{ snmp_exporter_version }}/snmp_exporter-{{ snmp_exporter_version }}.linux-amd64.tar.gz"
-    dest: "/tmp/snmp_exporter.tar.gz"
-    mode: '0644'
-  when: not snmp_exporter_binary.stat.exists or (installed_version.stdout is defined and snmp_exporter_version not in installed_version.stdout)
-
-- name: Extract snmp_exporter
-  unarchive:
-    src: "/tmp/snmp_exporter.tar.gz"
-    dest: "/tmp/"
-    remote_src: yes
-  when: not snmp_exporter_binary.stat.exists or (installed_version.stdout is defined and snmp_exporter_version not in installed_version.stdout)
-
-- name: Install snmp_exporter binary
-  copy:
-    src: "/tmp/snmp_exporter-{{ snmp_exporter_version }}.linux-amd64/snmp_exporter"
-    dest: "{{ snmp_exporter_install_dir }}/snmp_exporter"
-    remote_src: yes
-    owner: "{{ snmp_exporter_user }}"
-    group: "{{ snmp_exporter_group }}"
-    mode: '0755'
-  when: not snmp_exporter_binary.stat.exists or (installed_version.stdout is defined and snmp_exporter_version not in installed_version.stdout)
-  notify: Restart snmp_exporter
-
-- name: Install default snmp.yml from release
-  copy:
-    src: "/tmp/snmp_exporter-{{ snmp_exporter_version }}.linux-amd64/snmp.yml"
-    dest: "{{ snmp_exporter_config_dir }}/snmp.yml"
-    remote_src: yes
-    owner: "{{ snmp_exporter_user }}"
-    group: "{{ snmp_exporter_group }}"
-    mode: '0644'
-  when: not snmp_exporter_binary.stat.exists or (installed_version.stdout is defined and snmp_exporter_version not in installed_version.stdout)
-  notify: Restart snmp_exporter
-
-- name: Add synology_v3 auth entry
-  lineinfile:
-    path: "{{ snmp_exporter_config_dir }}/snmp.yml"
-    insertafter: "^auths:"
-    line: "  synology_v3: {version: 3, security_level: {{ snmp_v3_security_level }}, username: {{ snmp_v3_username }}, password: {{ snmp_v3_auth_password }}, auth_protocol: {{ snmp_v3_auth_protocol }}, priv_protocol: {{ snmp_v3_priv_protocol }}, priv_password: {{ snmp_v3_priv_password }}}"
-  notify: Restart snmp_exporter
-
-- name: Add synology_v2 auth entry
-  lineinfile:
-    path: "{{ snmp_exporter_config_dir }}/snmp.yml"
-    insertafter: "^auths:"
-    line: "  synology_v2: {version: 2, community: {{ snmp_v2_community }}}"
-  notify: Restart snmp_exporter
-
-- name: Create snmp_exporter systemd service
-  template:
-    src: snmp_exporter.service.j2
-    dest: /etc/systemd/system/snmp_exporter.service
-    owner: root
-    group: root
-    mode: '0644'
-  notify: Restart snmp_exporter
-
-- name: Enable and start snmp_exporter service
-  systemd:
-    name: snmp_exporter
-    state: started
-    enabled: yes
-    daemon_reload: yes
-
-- name: Clean up downloaded files
-  file:
-    path: "{{ item }}"
-    state: absent
-  loop:
-    - "/tmp/snmp_exporter.tar.gz"
-    - "/tmp/snmp_exporter-{{ snmp_exporter_version }}.linux-amd64"

ansible/roles/snmp-exporter/templates/snmp-auth.yml.j2

@@ -1,18 +0,0 @@
-# SNMPv3 Authentication for Synology NAS
-# Appended to default snmp.yml by Ansible
-
-auths:
-{% if snmp_use_v3 | default(true) %}
-  synology_v3:
-    security_level: {{ snmp_v3_security_level }}
-    username: {{ snmp_v3_username }}
-    password: {{ snmp_v3_auth_password }}
-    auth_protocol: {{ snmp_v3_auth_protocol }}
-{% if snmp_v3_security_level == 'authPriv' %}
-    priv_protocol: {{ snmp_v3_priv_protocol }}
-    priv_password: {{ snmp_v3_priv_password }}
-{% endif %}
-{% else %}
-  synology_v2:
-    community: {{ snmp_v2_community }}
-{% endif %}

ansible/roles/snmp-exporter/templates/snmp_exporter.service.j2

@@ -1,26 +0,0 @@
-[Unit]
-Description=Prometheus SNMP Exporter
-Documentation=https://github.com/prometheus/snmp_exporter
-After=network-online.target
-Wants=network-online.target
-
-[Service]
-Type=simple
-User={{ snmp_exporter_user }}
-Group={{ snmp_exporter_group }}
-ExecStart={{ snmp_exporter_install_dir }}/snmp_exporter \
-  --config.file={{ snmp_exporter_config_dir }}/snmp.yml \
-  --web.listen-address=:{{ snmp_exporter_listen_port }}
-
-Restart=on-failure
-RestartSec=5s
-
-# Security hardening
-NoNewPrivileges=yes
-ProtectSystem=strict
-ProtectHome=yes
-PrivateTmp=yes
-ReadOnlyPaths=/
-
-[Install]
-WantedBy=multi-user.target

ansible/roles/unifi-poller/defaults/main.yml

@@ -1,34 +0,0 @@
----
-# UniFi Poller defaults
-
-# Version to install
-unifi_poller_version: "2.11.2"
-
-# Installation paths
-unifi_poller_bin_path: /usr/local/bin/unpoller
-unifi_poller_config_dir: /etc/unpoller
-unifi_poller_config_file: "{{ unifi_poller_config_dir }}/up.conf"
-
-# Service user
-unifi_poller_user: unpoller
-unifi_poller_group: unpoller
-
-# UniFi Controller settings (override in vars/vault)
-unifi_controller_url: "https://192.168.1.1"
-unifi_controller_user: ""
-unifi_controller_pass: ""
-unifi_controller_verify_ssl: false
-unifi_controller_site: "default"
-
-# Prometheus exporter settings
-unifi_poller_prometheus_port: 9130
-unifi_poller_prometheus_path: /metrics
-
-# Polling intervals
-unifi_poller_interval: 30s
-unifi_poller_save_sites: true
-unifi_poller_save_ids: false
-unifi_poller_save_events: false
-unifi_poller_save_alarms: false
-unifi_poller_save_anomalies: false
-unifi_poller_save_dpi: true

ansible/roles/unifi-poller/handlers/main.yml

@@ -1,6 +0,0 @@
----
-- name: Restart unpoller
-  systemd:
-    name: unpoller
-    state: restarted
-    daemon_reload: yes

ansible/roles/unifi-poller/tasks/main.yml

@@ -1,88 +0,0 @@
----
-# UniFi Poller installation and configuration
-
-- name: Create unpoller system user
-  user:
-    name: "{{ unifi_poller_user }}"
-    system: yes
-    shell: /usr/sbin/nologin
-    create_home: no
-
-- name: Create configuration directory
-  file:
-    path: "{{ unifi_poller_config_dir }}"
-    state: directory
-    owner: "{{ unifi_poller_user }}"
-    group: "{{ unifi_poller_group }}"
-    mode: '0750'
-
-- name: Check if unpoller is already installed
-  stat:
-    path: "{{ unifi_poller_bin_path }}"
-  register: unpoller_binary
-
-- name: Get installed version
-  command: "{{ unifi_poller_bin_path }} --version"
-  register: installed_version
-  changed_when: false
-  failed_when: false
-  when: unpoller_binary.stat.exists
-
-- name: Download UniFi Poller
-  get_url:
-    url: "https://github.com/unpoller/unpoller/releases/download/v{{ unifi_poller_version }}/unpoller_{{ unifi_poller_version }}_linux_amd64.tar.gz"
-    dest: /tmp/unpoller.tar.gz
-    mode: '0644'
-  when: not unpoller_binary.stat.exists or (installed_version.stdout is defined and unifi_poller_version not in installed_version.stdout)
-
-- name: Extract UniFi Poller
-  unarchive:
-    src: /tmp/unpoller.tar.gz
-    dest: /tmp/
-    remote_src: yes
-  when: not unpoller_binary.stat.exists or (installed_version.stdout is defined and unifi_poller_version not in installed_version.stdout)
-
-- name: Install UniFi Poller binary
-  copy:
-    src: /tmp/unpoller
-    dest: "{{ unifi_poller_bin_path }}"
-    remote_src: yes
-    owner: root
-    group: root
-    mode: '0755'
-  when: not unpoller_binary.stat.exists or (installed_version.stdout is defined and unifi_poller_version not in installed_version.stdout)
-  notify: Restart unpoller
-
-- name: Create configuration file
-  template:
-    src: up.conf.j2
-    dest: "{{ unifi_poller_config_file }}"
-    owner: "{{ unifi_poller_user }}"
-    group: "{{ unifi_poller_group }}"
-    mode: '0640'
-  notify: Restart unpoller
-
-- name: Create systemd service
-  template:
-    src: unpoller.service.j2
-    dest: /etc/systemd/system/unpoller.service
-    owner: root
-    group: root
-    mode: '0644'
-  notify: Restart unpoller
-
-- name: Enable and start unpoller service
-  systemd:
-    name: unpoller
-    state: started
-    enabled: yes
-    daemon_reload: yes
-
-- name: Clean up downloaded files
-  file:
-    path: "{{ item }}"
-    state: absent
-  loop:
-    - /tmp/unpoller.tar.gz
-    - /tmp/unpoller
-  changed_when: false

ansible/roles/unifi-poller/templates/unpoller.service.j2

@@ -1,24 +0,0 @@
-[Unit]
-Description=UniFi Poller - Prometheus Exporter for UniFi Network
-Documentation=https://github.com/unpoller/unpoller
-After=network-online.target
-Wants=network-online.target
-
-[Service]
-Type=simple
-User={{ unifi_poller_user }}
-Group={{ unifi_poller_group }}
-ExecStart={{ unifi_poller_bin_path }} --config {{ unifi_poller_config_file }}
-Restart=on-failure
-RestartSec=10s
-
-# Security hardening
-NoNewPrivileges=yes
-ProtectSystem=strict
-ProtectHome=yes
-ReadOnlyPaths=/
-ReadWritePaths={{ unifi_poller_config_dir }}
-PrivateTmp=yes
-
-[Install]
-WantedBy=multi-user.target

ansible/roles/unifi-poller/templates/up.conf.j2

@@ -1,42 +0,0 @@
-# UniFi Poller Configuration
-# Generated by Ansible
-
-[poller]
-  # Turns on debug messages
-  debug = false
-  # Turns off messages
-  quiet = false
-
-[prometheus]
-  # Whether to export metrics
-  disable = false
-  # HTTP listen address for /metrics endpoint
-  http_listen = "0.0.0.0:{{ unifi_poller_prometheus_port }}"
-  # Path for metrics endpoint
-  report_errors = false
-
-[unifi]
-  # UniFi Controller configurations
-  [[unifi.controller]]
-    # URL for the UniFi Controller
-    url = "{{ unifi_controller_url }}"
-    # Username
-    user = "{{ unifi_controller_user }}"
-    # Password
-    pass = "{{ unifi_controller_pass }}"
-    # Sites to poll (leave empty for all sites)
-    sites = ["{{ unifi_controller_site }}"]
-    # Verify SSL certificates
-    verify_ssl = {{ unifi_controller_verify_ssl | lower }}
-    # Save site information
-    save_sites = {{ unifi_poller_save_sites | lower }}
-    # Save device IDs
-    save_ids = {{ unifi_poller_save_ids | lower }}
-    # Save events
-    save_events = {{ unifi_poller_save_events | lower }}
-    # Save alarms
-    save_alarms = {{ unifi_poller_save_alarms | lower }}
-    # Save anomalies
-    save_anomalies = {{ unifi_poller_save_anomalies | lower }}
-    # Save DPI data
-    save_dpi = {{ unifi_poller_save_dpi | lower }}

kubernetes/apps/authentik/blueprints-configmap.yaml

@@ -1326,7 +1326,121 @@ data:
           name: Caddy docs.renner.uno
         state: absent
 
-  # Passkey-only (WebAuthn) login for selected apps, in two parts:
+      # ---------------------------------------------------------------
+      # wordpress-gate — www.renner.uno
+      #
+      # Unlike every other gate here, this one does NOT front a whole
+      # host. www.renner.uno is a public family blog; only /wp-admin and
+      # /wp-login.php are routed to this outpost (see
+      # kubernetes/apps/wordpress/ingress.yaml). The front end never
+      # touches it.
+      #
+      # The reverse carve-out — gate everything, skip_path_regex the
+      # public parts, as paperless-gate does for /static/ — is not
+      # possible here: the public set is every post URL, so the pattern
+      # would need a negative lookahead and the outpost's RE2 has none.
+      - id: wordpress-authors-group
+        model: authentik_core.group
+        identifiers:
+          name: WordPress Authors
+        state: present
+        attrs:
+          is_superuser: false
+          # Git owns membership, as with the other gates: this group is the
+          # explicit allowlist for who may publish on www.renner.uno.
+          # Add/remove members HERE, not in the UI — blueprint apply is
+          # authoritative for this M2M and reverts UI edits.
+          #
+          # s.renner is the Authentik account for the site's WordPress
+          # Editor, which still carries the pre-2019 login `s.birzer` — the
+          # two names do not match, and that is expected.
+          users:
+            - !Find [authentik_core.user, [username, k.renner]]
+            - !Find [authentik_core.user, [username, s.renner]]
+
+      - id: wordpress-gate-provider
+        model: authentik_providers_proxy.proxyprovider
+        identifiers:
+          name: wordpress-gate
+        state: present
+        attrs:
+          authorization_flow: !Find [authentik_flows.flow, [slug, default-provider-authorization-implicit-consent]]
+          invalidation_flow: !Find [authentik_flows.flow, [slug, default-provider-invalidation-flow]]
+          mode: proxy
+          internal_host: http://wordpress.wordpress.svc.cluster.local
+          internal_host_ssl_validation: false
+          external_host: https://www.renner.uno
+          cookie_domain: renner.uno
+          access_token_validity: hours=24
+          refresh_token_validity: days=30
+          intercept_header_auth: true
+          # No skip_path_regex: wp-admin's assets are only ever fetched by an
+          # already-authenticated session, and wp-admin has no manifest or
+          # service worker — the credentials:'omit' PWA problem that forced
+          # carve-outs on vscode-gate and retired music-assistant-gate does
+          # not arise here.
+
+      - model: authentik_core.application
+        identifiers:
+          slug: wordpress-gate
+        state: present
+        attrs:
+          name: WordPress (Gate)
+          slug: wordpress-gate
+          provider: !KeyOf wordpress-gate-provider
+          meta_launch_url: https://www.renner.uno/wp-admin/
+          meta_description: "Forward-auth gate in front of wp-admin on www.renner.uno"
+
+      # Only WordPress Authors pass the gate — see the group above for the
+      # membership list. Both accounts that exist on the migrated site are
+      # covered, so the gate cannot lock out the editor the site is
+      # currently being kept alive for.
+      - model: authentik_policies.policybinding
+        identifiers:
+          target: !Find [authentik_core.application, [slug, wordpress-gate]]
+          group: !KeyOf wordpress-authors-group
+          order: 0
+        state: present
+        attrs:
+          enabled: true
+          negate: false
+          failure_result: false
+
+      - model: authentik_outposts.outpost
+        identifiers:
+          name: wordpress-gate
+        state: present
+        attrs:
+          type: proxy
+          service_connection: !KeyOf k8s-conn
+          providers:
+            - !KeyOf wordpress-gate-provider
+          config:
+            log_level: info
+            authentik_host: "http://authentik-server.authentik.svc.cluster.local"
+            authentik_host_insecure: false
+            authentik_host_browser: "https://auth.renner.uno/"
+            refresh_interval: "minutes=5"
+            container_image: null
+            object_naming_template: "ak-outpost-%(name)s"
+            kubernetes_namespace: authentik
+            kubernetes_replicas: 1
+            kubernetes_service_type: ClusterIP
+            kubernetes_disabled_components: ["ingress"]
+            kubernetes_image_pull_secrets: []
+            kubernetes_ingress_annotations: {}
+            kubernetes_ingress_class_name: null
+            kubernetes_ingress_secret_name: "authentik-outpost-tls"
+            kubernetes_json_patches: null
+
+  # Passkey login for selected apps. Passkey is mandatory for accounts
+  # that hold one and unavailable to accounts that do not, so both flows
+  # branch on enrollment (policy passkey-user-has-webauthn): enrolled →
+  # WebAuthn, not enrolled → password + MFA. Enrollment status is a
+  # property of the identified account, so the weaker branch is never
+  # something a requester can select.
+  #
+  # Two parts:
   #
   # 1. A passwordless AUTHENTICATION flow (username → passkey, no
   #    password prompt) bound per-provider, so fresh logins to these
@@ -1335,7 +1449,13 @@ data:
   # 2. A step-up AUTHORIZATION flow — unlike the authentication flow,
   #    this runs on EVERY interactive authorize, so an existing
   #    password-only SSO session from another app cannot slip into
-  #    these apps without presenting a passkey.
+  #    these apps without presenting a passkey. Two things narrow that:
+  #    the stage is skipped outright for accounts with no passkey (they
+  #    have no way to satisfy it), and last_auth_threshold reduces it to
+  #    "without presenting a passkey in this browser within the window"
+  #    — one touch then covers every app, including a later
+  #    password-only session for the same user. The cookie is keyed to
+  #    that user's device, so it never carries across users.
   passkey-stepup.yaml: |
     version: 1
     metadata:
@@ -1356,43 +1476,99 @@ data:
           # flow this replaces.
           authentication: require_authenticated
 
-      - id: passkey-stepup-webauthn-validation
+      # ONE stage object, shared by the step-up authorization flow and
+      # the passwordless authentication flow below.
+      #
+      # `last_auth_threshold` is enforced with a single browser cookie
+      # (`authentik_mfa`) whose JWT is both signed with and carries the
+      # stage's PK, so a validation stage only ever honours a cookie it
+      # wrote itself. Two separate stage objects therefore each demand
+      # their own touch, and — sharing the one cookie name — overwrite
+      # each other, so neither window survives. Sharing this object is
+      # what lets the login touch satisfy the per-app step-up.
+      #
+      # Declared here, ahead of every !KeyOf reference to it: the
+      # blueprint importer resolves entries top-to-bottom.
+      - id: passkey-webauthn-validation
         model: authentik_stages_authenticator_validate.authenticatorvalidatestage
         identifiers:
-          name: passkey-stepup-webauthn-validation
+          name: passkey-webauthn-validation
         state: present
         attrs:
           device_classes:
             - webauthn
-          # `configure`, not `deny`: g.renner has Home Assistant access
-          # but no enrolled WebAuthn device (live check 2026-07-07), so
-          # a hard deny would lock her out with no self-service path.
-          # `configure` injects the default WebAuthn enrollment stage
-          # inline — first login without a passkey becomes "enroll one
-          # now, then validate it". Still passkey-only: there is no way
-          # through this stage without a WebAuthn credential.
-          not_configured_action: configure
-          configuration_stages:
-            - !Find [authentik_stages_authenticator_webauthn.authenticatorwebauthnstage, [name, default-authenticator-webauthn-setup]]
-          # Skip re-validation if a passkey was already presented within
-          # the window. Non-zero for two reasons: (1) no double-touch
-          # when the login itself used a passkey seconds earlier;
-          # (2) NetBird's dashboard silently re-authorizes via
-          # /silent-auth (prompt=none) on token expiry — a flow that
-          # always demands interaction would fail those and force a
-          # visible re-login every access-token lifetime.
+          # `deny`, not `configure`: the authentication flow reaches this
+          # stage with the user merely *identified* (a typed username)
+          # and no password stage in front of it, so inline enrollment
+          # would let anyone who knows a passkey-less username enroll
+          # their own passkey and take over that account. The policy
+          # binding means only accounts that already hold a passkey reach
+          # this stage at all, so `deny` is defence in depth rather than
+          # a path anyone walks — accounts without one branch to password
+          # + MFA instead.
+          not_configured_action: deny
+          # Window within which an already-presented passkey satisfies a
+          # later run of this stage. Non-zero for two reasons: (1) the
+          # per-app step-up must not demand a second touch seconds after
+          # the login touch; (2) NetBird's dashboard silently
+          # re-authorizes via /silent-auth (prompt=none) on token expiry —
+          # a stage that always demands interaction fails those and
+          # forces a visible re-login every access-token lifetime.
+          #
+          # The cookie is browser-scoped, so inside the window a re-login
+          # in that same browser is satisfied by the cookie alone and the
+          # username is all that gets typed. Accepted for a single touch
+          # per app.
           last_auth_threshold: hours=8
           webauthn_user_verification: preferred
 
-      - model: authentik_flows.flowstagebinding
+      # Gate on "this account has a passkey", so enrollment status picks
+      # the branch. Enrolled accounts must use WebAuthn — at login and at
+      # every authorize — while accounts with no passkey fall back to
+      # password + MFA. An attacker cannot pick the weaker branch: the
+      # user is already identified when this runs, so the branch follows
+      # the victim's enrollment, not the requester's choice.
+      #
+      # request.user is the flow's pending_user, not the (anonymous)
+      # requester: ReevaluateMarker builds the engine with
+      # plan.context[PLAN_CONTEXT_PENDING_USER] (flows/markers.py). That
+      # requires the bindings below to keep evaluate_on_plan false and
+      # re_evaluate_policies true — on-plan evaluation happens before the
+      # identification stage has resolved a user.
+      - id: policy-has-passkey
+        model: authentik_policies_expression.expressionpolicy
+        identifiers:
+          name: passkey-user-has-webauthn
+        state: present
+        attrs:
+          expression: |
+            return ak_user_has_authenticator(request.user, "webauthn")
+
+      - id: binding-stepup-webauthn
+        model: authentik_flows.flowstagebinding
         identifiers:
           target: !KeyOf passkey-stepup-flow
-          stage: !KeyOf passkey-stepup-webauthn-validation
+          stage: !KeyOf passkey-webauthn-validation
           order: 10
         state: present
+        attrs:
+          evaluate_on_plan: false
+          re_evaluate_policies: true
+
+      - model: authentik_policies.policybinding
+        identifiers:
+          policy: !KeyOf policy-has-passkey
+          target: !KeyOf binding-stepup-webauthn
+          order: 0
+        state: present
+        attrs:
+          enabled: true
+          negate: false
+          failure_result: false
 
       # Passwordless authentication flow: identification (username or
-      # email, NO password stage) → WebAuthn validation → login.
+      # email, NO password stage) → WebAuthn validation → login, with a
+      # password + MFA branch for accounts that have no passkey.
       - id: passkey-passwordless-flow
         model: authentik_flows.flow
         identifiers:
@@ -1403,37 +1579,6 @@ data:
           title: "Sign in with your passkey"
           designation: authentication
 
-      # SEPARATE stage from the step-up one, and deliberately `deny`:
-      # at this point in an authentication flow the user has only been
-      # *identified* (typed a username), not authenticated — `configure`
-      # here would let anyone who knows a passkey-less username enroll
-      # their own passkey and take over that account. Inline enrollment
-      # for passkey-less users happens exclusively in the authorization
-      # step-up stage above, where require_authenticated guarantees a
-      # fully authenticated session. A passkey-less user therefore
-      # bootstraps via password login on auth.renner.uno, then gets the
-      # enrollment prompt on first app access.
-      #
-      # Declared BEFORE the identification stage because the latter
-      # references it via !KeyOf (webauthn_stage), and the blueprint
-      # importer resolves entries top-to-bottom.
-      - id: passkey-passwordless-webauthn-validation
-        model: authentik_stages_authenticator_validate.authenticatorvalidatestage
-        identifiers:
-          name: passkey-passwordless-webauthn-validation
-        state: present
-        attrs:
-          device_classes:
-            - webauthn
-          not_configured_action: deny
-          # Small non-zero window so a passkey presented via the
-          # identification stage's autofill (webauthn_stage below) is
-          # not challenged a second time by this stage at order 20 in
-          # the same flow run. The skip is session-scoped, so a fresh
-          # (attacker) session never inherits it.
-          last_auth_threshold: minutes=5
-          webauthn_user_verification: preferred
-
       # webauthn_stage enables passkey autofill (conditional WebAuthn,
       # authentik >= 2025.12): the browser offers enrolled passkeys in
       # the username field's autofill — one tap, no username typed.
@@ -1449,7 +1594,7 @@ data:
           user_fields:
             - username
             - email
-          webauthn_stage: !KeyOf passkey-passwordless-webauthn-validation
+          webauthn_stage: !KeyOf passkey-webauthn-validation
 
       - model: authentik_flows.flowstagebinding
         identifiers:
@@ -1458,12 +1603,87 @@ data:
           order: 10
         state: present
 
-      - model: authentik_flows.flowstagebinding
+      # Order 20 and 21/22 are the two mutually exclusive branches.
+      - id: binding-passwordless-webauthn
+        model: authentik_flows.flowstagebinding
         identifiers:
           target: !KeyOf passkey-passwordless-flow
-          stage: !KeyOf passkey-passwordless-webauthn-validation
+          stage: !KeyOf passkey-webauthn-validation
           order: 20
         state: present
+        attrs:
+          evaluate_on_plan: false
+          re_evaluate_policies: true
+
+      - model: authentik_policies.policybinding
+        identifiers:
+          policy: !KeyOf policy-has-passkey
+          target: !KeyOf binding-passwordless-webauthn
+          order: 0
+        state: present
+        attrs:
+          enabled: true
+          negate: false
+          failure_result: false
+
+      # The negated branch. failure_result: true is the safety net — the
+      # UserLoginStage defaults a missing authentication backend to
+      # BACKEND_INBUILT instead of refusing (user_login/stage.py), so a
+      # flow that drops BOTH branches would log the user in on a typed
+      # username alone. A PolicyException is resolved to failure_result
+      # before negate is applied (policies/process.py), so a broken
+      # expression pulls the password stage IN and degrades everyone to
+      # password + MFA rather than to no factor at all.
+      - id: binding-passwordless-password
+        model: authentik_flows.flowstagebinding
+        identifiers:
+          target: !KeyOf passkey-passwordless-flow
+          stage: !Find [authentik_stages_password.passwordstage, [name, default-authentication-password]]
+          order: 21
+        state: present
+        attrs:
+          evaluate_on_plan: false
+          re_evaluate_policies: true
+
+      - model: authentik_policies.policybinding
+        identifiers:
+          policy: !KeyOf policy-has-passkey
+          target: !KeyOf binding-passwordless-password
+          order: 0
+        state: present
+        attrs:
+          enabled: true
+          negate: true
+          failure_result: true
+
+      # Same MFA validation the default authentication flow runs at its
+      # order 30, so the fallback is not a weaker way into these apps
+      # than auth.renner.uno's own login already is. It carries
+      # not_configured_action: configure, which is safe here and nowhere
+      # else in this flow: the password stage above has established a
+      # first factor by the time it runs, so it doubles as the enrollment
+      # path for accounts holding no authenticator at all.
+      - id: binding-passwordless-mfa
+        model: authentik_flows.flowstagebinding
+        identifiers:
+          target: !KeyOf passkey-passwordless-flow
+          stage: !Find [authentik_stages_authenticator_validate.authenticatorvalidatestage, [name, default-authentication-mfa-validation]]
+          order: 22
+        state: present
+        attrs:
+          evaluate_on_plan: false
+          re_evaluate_policies: true
+
+      - model: authentik_policies.policybinding
+        identifiers:
+          policy: !KeyOf policy-has-passkey
+          target: !KeyOf binding-passwordless-mfa
+          order: 0
+        state: present
+        attrs:
+          enabled: true
+          negate: true
+          failure_result: true
 
       - model: authentik_flows.flowstagebinding
         identifiers:
@@ -1577,3 +1797,19 @@ data:
           redirect_uris:
             - url: https://dash.renner.uno/login/generic_oauth
               matching_mode: strict
+
+      # Superseded by the single passkey-webauthn-validation stage above.
+      # Their flowstagebindings are FK-CASCADEd away with them; the
+      # identification stage's webauthn_stage FK is SET_NULL and is
+      # re-pointed above, so no dangling reference survives. Declared
+      # last: the importer applies entries top-to-bottom inside one
+      # transaction, so the replacements exist before these go.
+      - model: authentik_stages_authenticator_validate.authenticatorvalidatestage
+        identifiers:
+          name: passkey-stepup-webauthn-validation
+        state: absent
+
+      - model: authentik_stages_authenticator_validate.authenticatorvalidatestage
+        identifiers:
+          name: passkey-passwordless-webauthn-validation
+        state: absent

kubernetes/apps/authentik/kustomization.yaml

@@ -14,6 +14,7 @@ resources:
   - zone-configurator-gate-egress.yaml
   - alertmanager-gate-egress.yaml
   - prometheus-gate-egress.yaml
+  - wordpress-gate-egress.yaml
   - outpost-servicemonitors.yaml
   - ldap-outpost-cert.yaml
   - ldap-outpost-lb.yaml

kubernetes/apps/authentik/wordpress-gate-egress.yaml

@@ -0,0 +1,28 @@
+---
+# Allow the dedicated wordpress-gate outpost to egress into the wordpress
+# namespace and reach WordPress (:80). Without this, the authentik-namespace
+# baseline CNP has no egress rule to wordpress, so the gate outpost's
+# forwarded requests are dropped. Scoped to the gate outpost's pod label so no
+# other authentik component can reach it. Mirrors esphome-gate-egress.yaml.
+#
+# Note this gate only ever sees /wp-admin and /wp-login.php — the public site
+# is routed straight to the same Service by a separate Ingress, so this rule
+# is not the only path into the wordpress namespace, unlike the paperless case.
+apiVersion: cilium.io/v2
+kind: CiliumNetworkPolicy
+metadata:
+  name: wordpress-gate-egress
+  namespace: authentik
+spec:
+  endpointSelector:
+    matchLabels:
+      goauthentik.io/outpost-name: wordpress-gate
+  egress:
+    - toEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": wordpress
+            app.kubernetes.io/name: wordpress
+      toPorts:
+        - ports:
+            - port: "80"
+              protocol: TCP

kubernetes/apps/gatus/configmap.yaml

@@ -160,6 +160,24 @@ data:
         conditions:
           - "[STATUS] == any(200, 301, 302)"
 
+      # Terminus admin UI. Checks the TLS route only — the cleartext
+      # device endpoint (trmnl-device.renner.uno) is the same pod, so a
+      # failure here covers both. Terminus exposes no health endpoint, so
+      # this is a root-page check: the login page answers 200, and an
+      # unauthenticated request may redirect.
+      - name: Web Service 10
+        group: services
+        url: https://trmnl.renner.uno
+        interval: 60s
+        ui:
+          hide-hostname: true
+          hide-url: true
+          hide-port: true
+          hide-conditions: true
+          hide-errors: true
+        conditions:
+          - "[STATUS] == any(200, 301, 302)"
+
       # Removed 2026-05-21: Jellyfin (media.renner.uno), Wiki (wiki.renner.uno),
       # InfluxDB (influx.renner.uno) — hostnames NXDOMAIN, no Ingress; services
       # no longer deployed. Re-add when/if they come back.

kubernetes/apps/home-assistant/deployment.yaml

@@ -38,7 +38,7 @@ spec:
       # gets the heavy HA pod (8Gi) off the soon-to-be-tiny edge box.
       containers:
         - name: home-assistant
-          image: ghcr.io/home-assistant/home-assistant:2026.8.1 # {"$imagepolicy": "flux-system:home-assistant"}
+          image: ghcr.io/home-assistant/home-assistant:2026.8.2 # {"$imagepolicy": "flux-system:home-assistant"}
           env:
             - name: TZ
               value: Europe/Berlin

kubernetes/apps/ical-proxy/deployment.yaml

@@ -17,7 +17,33 @@ spec:
     spec:
       containers:
         - name: ical-proxy
-          image: ghcr.io/konairius/ical-proxy:latest
+          # Pinned to a commit build rather than :latest.
+          #
+          # :latest never rolls. The tag moves when a build publishes, but
+          # nothing tells the Deployment about it: the pod template is
+          # unchanged, so Kubernetes has no reason to replace anything, and
+          # this image is deliberately NOT enrolled in Flux image automation
+          # — see kubernetes/infrastructure/image-automation/README.md, which
+          # lists ghcr.io/konairius/ical-proxy as having "only main, latest,
+          # and main-<sha> commit builds; no release tags", so there is no
+          # ImagePolicy watching it. The pods therefore sat on a 21-day-old
+          # image while :latest moved underneath them, and the only way to
+          # pick up a new build was a manual restart.
+          #
+          # A pinned tag makes the rollout a git change: Flux applies the new
+          # pod template and Kubernetes rolls it, which is also what makes
+          # what is RUNNING recoverable from this repo rather than "whatever
+          # :latest meant the last time a pod happened to restart".
+          #
+          # main-99a912b is the merge of konairius/ical-proxy#12, which adds
+          # the summary_match/summary_replace parameters this deployment
+          # needs to strip "| Abfuhrkalender - Landkreis Amberg-Sulzbach"
+          # from every waste collection.
+          #
+          # Bumping this is a manual, reviewed edit. The tag format is
+          # main-<short-sha> (docker/metadata-action type=sha with the branch
+          # as prefix).
+          image: ghcr.io/konairius/ical-proxy:main-99a912b
           ports:
             - containerPort: 8080
               name: http

kubernetes/apps/music-assistant/deployment.yaml

@@ -52,7 +52,7 @@ spec:
         # stdlib only; Flux keeps the tag in lockstep via the shared image
         # policy) — no extra dependency.
         - name: reconcile-publish-ip
-          image: ghcr.io/music-assistant/server:2.9.11 # {"$imagepolicy": "flux-system:music-assistant"}
+          image: ghcr.io/music-assistant/server:2.9.13 # {"$imagepolicy": "flux-system:music-assistant"}
           command: ["/app/venv/bin/python", "/scripts/reconcile.py"]
           volumeMounts:
             - name: data
@@ -62,7 +62,7 @@ spec:
               readOnly: true
       containers:
         - name: music-assistant
-          image: ghcr.io/music-assistant/server:2.9.11 # {"$imagepolicy": "flux-system:music-assistant"}
+          image: ghcr.io/music-assistant/server:2.9.13 # {"$imagepolicy": "flux-system:music-assistant"}
           env:
             - name: LOG_LEVEL
               value: info

kubernetes/apps/paperless/gotenberg.yaml

@@ -22,7 +22,7 @@ spec:
           type: RuntimeDefault
       containers:
         - name: gotenberg
-          image: gotenberg/gotenberg:8.35.0 # {"$imagepolicy": "flux-system:gotenberg"}
+          image: gotenberg/gotenberg:8.36.0 # {"$imagepolicy": "flux-system:gotenberg"}
           args:
             - "gotenberg"
             - "--chromium-disable-javascript=true"

kubernetes/apps/trmnl/cluster.yaml

@@ -0,0 +1,41 @@
+apiVersion: postgresql.cnpg.io/v1
+kind: Cluster
+metadata:
+  name: tr-pg
+  namespace: trmnl
+spec:
+  instances: 1
+  smartShutdownTimeout: 10
+  stopDelay: 60
+  # PG 18, NOT the repo-standard 16. Terminus ships a committed
+  # config/db/structure.sql containing `SET transaction_timeout = 0;` — a GUC
+  # that does not exist before PostgreSQL 17, so schema load against a 16
+  # server dies with `unrecognized configuration parameter`. Upstream's
+  # compose.yml develops against 18.4, so 18 is the tested target; 17 is the
+  # bare minimum. Bare major tag matches the other clusters (minor security
+  # bumps flow automatically, major upgrades stay an explicit decision).
+  imageName: ghcr.io/cloudnative-pg/postgresql:18
+  enableSuperuserAccess: true
+  storage:
+    storageClass: ceph-block
+    size: 5Gi
+  plugins:
+    - name: barman-cloud.cloudnative-pg.io
+      isWALArchiver: true
+      parameters:
+        barmanObjectName: tr-pg-store
+  bootstrap:
+    initdb:
+      database: terminus
+      owner: terminus
+  postgresql:
+    parameters:
+      shared_buffers: "128MB"
+      max_connections: "50"
+      wal_compression: "on"
+  resources:
+    requests:
+      cpu: 100m
+      memory: 256Mi
+    limits:
+      memory: 512Mi

kubernetes/apps/trmnl/device-service.yaml

@@ -0,0 +1,48 @@
+# Device-facing endpoint. Cleartext HTTP, deliberately.
+#
+# The TRMNL firmware initiates every request and cannot be handed the
+# internal CA, so terminating TLS here would simply break polling. Scope of
+# the exposure: this Service is LAN-only (no public DNS label, no Cloudflare
+# record), it carries no admin session — the device authenticates with its
+# MAC in the `ID` header plus a per-device api_key — and the Terminus admin
+# UI stays on the TLS route at trmnl.renner.uno.
+#
+# Same shape as home-assistant-bgp: a dedicated address on the Cilium BGP
+# pool rather than a TLS-less Ingress, which would be a new pattern for this
+# repo. The device is provisioned with the HOSTNAME, not the address.
+#
+# The address is requested explicitly, matching every other LB service here.
+# Left to allocate freely, Cilium IPAM handed out the first free address in
+# the pool block (10.2.99.0-10.2.99.255) — which is 10.2.99.0, the base
+# address of the range. Each LB address is advertised to the UDM as its own
+# BGP route so that is routable in principle, but a .0 host address is the
+# kind of thing constrained embedded IP stacks get wrong, and the TRMNL
+# firmware is the single client that has to work here. Not worth the risk
+# for an address nobody chose on purpose.
+#
+# .89 continues the existing sequence (.81-.88 are in use) and matches the
+# fd99::89 that IPAM had already picked for v6.
+apiVersion: v1
+kind: Service
+metadata:
+  name: terminus-device
+  namespace: trmnl
+  labels:
+    app.kubernetes.io/name: terminus
+    lb-pool: bgp
+  annotations:
+    external-dns.alpha.kubernetes.io/hostname: trmnl-device.renner.uno
+    lbipam.cilium.io/ips: "10.2.99.89,fd99::89"
+spec:
+  type: LoadBalancer
+  externalTrafficPolicy: Local
+  ipFamilyPolicy: PreferDualStack
+  ipFamilies:
+    - IPv4
+    - IPv6
+  selector:
+    app.kubernetes.io/name: terminus
+  ports:
+    - name: http
+      port: 80
+      targetPort: http

kubernetes/apps/trmnl/externalsecret.yaml

@@ -0,0 +1,18 @@
+apiVersion: external-secrets.io/v1
+kind: ExternalSecret
+metadata:
+  name: terminus-secrets
+  namespace: trmnl
+spec:
+  refreshInterval: 1h
+  secretStoreRef:
+    kind: ClusterSecretStore
+    name: openbao
+  target:
+    name: terminus-secrets
+  # Holds APP_SECRET only (session/cookie/CSRF key). DATABASE_URL is not in
+  # here — it comes straight off the CNPG-generated `tr-pg-app` Secret's
+  # `uri` key, so the password rotates with the cluster and is never copied.
+  dataFrom:
+    - extract:
+        key: apps/trmnl

kubernetes/apps/trmnl/hetzner-creds-externalsecret.yaml

@@ -0,0 +1,15 @@
+apiVersion: external-secrets.io/v1
+kind: ExternalSecret
+metadata:
+  name: hetzner-s3-creds
+  namespace: trmnl
+spec:
+  refreshInterval: 1h
+  secretStoreRef:
+    kind: ClusterSecretStore
+    name: openbao
+  target:
+    name: hetzner-s3-creds
+  dataFrom:
+    - extract:
+        key: seed/trmnl/hetzner-creds

kubernetes/apps/trmnl/ingress.yaml

@@ -0,0 +1,28 @@
+# Admin UI only — login, device registration, playlists, screen previews.
+# The device never touches this route: the firmware speaks plain HTTP and
+# would reject the internal CA, so it uses the cleartext LoadBalancer in
+# device-service.yaml instead. Both front the same Service.
+apiVersion: networking.k8s.io/v1
+kind: Ingress
+metadata:
+  name: terminus
+  namespace: trmnl
+  annotations:
+    cert-manager.io/cluster-issuer: letsencrypt-cluster
+spec:
+  ingressClassName: traefik
+  tls:
+    - hosts:
+        - trmnl.renner.uno
+      secretName: trmnl-tls
+  rules:
+    - host: trmnl.renner.uno
+      http:
+        paths:
+          - path: /
+            pathType: Prefix
+            backend:
+              service:
+                name: terminus
+                port:
+                  number: 80

kubernetes/apps/trmnl/kustomization.yaml

@@ -0,0 +1,20 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+resources:
+  - externalsecret.yaml
+  - pvcs.yaml
+  - valkey.yaml
+  - terminus.yaml
+  - ingress.yaml
+  - device-service.yaml
+  # Headless HA screenshotter. Lives in this namespace because Terminus
+  # FETCHES rendered screens back from it (uri delivery mode), so the
+  # two need the intra-namespace path.
+  - renderer.yaml
+  - renderer-externalsecret.yaml
+  # tr-pg CNPG Postgres + offsite Barman backup. Note this cluster runs
+  # PostgreSQL 18, not the repo-standard 16 — see the comment in cluster.yaml.
+  - cluster.yaml
+  - objectstore.yaml
+  - scheduled-backup.yaml
+  - hetzner-creds-externalsecret.yaml

kubernetes/apps/trmnl/objectstore.yaml

@@ -0,0 +1,30 @@
+---
+apiVersion: barmancloud.cnpg.io/v1
+kind: ObjectStore
+metadata:
+  name: tr-pg-store
+  namespace: trmnl
+spec:
+  configuration:
+    # Offsite Hetzner Object Storage (fsn1), same bucket as the other CNPG
+    # clusters. No SSE: Hetzner supports only SSE-C, unusable by barman-cloud
+    # (plugin #646), so at-rest encryption is unavailable; transit is TLS.
+    destinationPath: s3://renner-uno-k8s-cnpg/tr-pg
+    endpointURL: https://fsn1.your-objectstorage.com
+    s3Credentials:
+      accessKeyId:
+        name: hetzner-s3-creds
+        key: ACCESS_KEY_ID
+      secretAccessKey:
+        name: hetzner-s3-creds
+        key: ACCESS_SECRET_KEY
+      # Region is required (REGION=fsn1) — without it boto3's HeadBucket
+      # (barman-cloud-check-wal-archive during bootstrap.recovery) 400s.
+      region:
+        name: hetzner-s3-creds
+        key: REGION
+    wal:
+      compression: gzip
+    data:
+      compression: gzip
+  retentionPolicy: "30d"

kubernetes/apps/trmnl/pvcs.yaml

@@ -0,0 +1,57 @@
+---
+# Uploads + synchronised fonts for the Terminus pod. Mounted twice by
+# subPath (uploads/, fonts/) into both the web and sidekiq containers, which
+# share one pod — so RWO is sufficient and the Deployment uses `Recreate`.
+#
+# Contents: rendered device screens and the font set pulled by the
+# FONT_SYNCHRONIZER job. Persisting fonts here (rather than an emptyDir)
+# means a pod restart does not depend on upstream being reachable.
+apiVersion: v1
+kind: PersistentVolumeClaim
+metadata:
+  name: terminus-data
+  namespace: trmnl
+spec:
+  accessModes:
+    - ReadWriteOnce
+  storageClassName: ceph-block
+  resources:
+    requests:
+      storage: 5Gi
+---
+# Valkey RDB snapshots. The Sidekiq queue is reconstructable — the
+# synchroniser jobs re-enqueue on boot — so this is small and losing it is
+# an inconvenience, not data loss.
+apiVersion: v1
+kind: PersistentVolumeClaim
+metadata:
+  name: terminus-valkey-data
+  namespace: trmnl
+spec:
+  accessModes:
+    - ReadWriteOnce
+  storageClassName: ceph-block
+  resources:
+    requests:
+      storage: 1Gi
+---
+# Renderer state: schedules.json plus the rendered screens under
+# output/, which Terminus fetches back over HTTP in uri delivery mode.
+#
+# This mount is REQUIRED, not just nice to have. The app picks its data
+# directory by testing whether /data exists — with nothing mounted it
+# silently falls back to /app/data inside the container, so schedules
+# and rendered output would vanish on every restart and the panel would
+# go blank until the next render.
+apiVersion: v1
+kind: PersistentVolumeClaim
+metadata:
+  name: trmnl-renderer-data
+  namespace: trmnl
+spec:
+  accessModes:
+    - ReadWriteOnce
+  storageClassName: ceph-block
+  resources:
+    requests:
+      storage: 2Gi

kubernetes/apps/trmnl/renderer-externalsecret.yaml

@@ -0,0 +1,27 @@
+apiVersion: external-secrets.io/v1
+kind: ExternalSecret
+metadata:
+  name: trmnl-renderer-secrets
+  namespace: trmnl
+spec:
+  refreshInterval: 1h
+  secretStoreRef:
+    kind: ClusterSecretStore
+    name: openbao
+  target:
+    name: trmnl-renderer-secrets
+  # Kept separate from apps/trmnl (which holds Terminus's APP_SECRET)
+  # because these are credentials the RENDERER holds about other
+  # systems, with different rotation stories:
+  #
+  #   HA_ACCESS_TOKEN     — a dedicated Home Assistant long-lived token,
+  #                         not the operator's personal one, so the
+  #                         renderer can be revoked in isolation.
+  #   BYOS_ACCESS_TOKEN   — Terminus JWTs, minted once out of band from
+  #   BYOS_REFRESH_TOKEN    a Terminus login. The PASSWORD deliberately
+  #                         never enters the cluster; stock Terminus
+  #                         disables session expiry, so these are
+  #                         effectively permanent.
+  dataFrom:
+    - extract:
+        key: apps/trmnl/renderer

kubernetes/apps/trmnl/renderer.yaml

@@ -0,0 +1,385 @@
+# Headless renderer: screenshots the Home Assistant e-ink dashboard,
+# dithers it to 16 greys and hands it to Terminus.
+#
+# Upstream ships this as a Home Assistant add-on. Our HA is a plain
+# container with no Supervisor, so it runs standalone here instead —
+# viable because the app prefers HOME_ASSISTANT_URL/ACCESS_TOKEN env
+# vars over /data/options.json and authenticates with an ordinary
+# long-lived token. SUPERVISOR_TOKEN is deliberately NOT set: it only
+# flips an internal isAddOn flag, and setting it would switch the app
+# to add-on behaviour (fixed port, /data assumptions).
+#
+# DIRECTION OF TRAVEL, which is the non-obvious part: with the
+# byos-hanami webhook in `uri` delivery mode the renderer does NOT
+# push image bytes. It POSTs a URL pointing at its own /output/<file>
+# endpoint, and TERMINUS FETCHES THE IMAGE BACK from it. That is why
+# both live in this namespace — the pull is intra-namespace and is
+# already permitted by the default-deny-ingress rule.
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: trmnl-renderer-schedule
+  namespace: trmnl
+data:
+  # Template for /data/schedules.json. The init container splices the
+  # BYOS tokens in; everything non-secret is declared here so the
+  # schedule is reviewable in git rather than clicked into the app's
+  # web UI (which has no authentication and is not exposed).
+  #
+  # The app re-reads schedules.json every 60 s, so edits here take
+  # effect without a restart once the init container has rewritten it.
+  #
+  # lang is "de", not null. It calls Home Assistant's own
+  # _selectLanguage() for the screenshot session, so the rendered page
+  # is localised by HA rather than by hand. With it left null the
+  # dashboard grew a German weekday array, a German month array and a
+  # fifteen-entry weather-condition dict — reimplementing translations
+  # HA already ships. The instance itself stays language "en"; this
+  # only affects the render session.
+  #
+  # model_id is 1 — TRMNL X (name "v2", 1872x1404, 16 colours) — and it
+  # is an INTEGER, not a name. Confirmed against the running server:
+  # GET /api/screens returns model_id as a number, and upstream's
+  # example value "og_1bit" matches no model on this instance (the
+  # 2-bit OG model here is id 11, name "og_plus"). Getting this wrong
+  # shows up as Terminus rejecting the pushed screen with a 422.
+  #
+  # All three auth fields below are placeholders the init container
+  # replaces — including obtained_at, which must end up a real epoch-MS
+  # stamp. The 0 here is inert only because it is always overwritten;
+  # left as 0 the scheduler reads the tokens as ancient and refuses to
+  # push at all ("expired beyond the refresh window"), with no render
+  # attempt and no error beyond one startup warning.
+  schedules.template.json: |
+    [
+      {
+        "id": "trmnl-entrance",
+        "name": "Entrance e-ink panel",
+        "enabled": true,
+        "cron": "* * * * *",
+        "interval_minutes": 1,
+        "ha_mode": true,
+        "dashboard_path": "/trmnl-entrance/home",
+        "target_url": null,
+        "viewport": { "width": 1872, "height": 1404 },
+        "device": "schedule_trmnl_x",
+        "crop": { "enabled": false, "x": 0, "y": 0, "width": 0, "height": 0 },
+        "format": "png",
+        "rotate": null,
+        "zoom": 1,
+        "wait": null,
+        "theme": "Graphite E-ink Light",
+        "lang": "de",
+        "dark": false,
+        "invert": false,
+        "timestamp": false,
+        "dithering": {
+          "enabled": true,
+          "method": "floyd-steinberg",
+          "palette": "gray-16",
+          "bitDepth": 4,
+          "gammaCorrection": true,
+          "blackLevel": 0,
+          "whiteLevel": 100,
+          "normalize": true,
+          "saturationBoost": false
+        },
+        "webhook_url": "http://terminus.trmnl.svc.cluster.local/api/screens",
+        "webhook_headers": {},
+        "webhook_format": {
+          "format": "byos-hanami",
+          "byosConfig": {
+            "label": "Eingang",
+            "name": "trmnl-entrance",
+            "model_id": 1,
+            "preprocessed": true,
+            "addon_base_url": "http://trmnl-renderer.trmnl.svc.cluster.local:10000",
+            "delivery_mode": "uri",
+            "auth": {
+              "enabled": true,
+              "access_token": "__BYOS_ACCESS_TOKEN__",
+              "refresh_token": "__BYOS_REFRESH_TOKEN__",
+              "obtained_at": 0
+            }
+          }
+        },
+        "createdAt": "2026-08-11T00:00:00.000Z",
+        "updatedAt": "2026-08-11T00:00:00.000Z"
+      }
+    ]
+---
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: trmnl-renderer
+  namespace: trmnl
+  labels:
+    app.kubernetes.io/name: trmnl-renderer
+  annotations:
+    # BOTH sources need this, and for the same reason: the init
+    # container writes /data/schedules.json ONCE, at pod start. Nothing
+    # re-reads either source afterwards.
+    #
+    # The app re-reads schedules.json every 60s, which makes it look
+    # self-updating — but only the FILE, and the file is only rewritten
+    # by the init container. So a ConfigMap or Secret change lands in
+    # git, updates the object, and the running renderer keeps the old
+    # values indefinitely.
+    #
+    # Both halves were hit repeatedly in practice and each time needed
+    # `kubectl delete pod` by hand: the theme name, the schedule's lang
+    # field, and every re-mint of the BYOS tokens. A fix that depends on
+    # someone remembering to restart a pod is not deployed.
+    #
+    # Named per-resource rather than the blanket
+    # reloader.stakater.com/auto: "true", matching terminus.yaml and
+    # mosquitto-deployment.yaml.
+    configmap.reloader.stakater.com/reload: trmnl-renderer-schedule
+    secret.reloader.stakater.com/reload: trmnl-renderer-secrets
+spec:
+  replicas: 1
+  # RWO PVC, and two renderers would push conflicting screens for the
+  # same device.
+  strategy:
+    type: Recreate
+  selector:
+    matchLabels:
+      app.kubernetes.io/name: trmnl-renderer
+  template:
+    metadata:
+      labels:
+        app.kubernetes.io/name: trmnl-renderer
+    spec:
+      enableServiceLinks: false
+      securityContext:
+        runAsNonRoot: true
+        runAsUser: 1000
+        runAsGroup: 1000
+        fsGroup: 1000
+        fsGroupChangePolicy: OnRootMismatch
+        seccompProfile:
+          type: RuntimeDefault
+      initContainers:
+        # Writes /data/schedules.json from the ConfigMap template with
+        # the BYOS tokens injected.
+        #
+        # jq rather than sed: the tokens are JWTs and go into a JSON
+        # string, so they need JSON-safe escaping. A sed substitution
+        # would corrupt on any character sed treats as special.
+        #
+        # No HTTP here on purpose. The app does expose a /api/byos/login
+        # proxy that exchanges a Terminus username+password for these
+        # tokens, but that would mean shipping the Terminus PASSWORD
+        # into the cluster and doing a network call at every pod start.
+        # Tokens are therefore minted once out of band and kept in
+        # OpenBao — the password never enters the cluster.
+        #
+        # That only works because Terminus is configured with
+        # API_ACCESS_TOKEN_PERIOD = 10 years (see terminus.yaml). It is
+        # NOT, as an earlier version of this comment claimed, because
+        # stock Terminus issues permanent tokens — the stock default is
+        # 1800s, and this deployment demonstrated it by having the
+        # webhook start failing with "expired JWT access token" about
+        # thirty minutes after seeding.
+        - name: render-schedule
+          image: gitea.renner.uno/gitea_admin/alpine-jq:2026-07-30
+          command:
+            - sh
+            - -c
+            - |
+              set -eu
+              # obtained_at is stamped at write time, in MILLISECONDS.
+              # It is not cosmetic: the scheduler derives token age from
+              # it, and a literal 0 made the app treat freshly-minted
+              # tokens as ancient and skip the webhook entirely —
+              #   "BYOS tokens ... expired beyond the refresh window"
+              #
+              # Stamping it "now" is only safe because Terminus is
+              # configured with API_ACCESS_TOKEN_PERIOD = 10 years (see
+              # terminus.yaml). Against the 30-minute default it would be
+              # actively harmful: the app would believe a stale token had
+              # 30 fresh minutes left, skip its proactive refresh, and
+              # then be unable to recover, because refreshing requires a
+              # still-valid token. If that setting is ever reverted, this
+              # must become a real mint time carried alongside the tokens.
+              jq --arg at "$BYOS_ACCESS_TOKEN" \
+                 --arg rt "$BYOS_REFRESH_TOKEN" \
+                 '(.[].webhook_format.byosConfig.auth.access_token) = $at
+                  | (.[].webhook_format.byosConfig.auth.refresh_token) = $rt
+                  | (.[].webhook_format.byosConfig.auth.obtained_at) = (now * 1000 | floor)' \
+                 /template/schedules.template.json > /data/schedules.json
+              echo "wrote /data/schedules.json ($(wc -c < /data/schedules.json) bytes)"
+          env:
+            - name: BYOS_ACCESS_TOKEN
+              valueFrom:
+                secretKeyRef:
+                  name: trmnl-renderer-secrets
+                  key: BYOS_ACCESS_TOKEN
+            - name: BYOS_REFRESH_TOKEN
+              valueFrom:
+                secretKeyRef:
+                  name: trmnl-renderer-secrets
+                  key: BYOS_REFRESH_TOKEN
+          volumeMounts:
+            - name: data
+              mountPath: /data
+            - name: schedule-template
+              mountPath: /template
+              readOnly: true
+          securityContext:
+            allowPrivilegeEscalation: false
+            readOnlyRootFilesystem: true
+            capabilities:
+              drop: ["ALL"]
+      containers:
+        # Pinned, not enrolled in image automation — same reasoning as
+        # Terminus itself. This drives a wall-mounted panel and upstream
+        # is pre-1.0; an unattended bump that changes the dithering
+        # pipeline or the webhook payload shows up as a corrupted
+        # screen, not as a failed rollout.
+        #
+        # Note the image is ARCH-SUFFIXED: upstream publishes no
+        # multi-arch manifest, only trmnl-ha-amd64 and trmnl-ha-aarch64.
+        - name: renderer
+          image: ghcr.io/usetrmnl/trmnl-ha-amd64:0.9.5
+          ports:
+            - name: http
+              containerPort: 10000
+          env:
+            # The in-cluster Service name, NOT ha.renner.uno and not
+            # home.renner.uno. Both public names are unusable from a
+            # pod: ha.renner.uno resolves to the home-assistant-bgp
+            # LoadBalancer's external address (10.2.99.87), which is not
+            # backed by an Ingress, so CoreDNS's k8s_gateway delegation
+            # does not intercept it and the pod hairpins off the LB.
+            #
+            # This cluster has hit that exact failure before — see the
+            # comment on kubernetes/infrastructure/openbao/rotation/
+            # home-assistant.yaml, which reaches HA the same way and
+            # states it outright: "The public name is unusable from a
+            # pod (ingress-LB hairpin)."
+            #
+            # FAILS QUIETLY if wrong: the schedule keeps firing every
+            # minute, each render errors on connect, and the panel just
+            # never updates. Nothing goes CrashLoopBackOff.
+            - name: HOME_ASSISTANT_URL
+              value: "http://home-assistant-bgp.home-assistant.svc.cluster.local:8123"
+            - name: TZ
+              value: "Europe/Berlin"
+            # Chromium will not start without a writable HOME. Running
+            # as uid 1000 with no matching passwd entry leaves HOME="/",
+            # which is root-owned 0755, and the browser dies on launch
+            # with
+            #   chrome_crashpad_handler: --database is required
+            # because crashpad cannot create its database directory.
+            #
+            # Diagnosed on the running pod, not guessed: HOME=/, uid=1000,
+            # / not writable, /tmp writable. Crash dumps are throwaway, so
+            # /tmp is the right home for them rather than the PVC.
+            #
+            # This is a consequence of OUR non-root securityContext, not
+            # of the image — so it must move together with runAsUser.
+            - name: HOME
+              value: "/tmp"
+            # Dedicated long-lived token, NOT the operator's personal
+            # one, so the renderer's access can be revoked on its own.
+            # It is injected into the headless browser's localStorage,
+            # which is how the app authenticates to the HA frontend.
+            - name: ACCESS_TOKEN
+              valueFrom:
+                secretKeyRef:
+                  name: trmnl-renderer-secrets
+                  key: HA_ACCESS_TOKEN
+          volumeMounts:
+            - name: data
+              mountPath: /data
+            # Bounded scratch backing HOME. Chromium writes crashpad
+            # state and profile data here on every launch, and it now
+            # launches once a minute for the life of the pod — on the
+            # container's writable layer that is unbounded growth
+            # against node disk, with no ephemeral-storage limit to
+            # catch it. A sizeLimit converts a slow, cluster-wide
+            # node-disk-pressure problem into a bounded and clearly
+            # attributable eviction of this pod.
+            - name: tmp
+              mountPath: /tmp
+          # HTTP probes here, unlike terminus.yaml's TCP-only ones, because
+          # /health is confirmed rather than assumed: the upstream image's
+          # own Dockerfile declares
+          #   HEALTHCHECK ... curl -f http://localhost:10000/health
+          # so the route is part of the image's published contract. Terminus
+          # got TCP probes precisely because it publishes no such endpoint.
+          livenessProbe:
+            httpGet:
+              path: /health
+              port: http
+            initialDelaySeconds: 30
+            periodSeconds: 30
+            timeoutSeconds: 10
+          readinessProbe:
+            httpGet:
+              path: /health
+              port: http
+            initialDelaySeconds: 10
+            periodSeconds: 10
+            timeoutSeconds: 5
+          resources:
+            requests:
+              cpu: 250m
+              memory: 768Mi
+            # Chromium is a SEPARATE process from the app, and the app's
+            # own watchdog only measures its own RSS (restarting itself
+            # past 900 MB). The limit therefore has to cover the app's
+            # ceiling PLUS a browser rendering a 1872x1404 page, which
+            # is well above the 512 MB upstream's troubleshooting note
+            # suggests.
+            limits:
+              memory: 2Gi
+          # Chromium normally needs either SYS_ADMIN or a user namespace to
+          # start its own sandbox, which would make dropping ALL fatal. It
+          # is safe here because the app launches the browser with
+          # --no-sandbox --disable-setuid-sandbox --no-zygote (and
+          # --disable-dev-shm-usage, which is why no /dev/shm emptyDir is
+          # mounted either). Those flags are baked into the app's own
+          # PUPPETEER_ARGS, not something we pass, so this holds as long as
+          # the pinned tag does — worth re-checking on any bump.
+          #
+          # Confirmed on the running pod: the capability drop is genuinely
+          # not what broke the browser. The first deploy DID crash it, but
+          # on a writable-HOME problem caused by runAsUser — see the HOME
+          # env above. Do not "fix" a future Chromium launch failure by
+          # loosening this block before checking HOME and /tmp.
+          securityContext:
+            allowPrivilegeEscalation: false
+            capabilities:
+              drop: ["ALL"]
+      volumes:
+        - name: data
+          persistentVolumeClaim:
+            claimName: trmnl-renderer-data
+        - name: schedule-template
+          configMap:
+            name: trmnl-renderer-schedule
+        - name: tmp
+          emptyDir:
+            sizeLimit: 512Mi
+---
+apiVersion: v1
+kind: Service
+metadata:
+  name: trmnl-renderer
+  namespace: trmnl
+  labels:
+    app.kubernetes.io/name: trmnl-renderer
+spec:
+  # ClusterIP only, and deliberately no Ingress. The app's web UI on
+  # this port has NO authentication of any kind — upstream says so
+  # outright — so it must not be routable from the LAN. Terminus
+  # reaches it in-namespace to fetch rendered screens; nothing else
+  # needs to.
+  selector:
+    app.kubernetes.io/name: trmnl-renderer
+  ports:
+    - name: http
+      port: 10000
+      targetPort: http

kubernetes/apps/trmnl/scheduled-backup.yaml

@@ -0,0 +1,16 @@
+apiVersion: postgresql.cnpg.io/v1
+kind: ScheduledBackup
+metadata:
+  name: tr-pg-daily
+  namespace: trmnl
+spec:
+  # 03:00 local, same slot as the other CNPG clusters (the Velero PVC
+  # schedule for this namespace runs later, at 08:15).
+  schedule: "0 0 3 * * *"
+  immediate: true
+  backupOwnerReference: self
+  cluster:
+    name: tr-pg
+  method: plugin
+  pluginConfiguration:
+    name: barman-cloud.cloudnative-pg.io

kubernetes/apps/trmnl/terminus.yaml

@@ -0,0 +1,256 @@
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: terminus-config
+  namespace: trmnl
+data:
+  TZ: "Europe/Berlin"
+  HANAMI_PORT: "2345"
+  # What the device is told to fetch from. Terminus bakes this into the
+  # absolute `image_url` it hands back on /api/display, so it MUST be the
+  # cleartext LAN name, not trmnl.renner.uno — the firmware speaks plain
+  # HTTP and would reject the internal CA. No trailing slash: the same
+  # string is typed into the captive portal at provisioning time.
+  API_URI: "http://trmnl-device.renner.uno"
+  KEYVALUE_URL: "redis://terminus-valkey:6379"
+  # Rack::Attack rate-limit bypass. Beyond upstream's RFC1918 defaults this
+  # adds fc00::/7, which covers the pod CIDR (fd10:244::/48), the Cilium LB
+  # pool (fd99::/120) and the LAN ULA — without it every IPv6 client,
+  # including the device and Traefik, is rate-limited as if it were remote.
+  RACK_ATTACK_ALLOWED_SUBNETS: "10.0.0.0/8,172.16.0.0/12,192.168.0.0/16,127.0.0.1,::1,fc00::/7"
+  # 10 years, against a default of 1800s (30 minutes).
+  #
+  # Documented upstream in doc/configuration.adoc:
+  #   API_ACCESS_TOKEN_PERIOD — "The amount of time, in seconds, the JWT
+  #   access token remains valid before needing to refresh your access
+  #   token." Default: 1,800 seconds (30 minutes).
+  # Corroborated on this instance: a token minted here was rejected about
+  # thirty minutes later, i.e. the documented default behaving as written.
+  #
+  # This exists for the renderer, which pushes screens to /api/screens with
+  # a JWT held in OpenBao. At the default period that arrangement cannot
+  # work unattended, and it fails in a way that has no recovery: minting a
+  # replacement token via POST /api/jwt requires a still-VALID access token
+  # in the Authorization header, so once one expires it cannot be refreshed
+  # from itself. Observed exactly that on first deploy —
+  #   [webhook] Webhook failed: 400 {"error":"expired JWT access token"}
+  #   [webhook] BYOS auth: refresh failed (400) {"error":"expired JWT ..."}
+  # — the refresh attempt being rejected by the very expiry it was trying
+  # to repair.
+  #
+  # The renderer's own code assumes "stock Terminus disables session
+  # expiration and issues effectively permanent tokens". That is simply not
+  # true of this deployment, so the assumption is made true here rather
+  # than worked around in the client.
+  #
+  # The trade is deliberate: a long-lived JWT with full Terminus API rights
+  # now sits in OpenBao. Accepted because the alternative was putting the
+  # Terminus account PASSWORD into the cluster so an init container could
+  # mint a short-lived token on every pod start, which is a strictly worse
+  # secret to hold. The token is revocable by rotating the account's
+  # credentials; it grants access only to this LAN-only instance.
+  #
+  # NOTE: the period is fixed at mint time. Changing this does not extend
+  # tokens that already exist — they must be re-minted afterwards.
+  API_ACCESS_TOKEN_PERIOD: "315360000"
+  # The OTHER half of the same problem, and the reason the first fix
+  # did not hold. Terminus has TWO independent expiries: the JWT's own
+  # lifetime (above) and the SESSION behind it. Raising only the token
+  # period left the session at its defaults, so pushes kept working for
+  # a while and then failed with
+  #   HTTP 401 {"error":"This session has expired, please login again"}
+  # which is a different error, from a different mechanism, than the
+  # "expired JWT access token" the first fix addressed.
+  #
+  # Documented in doc/configuration.adoc:
+  #   SESSION_EXPIRATION_ENABLED — "Determines if your browser session
+  #   or JWT API token is expired or not." Default: true.
+  #   SESSION_INACTIVITY_LIMIT 1800s, SESSION_LIFETIME_LIMIT 86400s.
+  #
+  # Disabling it is exactly the state the renderer's own code assumes
+  # ("stock Terminus disables session expiration and issues ~100-year
+  # tokens") — an assumption that is false by default and has to be
+  # made true here, same as the token period.
+  #
+  # Same trade as above, and it is the deciding one for an unattended
+  # appliance: a session that expires needs a human with a password to
+  # restore it, and this panel has neither.
+  SESSION_EXPIRATION_ENABLED: "false"
+---
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: terminus
+  namespace: trmnl
+  labels:
+    app.kubernetes.io/name: terminus
+  annotations:
+    # terminus-config is consumed via envFrom, and env vars are read ONCE at
+    # container start — Kubernetes does not hot-reload them. The ConfigMap is
+    # a plain resource, not a configMapGenerator, so editing it does not
+    # change the pod template hash and Flux will not roll the pods either.
+    # Without this annotation a ConfigMap-only change (such as raising
+    # API_ACCESS_TOKEN_PERIOD) applies to git and to the ConfigMap object
+    # while the running app keeps the old value indefinitely — a silent
+    # no-op, with the manifest and reality disagreeing.
+    #
+    # Named per-ConfigMap rather than the blanket `reloader.stakater.com/
+    # auto: "true"`, matching mosquitto-deployment.yaml: enrolling anything
+    # further stays a visible, reviewable edit to this line.
+    configmap.reloader.stakater.com/reload: terminus-config
+spec:
+  replicas: 1
+  # RWO PVC shared by both containers in this pod, and APP_SETUP runs
+  # migrations on boot — two concurrent web pods would race on both.
+  strategy:
+    type: Recreate
+  selector:
+    matchLabels:
+      app.kubernetes.io/name: terminus
+  template:
+    metadata:
+      labels:
+        app.kubernetes.io/name: terminus
+    spec:
+      enableServiceLinks: false
+      securityContext:
+        runAsNonRoot: true
+        runAsUser: 1000
+        runAsGroup: 1000
+        fsGroup: 1000
+        fsGroupChangePolicy: OnRootMismatch
+        seccompProfile:
+          type: RuntimeDefault
+      containers:
+        # Tag pinned deliberately and NOT enrolled in Flux image automation.
+        # Terminus is pre-1.0 with no GitHub Releases and no changelog; an
+        # unattended bump that changes the firmware or /api/display contract
+        # is a device-bricking class of risk. Bumps are a reviewed, manual
+        # act — see kubernetes/infrastructure/image-automation/README.md.
+        - name: terminus
+          image: ghcr.io/usetrmnl/terminus:0.68.0
+          ports:
+            - name: http
+              containerPort: 2345
+          envFrom:
+            - configMapRef:
+                name: terminus-config
+          env:
+            # Runs `hanami assets compile` + `hanami db migrate` in the
+            # entrypoint before exec'ing puma. Set on the web container
+            # only — the sidekiq sidecar must not race the migration.
+            - name: APP_SETUP
+              value: "true"
+            # CNPG generates this Secret; `uri` already carries the rotated
+            # password and points at the tr-pg-rw service.
+            - name: DATABASE_URL
+              valueFrom:
+                secretKeyRef:
+                  name: tr-pg-app
+                  key: uri
+            - name: APP_SECRET
+              valueFrom:
+                secretKeyRef:
+                  name: terminus-secrets
+                  key: APP_SECRET
+          volumeMounts:
+            - name: data
+              mountPath: /app/public/uploads
+              subPath: uploads
+            - name: data
+              mountPath: /app/public/fonts
+              subPath: fonts
+          # TCP probes, not HTTP: Terminus exposes no documented health
+          # endpoint, and every real route is behind either auth or
+          # Rack::Attack. Tighten to an HTTP GET once the route set is
+          # confirmed against the running app.
+          startupProbe:
+            tcpSocket:
+              port: http
+            periodSeconds: 10
+            # ~10 min: first boot compiles assets and migrates a fresh DB.
+            failureThreshold: 60
+          livenessProbe:
+            tcpSocket:
+              port: http
+            periodSeconds: 30
+            timeoutSeconds: 5
+          readinessProbe:
+            tcpSocket:
+              port: http
+            periodSeconds: 10
+            timeoutSeconds: 5
+          resources:
+            requests:
+              cpu: 200m
+              memory: 512Mi
+            limits:
+              memory: 1Gi
+          securityContext:
+            allowPrivilegeEscalation: false
+            capabilities:
+              drop: ["ALL"]
+        - name: sidekiq
+          image: ghcr.io/usetrmnl/terminus:0.68.0
+          command: ["bundle", "exec", "sidekiq", "-r", "./config/sidekiq.rb"]
+          envFrom:
+            - configMapRef:
+                name: terminus-config
+          # EXPECTED on a fresh deploy: this container loses the race against
+          # the web container's migration and exits 1 with
+          # `PG::UndefinedTable: relation "account" does not exist`, then
+          # comes up healthy on the restart once the schema exists.
+          # Observed on first deploy 2026-08-10; converged in ~1s. There is
+          # no upstream ordering hook, and gating it behind an initContainer
+          # that probes for a specific table would couple us to Terminus's
+          # schema. So this is left to Kubernetes' restart, deliberately —
+          # a single sidekiq restart shortly after a rollout is benign. More
+          # than a couple, or restarts long after a rollout, are NOT this.
+          env:
+            # No APP_SETUP here — see the web container.
+            - name: DATABASE_URL
+              valueFrom:
+                secretKeyRef:
+                  name: tr-pg-app
+                  key: uri
+            - name: APP_SECRET
+              valueFrom:
+                secretKeyRef:
+                  name: terminus-secrets
+                  key: APP_SECRET
+          volumeMounts:
+            - name: data
+              mountPath: /app/public/uploads
+              subPath: uploads
+            - name: data
+              mountPath: /app/public/fonts
+              subPath: fonts
+          resources:
+            requests:
+              cpu: 100m
+              memory: 256Mi
+            limits:
+              memory: 512Mi
+          securityContext:
+            allowPrivilegeEscalation: false
+            capabilities:
+              drop: ["ALL"]
+      volumes:
+        - name: data
+          persistentVolumeClaim:
+            claimName: terminus-data
+---
+apiVersion: v1
+kind: Service
+metadata:
+  name: terminus
+  namespace: trmnl
+  labels:
+    app.kubernetes.io/name: terminus
+spec:
+  selector:
+    app.kubernetes.io/name: terminus
+  ports:
+    - name: http
+      port: 80
+      targetPort: http

kubernetes/apps/trmnl/valkey.yaml

@@ -0,0 +1,79 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: terminus-valkey
+  namespace: trmnl
+  labels:
+    app.kubernetes.io/name: terminus-valkey
+spec:
+  replicas: 1
+  strategy:
+    type: Recreate
+  selector:
+    matchLabels:
+      app.kubernetes.io/name: terminus-valkey
+  template:
+    metadata:
+      labels:
+        app.kubernetes.io/name: terminus-valkey
+    spec:
+      securityContext:
+        runAsNonRoot: true
+        runAsUser: 999
+        fsGroup: 999
+        seccompProfile:
+          type: RuntimeDefault
+      containers:
+        # Valkey major version tracks upstream's compose.yml (9.x). Sidekiq
+        # is the only consumer; a major bump is a separate migration.
+        - name: valkey
+          image: valkey/valkey:9.1.1-alpine # {"$imagepolicy": "flux-system:valkey-trmnl"}
+          args:
+            - valkey-server
+            - --save
+            - "60"
+            - "1"
+            - --loglevel
+            - warning
+          ports:
+            - name: valkey
+              containerPort: 6379
+          volumeMounts:
+            - name: data
+              mountPath: /data
+          livenessProbe:
+            tcpSocket:
+              port: valkey
+            initialDelaySeconds: 10
+          readinessProbe:
+            exec:
+              command: ["valkey-cli", "ping"]
+            initialDelaySeconds: 5
+            periodSeconds: 10
+          resources:
+            requests:
+              cpu: 20m
+              memory: 64Mi
+            limits:
+              memory: 256Mi
+          securityContext:
+            allowPrivilegeEscalation: false
+            readOnlyRootFilesystem: true
+            capabilities:
+              drop: ["ALL"]
+      volumes:
+        - name: data
+          persistentVolumeClaim:
+            claimName: terminus-valkey-data
+---
+apiVersion: v1
+kind: Service
+metadata:
+  name: terminus-valkey
+  namespace: trmnl
+spec:
+  selector:
+    app.kubernetes.io/name: terminus-valkey
+  ports:
+    - port: 6379
+      targetPort: valkey

kubernetes/apps/whisper/deployment.yaml

@@ -18,7 +18,7 @@ spec:
     spec:
       containers:
         - name: whisper
-          image: rhasspy/wyoming-whisper:3.5.0 # {"$imagepolicy": "flux-system:wyoming-whisper"}
+          image: rhasspy/wyoming-whisper:3.6.0 # {"$imagepolicy": "flux-system:wyoming-whisper"}
           args:
             - --model=base-int8
             - --language=de

kubernetes/apps/wordpress/deployment.yaml

@@ -0,0 +1,115 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: wordpress
+  namespace: wordpress
+  annotations:
+    secret.reloader.stakater.com/reload: wordpress-secrets
+spec:
+  replicas: 1
+  strategy:
+    # RWO volume, single writer — the rolling default would deadlock on the
+    # new pod waiting for a volume the old pod still holds.
+    type: Recreate
+  selector:
+    matchLabels:
+      app.kubernetes.io/name: wordpress
+  template:
+    metadata:
+      labels:
+        app.kubernetes.io/name: wordpress
+    spec:
+      securityContext:
+        # wp-content must be writable by Apache (www-data, uid/gid 33) for
+        # media uploads and plugin installs.
+        fsGroup: 33
+      containers:
+        - name: wordpress
+          image: wordpress:7.0.4-php8.3-apache
+          ports:
+            - name: http
+              containerPort: 80
+          env:
+            # The operator names the base Service after the MariaDB resource;
+            # the CRD's separate primaryService/secondaryService fields are the
+            # suffixed variants for replicated setups, which this is not
+            # (replicas: 1). Not verified against a running cluster — if it is
+            # wrong the pod fails loudly on connect at startup rather than
+            # degrading quietly. Confirm with `kubectl get svc -n wordpress`
+            # on first deploy.
+            - name: WORDPRESS_DB_HOST
+              value: wordpress-mariadb.wordpress.svc.cluster.local:3306
+            - name: WORDPRESS_DB_NAME
+              value: wordpress
+            - name: WORDPRESS_DB_USER
+              value: wordpress
+            - name: WORDPRESS_DB_PASSWORD
+              valueFrom:
+                secretKeyRef:
+                  # Generated by the operator from the MariaDB CR — never
+                  # stored in Git or OpenBao.
+                  name: wordpress-mariadb-app
+                  key: password
+            - name: WORDPRESS_CONFIG_EXTRA
+              value: |
+                define('WP_HOME', 'https://www.renner.uno');
+                define('WP_SITEURL', 'https://www.renner.uno');
+                define('FORCE_SSL_ADMIN', true);
+                /* Traefik terminates TLS; without this WordPress sees plain
+                   HTTP and builds http:// URLs into every page and redirect. */
+                if (isset($_SERVER['HTTP_X_FORWARDED_PROTO']) && $_SERVER['HTTP_X_FORWARDED_PROTO'] === 'https') {
+                    $_SERVER['HTTPS'] = 'on';
+                }
+                /* Core lives in the image, not on the volume — the updater
+                   would only fail against a read-only docroot. Plugin and
+                   theme installs are unaffected. */
+                define('WP_AUTO_UPDATE_CORE', false);
+                /* No add_filter() here. WORDPRESS_CONFIG_EXTRA is eval'd inside
+                   wp-config.php, long before wp-settings.php loads the plugin
+                   API, so any add_filter call is a fatal error on every request
+                   — which is exactly what happened on first deploy. xmlrpc is
+                   handled at the ingress instead, by routing /xmlrpc.php
+                   through the Authentik gate. */
+          envFrom:
+            # The eight WordPress salts. Rotating them in OpenBao invalidates
+            # every session — which is the intended lever, not a side effect.
+            - secretRef:
+                name: wordpress-secrets
+          volumeMounts:
+            - name: content
+              mountPath: /var/www/html/wp-content
+          readinessProbe:
+            httpGet:
+              path: /
+              port: http
+            initialDelaySeconds: 10
+            periodSeconds: 10
+          livenessProbe:
+            httpGet:
+              path: /
+              port: http
+            initialDelaySeconds: 60
+            periodSeconds: 30
+          resources:
+            requests:
+              cpu: 50m
+              memory: 256Mi
+            limits:
+              memory: 768Mi
+      volumes:
+        - name: content
+          persistentVolumeClaim:
+            claimName: wordpress-content
+---
+apiVersion: v1
+kind: Service
+metadata:
+  name: wordpress
+  namespace: wordpress
+spec:
+  selector:
+    app.kubernetes.io/name: wordpress
+  ports:
+    - name: http
+      port: 80
+      targetPort: http

kubernetes/apps/wordpress/externalsecret.yaml

@@ -0,0 +1,46 @@
+# The eight WordPress salts (AUTH_KEY, SECURE_AUTH_KEY, LOGGED_IN_KEY,
+# NONCE_KEY, AUTH_SALT, SECURE_AUTH_SALT, LOGGED_IN_SALT, NONCE_SALT).
+# Expected keys in OpenBao at homelab/apps/wordpress, named exactly as the
+# image's env vars: WORDPRESS_AUTH_KEY, WORDPRESS_SECURE_AUTH_KEY, ...
+#
+# Deliberately NOT carried over from the old wp-config.php: the Hetzner host
+# is being decommissioned and its salts have been sitting on a VM whose
+# neighbour site was compromised this month. New salts cost one logout.
+apiVersion: external-secrets.io/v1
+kind: ExternalSecret
+metadata:
+  name: wordpress-secrets
+  namespace: wordpress
+spec:
+  refreshInterval: 1h
+  secretStoreRef:
+    kind: ClusterSecretStore
+    name: openbao
+  target:
+    name: wordpress-secrets
+  dataFrom:
+    - extract:
+        key: apps/wordpress
+---
+# S3 credentials for the MariaDB Backup.
+apiVersion: external-secrets.io/v1
+kind: ExternalSecret
+metadata:
+  name: hetzner-s3-creds
+  namespace: wordpress
+spec:
+  refreshInterval: 1h
+  secretStoreRef:
+    kind: ClusterSecretStore
+    name: openbao
+  target:
+    name: hetzner-s3-creds
+  dataFrom:
+    # Reuses the paperless copy of the shared Hetzner S3 credentials (identical
+    # ACCESS_KEY_ID / ACCESS_SECRET_KEY / REGION) to avoid seeding a new OpenBao
+    # path for this migration — same approach woodpecker took. FOLLOW-UP: give
+    # wordpress its own seed/wordpress/hetzner-creds path once OpenBao can be
+    # written, and repoint this here (the value is the same, so it's a drop-in
+    # swap).
+    - extract:
+        key: seed/paperless/hetzner-creds

kubernetes/apps/wordpress/gate-service.yaml

@@ -0,0 +1,15 @@
+# A plain Ingress cannot point at a Service in another namespace, and the
+# Authentik outposts all live in `authentik`. Same workaround as
+# zone-configurator and paperless: mirror the outpost Service in as an
+# ExternalName.
+apiVersion: v1
+kind: Service
+metadata:
+  name: ak-outpost-wordpress-gate
+  namespace: wordpress
+spec:
+  type: ExternalName
+  externalName: ak-outpost-wordpress-gate.authentik.svc.cluster.local
+  ports:
+    - port: 9000
+      targetPort: 9000

kubernetes/apps/wordpress/ingress.yaml

@@ -0,0 +1,110 @@
+# Two routers on one host. The site itself must stay readable by anyone --
+# it is a family blog -- while the editing surface sits behind Authentik.
+#
+# Why not one gated Ingress with skip_path_regex, the way paperless-gate
+# carves out /static/: that works when the PUBLIC paths are enumerable. Here
+# it is the reverse -- every post URL is public and only two paths are gated
+# -- which would need a negative lookahead. The outpost is Go, and RE2 has no
+# lookahead. So the split happens at the router instead.
+#
+# Traefik orders routers by rule length, so PathPrefix(`/wp-admin`) is
+# evaluated before PathPrefix(`/`); no explicit priority is needed.
+apiVersion: networking.k8s.io/v1
+kind: Ingress
+metadata:
+  name: wordpress
+  namespace: wordpress
+  labels:
+    dns.renner.uno/public: "true"
+  annotations:
+    cert-manager.io/cluster-issuer: letsencrypt-cluster
+spec:
+  ingressClassName: traefik
+  rules:
+    - host: www.renner.uno
+      http:
+        paths:
+          - path: /
+            pathType: Prefix
+            backend:
+              service:
+                name: wordpress
+                port:
+                  number: 80
+  tls:
+    - secretName: wordpress-tls
+      hosts:
+        - www.renner.uno
+---
+apiVersion: networking.k8s.io/v1
+kind: Ingress
+metadata:
+  name: wordpress-admin
+  namespace: wordpress
+  labels:
+    # Same label as the public Ingress on purpose. external-dns-unifi runs
+    # with --label-filter=dns.renner.uno/public!=true, so an unlabelled
+    # Ingress on this host would quietly gain a UDM A-record pointing LAN
+    # clients straight at the Traefik VIP -- a second, unintended path to
+    # the same host that skips the public route entirely.
+    dns.renner.uno/public: "true"
+  # No cert-manager annotation and no tls block: the public Ingress above
+  # already owns the certificate for this host. Declaring it twice would
+  # create two Certificates racing for the same wordpress-tls Secret.
+spec:
+  ingressClassName: traefik
+  rules:
+    - host: www.renner.uno
+      http:
+        paths:
+          # The outpost's own OAuth leg. Without this the gate is a dead end,
+          # not a gate: hitting /wp-admin redirects to
+          # /outpost.goauthentik.io/start on THIS host, which otherwise falls
+          # through to the public `/` router and lands on WordPress's 404
+          # page — verified exactly that way on first deploy. paperless needs
+          # the same route and has it as a separate Ingress on auth.renner.uno
+          # (apps/authentik/paperless-gate-ingress.yaml); here the flow stays
+          # on www.renner.uno, so it belongs in this rule set.
+          #
+          # It is not an authentication hole: these endpoints ARE the login
+          # flow, and the outpost serves them itself.
+          - path: /outpost.goauthentik.io
+            pathType: Prefix
+            backend:
+              service:
+                name: ak-outpost-wordpress-gate
+                port:
+                  number: 9000
+          # Covers admin-ajax.php too. That is only safe because the gift
+          # registry plugin -- the sole thing on this site that ever emitted
+          # an ajaxurl -- is being dropped in the migration. Verified against
+          # the live site: 121 admin-ajax hits in the current log window, not
+          # one of them a 200.
+          - path: /wp-admin
+            pathType: Prefix
+            backend:
+              service:
+                name: ak-outpost-wordpress-gate
+                port:
+                  number: 9000
+          - path: /wp-login.php
+            pathType: Exact
+            backend:
+              service:
+                name: ak-outpost-wordpress-gate
+                port:
+                  number: 9000
+          # xmlrpc.php would otherwise be served by the public `/` prefix
+          # router — an earlier version of this file claimed it was "blocked
+          # at the ingress" while nothing blocked it at all. There is no deny
+          # verb in an Ingress, so it goes through the gate instead: anonymous
+          # callers get Authentik, not the XML-RPC endpoint. Nothing on this
+          # site uses it (no Jetpack, no mobile app, no pingbacks worth
+          # keeping), so no legitimate caller is affected.
+          - path: /xmlrpc.php
+            pathType: Exact
+            backend:
+              service:
+                name: ak-outpost-wordpress-gate
+                port:
+                  number: 9000

kubernetes/apps/wordpress/kustomization.yaml

@@ -0,0 +1,9 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+resources:
+  - externalsecret.yaml
+  - mariadb.yaml
+  - pvc.yaml
+  - deployment.yaml
+  - gate-service.yaml
+  - ingress.yaml

kubernetes/apps/wordpress/mariadb.yaml

@@ -0,0 +1,86 @@
+# The database for www.renner.uno. Named `wordpress-mariadb`, not `wp-*`:
+# `wp-pg` in the woodpecker namespace is Woodpecker's Postgres, and reusing
+# the prefix for an unrelated engine invites exactly the wrong guess.
+#
+# Mirrors the CNPG convention: the operator lives in `database`, the database
+# resource lives with its consumer (ak-pg in authentik, im-pg in immich).
+apiVersion: k8s.mariadb.com/v1alpha1
+kind: MariaDB
+metadata:
+  name: wordpress-mariadb
+  namespace: wordpress
+  annotations:
+    # Data guard — a Kustomization rebuild must never take the database with it.
+    kustomize.toolkit.fluxcd.io/prune: disabled
+spec:
+  # generate: true lets the operator create both Secrets on first reconcile.
+  # Neither password is ever typed by a human or parked in OpenBao, so there
+  # is nothing to rotate by hand and nothing to leak through Git.
+  rootPasswordSecretKeyRef:
+    name: wordpress-mariadb-root
+    key: password
+    generate: true
+  username: wordpress
+  passwordSecretKeyRef:
+    name: wordpress-mariadb-app
+    key: password
+    generate: true
+  database: wordpress
+  port: 3306
+  replicas: 1
+
+  storage:
+    size: 5Gi
+    storageClassName: ceph-block
+  # The source database is 527 KB. 5Gi is not a sizing estimate, it is the
+  # smallest allocation that leaves room for binlogs and a decade of posts.
+
+  myCnf: |
+    [mariadb]
+    bind-address=*
+    default_storage_engine=InnoDB
+    innodb_buffer_pool_size=256M
+    # WordPress writes utf8mb4 throughout; the source dump carries
+    # utf8mb4_unicode_520_ci on every table that survives the migration.
+    character_set_server=utf8mb4
+    collation_server=utf8mb4_unicode_520_ci
+    max_allowed_packet=64M
+
+  resources:
+    requests:
+      cpu: 100m
+      memory: 512Mi
+    limits:
+      memory: 1Gi
+---
+# Logical backup to Hetzner Object Storage. Velero only captures PVC data,
+# which for a running database is crash-consistent, not transaction-consistent
+# — that is a snapshot, not a backup you would want to restore from.
+apiVersion: k8s.mariadb.com/v1alpha1
+kind: Backup
+metadata:
+  name: wordpress-mariadb-backup
+  namespace: wordpress
+spec:
+  mariaDbRef:
+    name: wordpress-mariadb
+  schedule:
+    # 02:45, ahead of the Velero window that starts at 03:15.
+    cron: "45 2 * * *"
+    suspend: false
+  maxRetention: 720h # 30d, matching the CNPG ObjectStore retentionPolicy
+  compression: gzip
+  storage:
+    s3:
+      bucket: renner-uno-k8s-mariadb
+      prefix: wordpress
+      endpoint: fsn1.your-objectstorage.com
+      region: fsn1
+      accessKeyIdSecretKeyRef:
+        name: hetzner-s3-creds
+        key: ACCESS_KEY_ID
+      secretAccessKeySecretKeyRef:
+        name: hetzner-s3-creds
+        key: ACCESS_SECRET_KEY
+      tls:
+        enabled: true

kubernetes/apps/wordpress/pvc.yaml

@@ -0,0 +1,35 @@
+# Only wp-content is persisted, not the whole docroot. WordPress core comes
+# from the pinned image, so upgrading the site means bumping a tag in Git
+# rather than clicking "update" in wp-admin — which is the entire point of
+# moving off the Hetzner VM, where core was a mutable pile of PHP on disk.
+#
+# Consequence to be aware of: WordPress's own core updater will fail against
+# a read-only core. Plugin and theme installs still work, because those land
+# in wp-content.
+apiVersion: v1
+kind: PersistentVolumeClaim
+metadata:
+  name: wordpress-content
+  namespace: wordpress
+  labels:
+    # The Velero Schedule selects ON this label rather than excluding the
+    # database PVC by one the mariadb-operator is merely assumed to set. If
+    # that assumption were wrong the exclusion would fail silently and Velero
+    # would snapshot the live database volume; selecting a label this repo
+    # owns cannot fail that way. Same direction as openbao-daily, which picks
+    # its snapshot PVC by app.kubernetes.io/component.
+    app.kubernetes.io/name: wordpress
+    app.kubernetes.io/component: content
+  annotations:
+    kustomize.toolkit.fluxcd.io/prune: disabled
+spec:
+  accessModes:
+    - ReadWriteOnce
+  # RWO is sufficient: a single replica with strategy Recreate. ceph-block is
+  # the cluster default; named explicitly so a default change can't silently
+  # move the media library onto another class.
+  storageClassName: ceph-block
+  resources:
+    requests:
+      # The migrated wp-content is 29 MB compressed, 13 MB of it uploads.
+      storage: 10Gi

kubernetes/clusters/homelab/infrastructure.yaml

@@ -48,6 +48,10 @@ spec:
       kind: HelmRelease
       name: cloudnative-pg # DB operator — seed jobs create databases
       namespace: database
+    - apiVersion: helm.toolkit.fluxcd.io/v2
+      kind: HelmRelease
+      name: mariadb-operator # the MySQL half — WordPress can't use CNPG
+      namespace: database
   # decryption block dropped 2026-07-13: no *.sops.yaml left under this
   # path — secrets come from OpenBao via ESO (MIGRATION-sops-openbao.md).
 ---

kubernetes/clusters/homelab/trmnl.yaml

@@ -0,0 +1,28 @@
+apiVersion: kustomize.toolkit.fluxcd.io/v1
+kind: Kustomization
+metadata:
+  name: trmnl
+  namespace: flux-system
+spec:
+  # Terminus (TRMNL BYOS server) plus its tr-pg CNPG Cluster + ObjectStore +
+  # ScheduledBackup, the Valkey queue, and the uploads/fonts PVC. Drives the
+  # entrance-way e-ink panel; the device polls the cleartext LoadBalancer.
+  interval: 10m
+  retryInterval: 1m
+  timeout: 30m
+  sourceRef:
+    kind: GitRepository
+    name: flux-system
+  path: ./kubernetes/apps/trmnl
+  prune: true
+  wait: true
+  dependsOn:
+    - name: infrastructure
+    - name: infrastructure-crds
+    - name: infrastructure-ceph-cluster
+    - name: infrastructure-cnpg-barman
+  healthCheckExprs:
+    - apiVersion: postgresql.cnpg.io/v1
+      kind: Cluster
+      failed: "has(status.phase) && status.phase.startsWith('Failed')"
+      current: "has(status.phase) && (status.phase == 'Cluster in healthy state' || status.readyInstances > 0)"

kubernetes/clusters/homelab/wordpress.yaml

@@ -0,0 +1,29 @@
+apiVersion: kustomize.toolkit.fluxcd.io/v1
+kind: Kustomization
+metadata:
+  name: wordpress
+  namespace: flux-system
+spec:
+  interval: 10m
+  retryInterval: 2m
+  timeout: 15m
+  sourceRef:
+    kind: GitRepository
+    name: flux-system
+  path: ./kubernetes/apps/wordpress
+  prune: true
+  dependsOn:
+    # infrastructure gates on the mariadb-operator HelmRelease being Ready,
+    # so the MariaDB CRD exists and has a controller before this applies.
+    - name: infrastructure
+    - name: infrastructure-crds
+    - name: infrastructure-ceph-cluster
+  # kstatus has no built-in rule for k8s.mariadb.com resources, so without
+  # this the Kustomization would report Ready the moment the manifests are
+  # accepted — before the database is actually serving. Same reasoning as
+  # the CephCluster expressions in infrastructure.yaml.
+  healthCheckExprs:
+    - apiVersion: k8s.mariadb.com/v1alpha1
+      kind: MariaDB
+      failed: "has(status.conditions) && status.conditions.exists(c, c.type == 'Ready' && c.status == 'False' && c.reason == 'Failed')"
+      current: "has(status.conditions) && status.conditions.exists(c, c.type == 'Ready' && c.status == 'True')"

kubernetes/infrastructure/image-automation/README.md

@@ -96,6 +96,29 @@ The security-scan agent will continue to surface vuln findings on
 these as "Skipped / Flux image-automation owns this" even without
 a policy.
 
+- `ghcr.io/usetrmnl/terminus` — publishes a usable `0.x.0` tag stream, so
+  unlike the three above this is **not** a tag-format problem. It is left
+  unenrolled because the blast radius is hardware: Terminus is pre-1.0,
+  cuts no GitHub Releases and ships no changelog, and it serves the
+  firmware and `/api/display` contract that the TRMNL X panel polls. An
+  unattended bump that changes either can leave a wall-mounted device
+  dark, with no rollback path that doesn't involve the captive portal.
+  Bumps are a reviewed, manual act: read the commit range, then edit the
+  tag in `kubernetes/apps/trmnl/terminus.yaml` (both containers).
+
+  The Valkey sidecar alongside it *is* enrolled (`valkey.yaml`) — it holds
+  only the Sidekiq queue and touches no device-facing contract.
+
+- `ghcr.io/usetrmnl/trmnl-ha-amd64` — the TRMNL screenshot renderer, for
+  the same reason and one more. It owns the dithering pipeline and the
+  webhook payload shape, so a bad bump surfaces as a *corrupted or blank
+  panel*, not as a failed rollout — there is no readiness signal that
+  catches "the image is now unreadable". Separately, upstream publishes
+  **no multi-arch manifest**: only the arch-suffixed `trmnl-ha-amd64` and
+  `trmnl-ha-aarch64` repositories exist, so an ImagePolicy here would also
+  be pinning us to one architecture in a way the tag does not make
+  obvious. Bump manually in `kubernetes/apps/trmnl/renderer.yaml`.
+
 ## Re-import hazard: cnpg-barman-plugin
 
 `kubernetes/infrastructure/cnpg-barman-plugin/manifest.yaml` is an

kubernetes/infrastructure/image-automation/kustomization.yaml

@@ -20,6 +20,7 @@ resources:
   - redis.yaml
   - shelly-manager.yaml
   - tika.yaml
+  - valkey.yaml
   - velero-plugin-for-aws.yaml
   - vsftpd.yaml
   - woodpecker.yaml

kubernetes/infrastructure/image-automation/valkey.yaml

@@ -0,0 +1,27 @@
+apiVersion: image.toolkit.fluxcd.io/v1
+kind: ImageRepository
+metadata:
+  name: valkey-trmnl
+  namespace: flux-system
+spec:
+  image: valkey/valkey
+  interval: 1h
+---
+# Alpine tags only, semver 9.x — matches the major version pinned in
+# upstream Terminus's compose.yml. Valkey holds nothing but the Sidekiq
+# queue here, so minor bumps are safe to take unattended; a 10.x cutover
+# is a deliberate migration.
+apiVersion: image.toolkit.fluxcd.io/v1
+kind: ImagePolicy
+metadata:
+  name: valkey-trmnl
+  namespace: flux-system
+spec:
+  imageRepositoryRef:
+    name: valkey-trmnl
+  filterTags:
+    pattern: '^(?P<ver>\d+\.\d+\.\d+)-alpine$'
+    extract: '$ver'
+  policy:
+    semver:
+      range: "9.x"

kubernetes/infrastructure/kustomization.yaml

@@ -15,6 +15,7 @@ resources:
   - rook-ceph/
   - ceph-csi-drivers/
   - cloudnative-pg/
+  - mariadb-operator/
   - netbird-ingress/
   - netbird-backup/
   - coredns-config/

kubernetes/infrastructure/mariadb-operator/helmrelease.yaml

@@ -0,0 +1,81 @@
+# MariaDB operator, for the one workload in this cluster that cannot use
+# CloudNativePG: WordPress. WordPress speaks MySQL natively — the PostgreSQL
+# route would mean pg4wp, an SQL-rewriting shim whose own README lists
+# WordPress 6.5.3 as its newest tested version while renner.uno runs 7.0.4.
+#
+# The charts are OCI artifacts (ghcr.io), so this follows the OCIRepository
+# pattern from infrastructure/spegel rather than the HelmRepository pattern
+# used for CNPG.
+#
+# CRDs ship as a SEPARATE chart that must be installed first — upstream is
+# explicit about the ordering, hence the dependsOn below rather than a single
+# release with crds: CreateReplace.
+apiVersion: source.toolkit.fluxcd.io/v1
+kind: OCIRepository
+metadata:
+  name: mariadb-operator-crds
+  namespace: database
+spec:
+  interval: 24h
+  url: oci://ghcr.io/mariadb-operator/charts/mariadb-operator-crds
+  ref:
+    # Calendar versioning: the major is the year. Upstream ships an upgrade
+    # guide with each minor (Helm repo deprecations, registry moves), so the
+    # range is capped at the year boundary instead of the usual open ">=".
+    semver: ">=26.6.0 <27.0.0"
+---
+apiVersion: helm.toolkit.fluxcd.io/v2
+kind: HelmRelease
+metadata:
+  name: mariadb-operator-crds
+  namespace: database
+spec:
+  interval: 30m
+  chartRef:
+    kind: OCIRepository
+    name: mariadb-operator-crds
+    namespace: database
+  install:
+    remediation:
+      retries: 3
+  upgrade:
+    remediation:
+      retries: 3
+---
+apiVersion: source.toolkit.fluxcd.io/v1
+kind: OCIRepository
+metadata:
+  name: mariadb-operator
+  namespace: database
+spec:
+  interval: 24h
+  url: oci://ghcr.io/mariadb-operator/charts/mariadb-operator
+  ref:
+    semver: ">=26.6.0 <27.0.0"
+---
+apiVersion: helm.toolkit.fluxcd.io/v2
+kind: HelmRelease
+metadata:
+  name: mariadb-operator
+  namespace: database
+spec:
+  interval: 30m
+  dependsOn:
+    - name: mariadb-operator-crds
+      namespace: database
+  chartRef:
+    kind: OCIRepository
+    name: mariadb-operator
+    namespace: database
+  install:
+    remediation:
+      retries: 3
+  upgrade:
+    remediation:
+      retries: 3
+  values:
+    # The operator watches all namespaces; the MariaDB CR itself lives with
+    # the app in kubernetes/apps/wordpress/, mirroring how CNPG Clusters sit
+    # in their consumer's namespace (ak-pg in authentik, im-pg in immich).
+    metrics:
+      enabled: false

kubernetes/infrastructure/mariadb-operator/kustomization.yaml

@@ -0,0 +1,4 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+resources:
+  - helmrelease.yaml

kubernetes/infrastructure/namespaces/namespaces.yaml

@@ -145,3 +145,18 @@ apiVersion: v1
 kind: Namespace
 metadata:
   name: external-secrets
+---
+# Terminus (TRMNL BYOS server) driving the entrance-way e-ink panel.
+apiVersion: v1
+kind: Namespace
+metadata:
+  name: trmnl
+---
+# WordPress for www.renner.uno, migrated off the Hetzner VM `wordpress-hochzeit`.
+# No privileged label needed: the cluster enforces `baseline`, which permits the
+# official Apache image's root master process binding :80. It does trip the
+# `restricted` audit/warn level — that is expected and not an error.
+apiVersion: v1
+kind: Namespace
+metadata:
+  name: wordpress

kubernetes/infrastructure/network-policies/home-assistant-ns.yaml

@@ -371,3 +371,36 @@ spec:
         - ports:
             - port: "8123"
               protocol: TCP
+---
+# The TRMNL e-ink renderer (trmnl ns) drives a headless browser against the
+# entrance dashboard, so it needs the HA frontend on :8123 — same in-cluster
+# path as the rotator above, and for the same reason: the public names
+# hairpin off the home-assistant-bgp LoadBalancer and are unusable from a
+# pod. It reaches HA at home-assistant-bgp.home-assistant.svc.cluster.local.
+#
+# Only the ingress half is needed here. Unlike the openbao rotator, the trmnl
+# namespace has no default-deny EGRESS covering this pod — terminus-egress is
+# scoped to the Terminus pods by label and does not select the renderer — so
+# there is no source-side rule to add.
+#
+# Source scoped to the renderer pod specifically, not the whole trmnl
+# namespace: Terminus itself sits in that namespace and has no business
+# reaching Home Assistant.
+apiVersion: cilium.io/v2
+kind: CiliumNetworkPolicy
+metadata:
+  name: home-assistant-trmnl-renderer-ingress
+  namespace: home-assistant
+spec:
+  endpointSelector:
+    matchLabels:
+      app.kubernetes.io/name: home-assistant
+  ingress:
+    - fromEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": trmnl
+            app.kubernetes.io/name: trmnl-renderer
+      toPorts:
+        - ports:
+            - port: "8123"
+              protocol: TCP

kubernetes/infrastructure/network-policies/kustomization.yaml

@@ -21,4 +21,6 @@ resources:
   - openbao-ns.yaml
   - external-secrets-ns.yaml
   - woodpecker-ci-ns.yaml
+  - trmnl-ns.yaml
+  - wordpress-ns.yaml
   - allow-cilium-health.yaml

kubernetes/infrastructure/network-policies/trmnl-ns.yaml

@@ -0,0 +1,177 @@
+apiVersion: cilium.io/v2
+kind: CiliumNetworkPolicy
+metadata:
+  name: default-deny-ingress
+  namespace: trmnl
+spec:
+  endpointSelector: {}
+  ingress:
+    # Intra-namespace — terminus ↔ sidekiq ↔ valkey ↔ tr-pg.
+    - fromEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": trmnl
+    # Admin UI via Traefik (trmnl.renner.uno). The device does NOT come
+    # this way — see allow-device-http below.
+    - fromEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": ingress
+            app.kubernetes.io/name: traefik
+    # Prometheus scrape.
+    - fromEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": monitoring
+    # Velero backups via PVC mount.
+    - fromEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": velero
+    # CNPG operator (in `database` ns) reconciles tr-pg — needs to reach
+    # the postgres pods for health, failover, role checks.
+    - fromEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": database
+    - fromEntities:
+        - host
+        - remote-node
+    # ICMP — all standard ICMPv4 + ICMPv6 from any cluster pod. See
+    # home-assistant-ns.yaml for the rationale. INGRESS ONLY.
+    - fromEndpoints:
+        - matchExpressions:
+            - key: io.kubernetes.pod.namespace
+              operator: Exists
+      icmps:
+        - fields:
+            - { type: 0,   family: IPv4 }
+            - { type: 3,   family: IPv4 }
+            - { type: 4,   family: IPv4 }
+            - { type: 5,   family: IPv4 }
+            - { type: 8,   family: IPv4 }
+            - { type: 9,   family: IPv4 }
+            - { type: 10,  family: IPv4 }
+            - { type: 11,  family: IPv4 }
+            - { type: 12,  family: IPv4 }
+            - { type: 13,  family: IPv4 }
+            - { type: 14,  family: IPv4 }
+            - { type: 15,  family: IPv4 }
+            - { type: 16,  family: IPv4 }
+            - { type: 17,  family: IPv4 }
+            - { type: 18,  family: IPv4 }
+            - { type: 30,  family: IPv4 }
+            - { type: 1,   family: IPv6 }
+            - { type: 2,   family: IPv6 }
+            - { type: 3,   family: IPv6 }
+            - { type: 4,   family: IPv6 }
+            - { type: 128, family: IPv6 }
+            - { type: 129, family: IPv6 }
+            - { type: 130, family: IPv6 }
+            - { type: 131, family: IPv6 }
+            - { type: 132, family: IPv6 }
+            - { type: 133, family: IPv6 }
+            - { type: 134, family: IPv6 }
+            - { type: 135, family: IPv6 }
+            - { type: 136, family: IPv6 }
+            - { type: 137, family: IPv6 }
+            - { type: 138, family: IPv6 }
+            - { type: 141, family: IPv6 }
+            - { type: 142, family: IPv6 }
+            - { type: 143, family: IPv6 }
+            - { type: 148, family: IPv6 }
+            - { type: 149, family: IPv6 }
+            - { type: 151, family: IPv6 }
+            - { type: 152, family: IPv6 }
+            - { type: 153, family: IPv6 }
+---
+# The TRMNL panel polls /api/display over cleartext HTTP on the
+# terminus-device LoadBalancer. The firmware cannot be handed the internal
+# CA, so this bypasses Traefik entirely and the namespace default-deny has
+# to be opened for it — same shape as paperless's allow-ftp-scanner.
+#
+# Scoped to the terminus pods and port 2345 only. Requests still have to
+# carry a known device MAC in the `ID` header plus its api_key; this rule
+# opens the port, not the application.
+apiVersion: cilium.io/v2
+kind: CiliumNetworkPolicy
+metadata:
+  name: allow-device-http
+  namespace: trmnl
+spec:
+  endpointSelector:
+    matchLabels:
+      app.kubernetes.io/name: terminus
+  ingress:
+    - fromEntities:
+        - world
+        - host
+        - remote-node
+      toPorts:
+        - ports:
+            - port: "2345"
+              protocol: TCP
+---
+# Egress allowlist for the Terminus pod only.
+#
+# Deliberately NOT applied to the tr-pg CNPG pods: selecting them here
+# would make their egress restricted too, and the barman-cloud WAL archive
+# path to Hetzner object storage is not worth putting behind an FQDN rule
+# that fails closed. They keep unrestricted egress, matching immich/authentik.
+#
+# Terminus needs exactly three outbound destinations, all HTTPS:
+#   - trmnl.com            → GET /api/models, /api/palettes (MODEL_SYNCHRONIZER,
+#                            which is what seeds the TRMNL X model definition —
+#                            1872x1404, 16 colours — into the Models list),
+#                            GET /api/firmware/latest (FIRMWARE_SYNCHRONIZER),
+#                            GET /fonts/<name> (FONT_SYNCHRONIZER)
+#   - trmnl-fw.s3.us-east-2.amazonaws.com
+#                          → the actual firmware binary. This host is NOT a
+#                            constant in the source: it is the `url` field of
+#                            the /api/firmware/latest response, so TRMNL can
+#                            move bucket/region without a Terminus release and
+#                            firmware sync would start failing closed here.
+#                            If firmware sync breaks, re-check that response
+#                            before assuming the policy is wrong.
+#
+# KNOWN LIMITATION: Terminus's "Extensions" (private plugin) feature polls
+# arbitrary admin-configured URLs rendered from a Liquid template stored in
+# the database. Those are unknowable from source and WILL be blocked by this
+# allowlist. We do not use Extensions — screen content is pushed in from the
+# Home Assistant renderer — so this is an accepted trade, not an oversight.
+apiVersion: cilium.io/v2
+kind: CiliumNetworkPolicy
+metadata:
+  name: terminus-egress
+  namespace: trmnl
+spec:
+  endpointSelector:
+    matchLabels:
+      app.kubernetes.io/name: terminus
+  egress:
+    # Intra-namespace — tr-pg-rw (postgres) and terminus-valkey.
+    - toEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": trmnl
+    # DNS. The L7 `rules.dns` block is mandatory, not decorative: Cilium's
+    # DNS proxy only learns an FQDN→IP mapping for names it has actually
+    # observed being resolved, so without this the toFQDNs rules below
+    # never match anything.
+    - toEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": kube-system
+            k8s-app: kube-dns
+      toPorts:
+        - ports:
+            - port: "53"
+              protocol: UDP
+            - port: "53"
+              protocol: TCP
+          rules:
+            dns:
+              - matchPattern: "*"
+    - toFQDNs:
+        - matchName: "trmnl.com"
+        - matchName: "trmnl-fw.s3.us-east-2.amazonaws.com"
+      toPorts:
+        - ports:
+            - port: "443"
+              protocol: TCP
+    - toEntities:
+        - host
+        - remote-node

kubernetes/infrastructure/network-policies/wordpress-ns.yaml

@@ -0,0 +1,167 @@
+apiVersion: cilium.io/v2
+kind: CiliumNetworkPolicy
+metadata:
+  name: default-deny-ingress
+  namespace: wordpress
+spec:
+  endpointSelector: {}
+  ingress:
+    - fromEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": wordpress
+    # Unlike paperless, traefik IS allowed here. www.renner.uno is a public
+    # blog: the front end must be reachable without authentication, so the
+    # public router terminates directly in this namespace.
+    #
+    # Be clear about what this does and does not buy: the public/admin split
+    # is enforced by the two Ingress objects' path rules (`/` straight to the
+    # Service, /wp-admin and /wp-login.php to the gate outpost) and by
+    # Authentik — NOT at the network layer. paperless can point a
+    # workload-scoped CNP at its gate because nothing else may reach it;
+    # here traefik must reach the pod directly for every public page, so the
+    # equivalent rule would be a no-op. Anyone auditing this namespace for
+    # defence in depth should know the network layer is not where the admin
+    # surface is protected.
+    - fromEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": ingress
+            app.kubernetes.io/name: traefik
+    # Scoped to the gate outpost rather than the whole authentik namespace:
+    # only that pod has business reaching wp-admin.
+    - fromEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": authentik
+            goauthentik.io/outpost-name: wordpress-gate
+    # The mariadb-operator reconciles the MariaDB CR from the database
+    # namespace — same cross-namespace requirement CNPG consumers have.
+    - fromEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": database
+    - fromEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": monitoring
+    - fromEntities:
+        - host
+        - remote-node
+    # ICMP — all standard ICMPv4 + ICMPv6 from any cluster pod. See
+    # home-assistant-ns.yaml for the rationale.
+    - fromEndpoints:
+        - matchExpressions:
+            - key: io.kubernetes.pod.namespace
+              operator: Exists
+      icmps:
+        - fields:
+            - { type: 0,   family: IPv4 }
+            - { type: 3,   family: IPv4 }
+            - { type: 4,   family: IPv4 }
+            - { type: 5,   family: IPv4 }
+            - { type: 8,   family: IPv4 }
+            - { type: 9,   family: IPv4 }
+            - { type: 10,  family: IPv4 }
+            - { type: 11,  family: IPv4 }
+            - { type: 12,  family: IPv4 }
+            - { type: 13,  family: IPv4 }
+            - { type: 14,  family: IPv4 }
+            - { type: 15,  family: IPv4 }
+            - { type: 16,  family: IPv4 }
+            - { type: 17,  family: IPv4 }
+            - { type: 18,  family: IPv4 }
+            - { type: 30,  family: IPv4 }
+            - { type: 1,   family: IPv6 }
+            - { type: 2,   family: IPv6 }
+            - { type: 3,   family: IPv6 }
+            - { type: 4,   family: IPv6 }
+            - { type: 128, family: IPv6 }
+            - { type: 129, family: IPv6 }
+            - { type: 130, family: IPv6 }
+            - { type: 131, family: IPv6 }
+            - { type: 132, family: IPv6 }
+            - { type: 133, family: IPv6 }
+            - { type: 134, family: IPv6 }
+            - { type: 135, family: IPv6 }
+            - { type: 136, family: IPv6 }
+            - { type: 137, family: IPv6 }
+            - { type: 138, family: IPv6 }
+            - { type: 141, family: IPv6 }
+            - { type: 142, family: IPv6 }
+            - { type: 143, family: IPv6 }
+            - { type: 148, family: IPv6 }
+            - { type: 149, family: IPv6 }
+            - { type: 151, family: IPv6 }
+            - { type: 152, family: IPv6 }
+            - { type: 153, family: IPv6 }
+  egress:
+    - toEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": wordpress
+    - toEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": authentik
+    - toEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": ingress
+            app.kubernetes.io/name: traefik
+    - toEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": rook-ceph
+    - toEndpoints:
+        - matchLabels:
+            "k8s:io.kubernetes.pod.namespace": kube-system
+            k8s-app: kube-dns
+      toPorts:
+        - ports:
+            - port: "53"
+              protocol: UDP
+    # world covers two things WordPress genuinely needs outbound: the
+    # wordpress.org update/plugin API, and the Hetzner Object Storage
+    # endpoint the MariaDB Backup writes to.
+    - toEntities:
+        - host
+        - remote-node
+        - world
+    # ICMP — egress mirror.
+    - toEndpoints:
+        - matchExpressions:
+            - key: io.kubernetes.pod.namespace
+              operator: Exists
+      icmps:
+        - fields:
+            - { type: 0,   family: IPv4 }
+            - { type: 3,   family: IPv4 }
+            - { type: 4,   family: IPv4 }
+            - { type: 5,   family: IPv4 }
+            - { type: 8,   family: IPv4 }
+            - { type: 9,   family: IPv4 }
+            - { type: 10,  family: IPv4 }
+            - { type: 11,  family: IPv4 }
+            - { type: 12,  family: IPv4 }
+            - { type: 13,  family: IPv4 }
+            - { type: 14,  family: IPv4 }
+            - { type: 15,  family: IPv4 }
+            - { type: 16,  family: IPv4 }
+            - { type: 17,  family: IPv4 }
+            - { type: 18,  family: IPv4 }
+            - { type: 30,  family: IPv4 }
+            - { type: 1,   family: IPv6 }
+            - { type: 2,   family: IPv6 }
+            - { type: 3,   family: IPv6 }
+            - { type: 4,   family: IPv6 }
+            - { type: 128, family: IPv6 }
+            - { type: 129, family: IPv6 }
+            - { type: 130, family: IPv6 }
+            - { type: 131, family: IPv6 }
+            - { type: 132, family: IPv6 }
+            - { type: 133, family: IPv6 }
+            - { type: 134, family: IPv6 }
+            - { type: 135, family: IPv6 }
+            - { type: 136, family: IPv6 }
+            - { type: 137, family: IPv6 }
+            - { type: 138, family: IPv6 }
+            - { type: 141, family: IPv6 }
+            - { type: 142, family: IPv6 }
+            - { type: 143, family: IPv6 }
+            - { type: 148, family: IPv6 }
+            - { type: 149, family: IPv6 }
+            - { type: 151, family: IPv6 }
+            - { type: 152, family: IPv6 }
+            - { type: 153, family: IPv6 }

kubernetes/infrastructure/openbao/rotation/kustomization.yaml

@@ -59,3 +59,22 @@ resources:
   - review-bot.yaml
   - claude-oauth.yaml
   - mosquitto-health.yaml
+# NOT ROTATED, deliberately — recorded here because the whole point of this
+# directory is that unrotated credentials should be a visible choice:
+#
+#   homelab/apps/trmnl/renderer → BYOS_ACCESS_TOKEN / BYOS_REFRESH_TOKEN
+#     Terminus API JWTs used by the e-ink renderer to push screens. Terminus
+#     is configured with API_ACCESS_TOKEN_PERIOD = 10 years
+#     (kubernetes/apps/trmnl/terminus.yaml), so nothing will ever expire
+#     these or prompt a rotation. That is deliberate: the alternative was
+#     holding the Terminus account password in-cluster to mint short-lived
+#     tokens per pod start. There is no rotator because minting requires the
+#     account password, which is exactly what we declined to store.
+#     Rotate BY HAND whenever the Terminus account password changes, by
+#     re-running POST /login and re-seeding both keys.
+#
+#     HA_ACCESS_TOKEN in the same path is a dedicated Home Assistant
+#     long-lived token. `ha-token` mode in secret-rotator already handles
+#     this class (see home-assistant.yaml) — enrolling it is a genuine
+#     follow-up, not a limitation, and would leave only the BYOS pair
+#     unrotated.

kubernetes/infrastructure/post-crds/openbao-clustersecretstore.yaml

@@ -28,7 +28,9 @@ spec:
         - monitoring
         - netbird-backup # netbird-backup-ssh: the VPS-pull backup key
         - paperless
+        - trmnl
         - velero
+        - wordpress
   provider:
     vault: # ESO's provider name; speaks OpenBao fine
       server: http://openbao.openbao.svc.cluster.local:8200

kubernetes/infrastructure/post-crds/velero-schedules.yaml

@@ -283,3 +283,67 @@ spec:
     storageLocation: default
     ttl: 720h0m0s
     snapshotMoveData: true
+---
+apiVersion: velero.io/v1
+kind: Schedule
+metadata:
+  name: trmnl-daily
+  namespace: velero
+  labels:
+    app.kubernetes.io/part-of: homelab-backup
+spec:
+  # Daily at 08:15 local (next free stagger slot after netbird-backup-daily
+  # at 07:45). Selects the whole namespace minus the CNPG-labelled PVCs, so
+  # it picks up terminus-data (device screens + synchronised fonts),
+  # terminus-valkey-data and trmnl-renderer-data (the renderer's schedule
+  # and rendered output) without needing to be edited per PVC. The tr-pg
+  # CNPG cluster is covered separately by Barman, hence the exclusion.
+  schedule: "15 8 * * *"
+  useOwnerReferencesInBackup: false
+  template:
+    includedNamespaces:
+      - trmnl
+    includedResources:
+      - persistentvolumeclaims
+    storageLocation: default
+    ttl: 720h0m0s
+    snapshotMoveData: true
+    labelSelector:
+      matchExpressions:
+        - key: cnpg.io/cluster
+          operator: DoesNotExist
+---
+apiVersion: velero.io/v1
+kind: Schedule
+metadata:
+  name: wordpress-daily
+  namespace: velero
+  labels:
+    app.kubernetes.io/part-of: homelab-backup
+spec:
+  # Daily at 08:45 local (next free stagger slot after trmnl-daily at 08:15).
+  # Covers the wordpress-content PVC: the media library and any plugins or
+  # themes installed through wp-admin. WordPress core is not in here — it
+  # comes from the pinned image, so Git is its backup.
+  schedule: "45 8 * * *"
+  useOwnerReferencesInBackup: false
+  template:
+    includedNamespaces:
+      - wordpress
+    includedResources:
+      - persistentvolumeclaims
+    storageLocation: default
+    ttl: 720h0m0s
+    snapshotMoveData: true
+    # Selects the content PVC by a label this repo sets, rather than
+    # excluding the database PVC by one the mariadb-operator is assumed to
+    # set — an exclusion built on a wrong assumption fails silently and
+    # sweeps the live database volume in. A crash-consistent snapshot of a
+    # running database is not a restorable backup; the operator's own Backup
+    # resource takes a logical dump to Hetzner Object Storage at 02:45, the
+    # same division of labour as CNPG/Barman elsewhere in this file.
+    # openbao-daily picks its PVC the same way.
+    labelSelector:
+      matchLabels:
+        app.kubernetes.io/name: wordpress
+        app.kubernetes.io/component: content

opentofu/object_storage.tf

@@ -42,6 +42,7 @@ locals {
     "renner-uno-k8s-velero",  # Velero PVC backups (Kopia data mover)
     "renner-uno-k8s-secrets", # SOPS talosconfig/kubeconfig for cluster bootstrap
     "renner-uno-tfstate",     # OpenTofu remote state (migrated off the nas-1 Garage S3)
+    "renner-uno-k8s-mariadb", # mariadb-operator logical dumps (wordpress-mariadb)
   ]
 }
 
@@ -95,8 +96,9 @@ locals {
   # precedent in this repo — left uncapped (keep every version) since
   # storage cost there is negligible.
   hetzner_buckets_noncurrent_expiry_days = {
-    "renner-uno-k8s-cnpg"   = 30
-    "renner-uno-k8s-velero" = 30
+    "renner-uno-k8s-cnpg"    = 30
+    "renner-uno-k8s-velero"  = 30
+    "renner-uno-k8s-mariadb" = 30 # Backup maxRetention: 720h (kubernetes/apps/wordpress/mariadb.yaml)
   }
 }