infra/stacks/headscale/modules/headscale/main.tf

522 lines
13 KiB
Terraform
Raw Normal View History

variable "tls_secret_name" {}
variable "tier" { type = string }
variable "headscale_config" {}
variable "headscale_acl" {}
variable "nfs_server" { type = string }
variable "homepage_token" {
type = string
default = ""
sensitive = true
}
variable "ui_cookie_secret" {
type = string
sensitive = true
}
variable "ui_api_key" {
type = string
sensitive = true
}
variable "headscale_derp_map" {
type = string
}
resource "kubernetes_namespace" "headscale" {
metadata {
name = "headscale"
labels = {
tier = var.tier
}
}
[infra] Suppress Goldilocks vpa-update-mode label drift on all namespaces [ci skip] ## Context Wave 3B-continued: the Goldilocks VPA dashboard (stacks/vpa) runs a Kyverno ClusterPolicy `goldilocks-vpa-auto-mode` that mutates every namespace with `metadata.labels["goldilocks.fairwinds.com/vpa-update-mode"] = "off"`. This is intentional — Terraform owns container resource limits, and Goldilocks should only provide recommendations, never auto-update. The label is how Goldilocks decides per-namespace whether to run its VPA in `off` mode. Effect on Terraform: every `kubernetes_namespace` resource shows the label as pending-removal (`-> null`) on every `scripts/tg plan`. Dawarich survey 2026-04-18 confirmed the drift. Cluster-side count: 88 namespaces carry the label (`kubectl get ns -o json | jq ... | wc -l`). Every TF-managed namespace is affected. This commit brings the intentional admission drift under the same `# KYVERNO_LIFECYCLE_V1` discoverability marker introduced in c9d221d5 for the ndots dns_config pattern. The marker now stands generically for any Kyverno admission-webhook drift suppression; the inline comment records which specific policy stamps which specific field so future grep audits show why each suppression exists. ## This change 107 `.tf` files touched — every stack's `resource "kubernetes_namespace"` resource gets: ```hcl lifecycle { # KYVERNO_LIFECYCLE_V1: goldilocks-vpa-auto-mode ClusterPolicy stamps this label on every namespace ignore_changes = [metadata[0].labels["goldilocks.fairwinds.com/vpa-update-mode"]] } ``` Injection was done with a brace-depth-tracking Python pass (`/tmp/add_goldilocks_ignore.py`): match `^resource "kubernetes_namespace" ` → track `{` / `}` until the outermost closing brace → insert the lifecycle block before the closing brace. The script is idempotent (skips any file that already mentions `goldilocks.fairwinds.com/vpa-update-mode`) so re-running is safe. Vault stack picked up 2 namespaces in the same file (k8s-users produces one, plus a second explicit ns) — confirmed via file diff (+8 lines). ## What is NOT in this change - `stacks/trading-bot/main.tf` — entire file is `/* … */` commented out (paused 2026-04-06 per user decision). Reverted after the script ran. - `stacks/_template/main.tf.example` — per-stack skeleton, intentionally minimal. User keeps it that way. Not touched by the script (file has no real `resource "kubernetes_namespace"` — only a placeholder comment). - `.terraform/` copies (e.g. `stacks/metallb/.terraform/modules/...`) — gitignored, won't commit; the live path was edited. - `terraform fmt` cleanup of adjacent pre-existing alignment issues in authentik, freedify, hermes-agent, nvidia, vault, meshcentral. Reverted to keep the commit scoped to the Goldilocks sweep. Those files will need a separate fmt-only commit or will be cleaned up on next real apply to that stack. ## Verification Dawarich (one of the hundred-plus touched stacks) showed the pattern before and after: ``` $ cd stacks/dawarich && ../../scripts/tg plan Before: Plan: 0 to add, 2 to change, 0 to destroy. # kubernetes_namespace.dawarich will be updated in-place (goldilocks.fairwinds.com/vpa-update-mode -> null) # module.tls_secret.kubernetes_secret.tls_secret will be updated in-place (Kyverno generate.* labels — fixed in 8d94688d) After: No changes. Your infrastructure matches the configuration. ``` Injection count check: ``` $ rg -c 'KYVERNO_LIFECYCLE_V1: goldilocks-vpa-auto-mode' stacks/ | awk -F: '{s+=$2} END {print s}' 108 ``` ## Reproduce locally 1. `git pull` 2. Pick any stack: `cd stacks/<name> && ../../scripts/tg plan` 3. Expect: no drift on the namespace's goldilocks.fairwinds.com/vpa-update-mode label. Closes: code-dwx Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-18 21:15:27 +00:00
lifecycle {
# KYVERNO_LIFECYCLE_V1: goldilocks-vpa-auto-mode ClusterPolicy stamps this label on every namespace
ignore_changes = [metadata[0].labels["goldilocks.fairwinds.com/vpa-update-mode"]]
}
}
module "tls_secret" {
source = "../../../../modules/kubernetes/setup_tls_secret"
namespace = kubernetes_namespace.headscale.metadata[0].name
tls_secret_name = var.tls_secret_name
}
module "nfs_data_host" {
source = "../../../../modules/kubernetes/nfs_volume"
name = "headscale-data-host"
namespace = kubernetes_namespace.headscale.metadata[0].name
nfs_server = "192.168.1.127"
nfs_path = "/srv/nfs/headscale"
}
feat(storage): migrate all sensitive services to proxmox-lvm-encrypted Reconcile Terraform with cluster state after manual encrypted PVC migrations and complete the remaining unfinished migrations. All services storing sensitive data now use LUKS2-encrypted block storage via the Proxmox CSI plugin. ## Context Only Technitium DNS was using encrypted storage in Terraform. Many services had been manually migrated to encrypted PVCs in the cluster, but Terraform was never updated — creating dangerous state drift where a `tg apply` could recreate unencrypted PVCs. ## This change Phase 0 — Infrastructure: - Add `proxmox-lvm-encrypted` StorageClass to Helm values (extraParameters) - Add ExternalSecret for LUKS encryption passphrase to Terraform - Fix CSI node plugin memory: `node.plugin.resources` (not `node.resources`) with 1280Mi limit for LUKS2 Argon2id key derivation Phase 1 — TF state reconciliation (zero downtime): - Health, Matrix, N8N, Forgejo, Vaultwarden, Mailserver: state rm + import - Redis, DBAAS MySQL, DBAAS PostgreSQL: Helm/CNPG value updates Phase 2 — Data migration (encrypted PVCs existed but unused): - Headscale, Frigate, MeshCentral: rsync + switchover - Nextcloud (20Gi): rsync + chart_values update Phase 3 — New encrypted PVCs: - Roundcube HTML, HackMD, Affine, DBAAS pgadmin: create + rsync + switchover Phase 4 — Cleanup: - Deleted 5 orphaned unencrypted PVCs ## Services migrated (18 PVCs across 14 namespaces) ``` vaultwarden → vaultwarden-data-encrypted dbaas → datadir-mysql-cluster-0, pg-cluster-{1,2}, dbaas-pgadmin-encrypted mailserver → mailserver-data-encrypted, roundcubemail-{enigma,html}-encrypted nextcloud → nextcloud-data-encrypted forgejo → forgejo-data-encrypted matrix → matrix-data-encrypted n8n → n8n-data-encrypted affine → affine-data-encrypted health → health-uploads-encrypted hackmd → hackmd-data-encrypted redis → redis-data-redis-node-{0,1} headscale → headscale-data-encrypted frigate → frigate-config-encrypted meshcentral → meshcentral-{data,files}-encrypted ``` Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-15 20:15:30 +00:00
resource "kubernetes_persistent_volume_claim" "data_encrypted" {
wait_until_bound = false
metadata {
feat(storage): migrate all sensitive services to proxmox-lvm-encrypted Reconcile Terraform with cluster state after manual encrypted PVC migrations and complete the remaining unfinished migrations. All services storing sensitive data now use LUKS2-encrypted block storage via the Proxmox CSI plugin. ## Context Only Technitium DNS was using encrypted storage in Terraform. Many services had been manually migrated to encrypted PVCs in the cluster, but Terraform was never updated — creating dangerous state drift where a `tg apply` could recreate unencrypted PVCs. ## This change Phase 0 — Infrastructure: - Add `proxmox-lvm-encrypted` StorageClass to Helm values (extraParameters) - Add ExternalSecret for LUKS encryption passphrase to Terraform - Fix CSI node plugin memory: `node.plugin.resources` (not `node.resources`) with 1280Mi limit for LUKS2 Argon2id key derivation Phase 1 — TF state reconciliation (zero downtime): - Health, Matrix, N8N, Forgejo, Vaultwarden, Mailserver: state rm + import - Redis, DBAAS MySQL, DBAAS PostgreSQL: Helm/CNPG value updates Phase 2 — Data migration (encrypted PVCs existed but unused): - Headscale, Frigate, MeshCentral: rsync + switchover - Nextcloud (20Gi): rsync + chart_values update Phase 3 — New encrypted PVCs: - Roundcube HTML, HackMD, Affine, DBAAS pgadmin: create + rsync + switchover Phase 4 — Cleanup: - Deleted 5 orphaned unencrypted PVCs ## Services migrated (18 PVCs across 14 namespaces) ``` vaultwarden → vaultwarden-data-encrypted dbaas → datadir-mysql-cluster-0, pg-cluster-{1,2}, dbaas-pgadmin-encrypted mailserver → mailserver-data-encrypted, roundcubemail-{enigma,html}-encrypted nextcloud → nextcloud-data-encrypted forgejo → forgejo-data-encrypted matrix → matrix-data-encrypted n8n → n8n-data-encrypted affine → affine-data-encrypted health → health-uploads-encrypted hackmd → hackmd-data-encrypted redis → redis-data-redis-node-{0,1} headscale → headscale-data-encrypted frigate → frigate-config-encrypted meshcentral → meshcentral-{data,files}-encrypted ``` Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-15 20:15:30 +00:00
name = "headscale-data-encrypted"
namespace = kubernetes_namespace.headscale.metadata[0].name
annotations = {
"resize.topolvm.io/threshold" = "10%"
"resize.topolvm.io/increase" = "100%"
"resize.topolvm.io/storage_limit" = "5Gi"
}
}
spec {
access_modes = ["ReadWriteOnce"]
feat(storage): migrate all sensitive services to proxmox-lvm-encrypted Reconcile Terraform with cluster state after manual encrypted PVC migrations and complete the remaining unfinished migrations. All services storing sensitive data now use LUKS2-encrypted block storage via the Proxmox CSI plugin. ## Context Only Technitium DNS was using encrypted storage in Terraform. Many services had been manually migrated to encrypted PVCs in the cluster, but Terraform was never updated — creating dangerous state drift where a `tg apply` could recreate unencrypted PVCs. ## This change Phase 0 — Infrastructure: - Add `proxmox-lvm-encrypted` StorageClass to Helm values (extraParameters) - Add ExternalSecret for LUKS encryption passphrase to Terraform - Fix CSI node plugin memory: `node.plugin.resources` (not `node.resources`) with 1280Mi limit for LUKS2 Argon2id key derivation Phase 1 — TF state reconciliation (zero downtime): - Health, Matrix, N8N, Forgejo, Vaultwarden, Mailserver: state rm + import - Redis, DBAAS MySQL, DBAAS PostgreSQL: Helm/CNPG value updates Phase 2 — Data migration (encrypted PVCs existed but unused): - Headscale, Frigate, MeshCentral: rsync + switchover - Nextcloud (20Gi): rsync + chart_values update Phase 3 — New encrypted PVCs: - Roundcube HTML, HackMD, Affine, DBAAS pgadmin: create + rsync + switchover Phase 4 — Cleanup: - Deleted 5 orphaned unencrypted PVCs ## Services migrated (18 PVCs across 14 namespaces) ``` vaultwarden → vaultwarden-data-encrypted dbaas → datadir-mysql-cluster-0, pg-cluster-{1,2}, dbaas-pgadmin-encrypted mailserver → mailserver-data-encrypted, roundcubemail-{enigma,html}-encrypted nextcloud → nextcloud-data-encrypted forgejo → forgejo-data-encrypted matrix → matrix-data-encrypted n8n → n8n-data-encrypted affine → affine-data-encrypted health → health-uploads-encrypted hackmd → hackmd-data-encrypted redis → redis-data-redis-node-{0,1} headscale → headscale-data-encrypted frigate → frigate-config-encrypted meshcentral → meshcentral-{data,files}-encrypted ``` Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-15 20:15:30 +00:00
storage_class_name = "proxmox-lvm-encrypted"
resources {
requests = {
storage = "1Gi"
}
}
}
}
resource "kubernetes_deployment" "headscale" {
metadata {
name = "headscale"
namespace = kubernetes_namespace.headscale.metadata[0].name
labels = {
app = "headscale"
tier = var.tier
# scare to try but probably non-http will fail
# "istio-injection" : "enabled"
}
annotations = {
"reloader.stakater.com/search" = "true"
}
}
spec {
replicas = 1
strategy {
type = "Recreate"
}
selector {
match_labels = {
app = "headscale"
}
}
template {
metadata {
labels = {
app = "headscale"
}
annotations = {
# "diun.enable" = "true"
"diun.enable" = "false"
"diun.include_tags" = "^\\d+(?:\\.\\d+)?(?:\\.\\d+)?$"
}
}
spec {
container {
image = "headscale/headscale:0.28.0"
# image = "headscale/headscale:0.28.0-debug" # -debug is for debug images
name = "headscale"
command = ["headscale", "serve"]
resources {
requests = {
cpu = "50m"
memory = "128Mi"
}
limits = {
memory = "128Mi"
}
}
port {
container_port = 8080
}
port {
container_port = 9090
}
port {
container_port = 41641
}
port {
container_port = 3479
protocol = "UDP"
}
liveness_probe {
http_get {
path = "/health"
port = 8080
}
initial_delay_seconds = 15
period_seconds = 30
timeout_seconds = 5
failure_threshold = 5
}
readiness_probe {
http_get {
path = "/health"
port = 8080
}
initial_delay_seconds = 5
period_seconds = 30
timeout_seconds = 5
failure_threshold = 3
}
volume_mount {
name = "config-volume"
mount_path = "/etc/headscale"
}
volume_mount {
mount_path = "/mnt"
name = "nfs-config"
}
}
volume {
name = "config-volume"
config_map {
name = "headscale-config"
items {
key = "config.yaml"
path = "config.yaml"
}
items {
key = "acl.yaml"
path = "acl.yaml"
}
items {
key = "derp.yaml"
path = "derp.yaml"
}
}
}
volume {
name = "nfs-config"
persistent_volume_claim {
feat(storage): migrate all sensitive services to proxmox-lvm-encrypted Reconcile Terraform with cluster state after manual encrypted PVC migrations and complete the remaining unfinished migrations. All services storing sensitive data now use LUKS2-encrypted block storage via the Proxmox CSI plugin. ## Context Only Technitium DNS was using encrypted storage in Terraform. Many services had been manually migrated to encrypted PVCs in the cluster, but Terraform was never updated — creating dangerous state drift where a `tg apply` could recreate unencrypted PVCs. ## This change Phase 0 — Infrastructure: - Add `proxmox-lvm-encrypted` StorageClass to Helm values (extraParameters) - Add ExternalSecret for LUKS encryption passphrase to Terraform - Fix CSI node plugin memory: `node.plugin.resources` (not `node.resources`) with 1280Mi limit for LUKS2 Argon2id key derivation Phase 1 — TF state reconciliation (zero downtime): - Health, Matrix, N8N, Forgejo, Vaultwarden, Mailserver: state rm + import - Redis, DBAAS MySQL, DBAAS PostgreSQL: Helm/CNPG value updates Phase 2 — Data migration (encrypted PVCs existed but unused): - Headscale, Frigate, MeshCentral: rsync + switchover - Nextcloud (20Gi): rsync + chart_values update Phase 3 — New encrypted PVCs: - Roundcube HTML, HackMD, Affine, DBAAS pgadmin: create + rsync + switchover Phase 4 — Cleanup: - Deleted 5 orphaned unencrypted PVCs ## Services migrated (18 PVCs across 14 namespaces) ``` vaultwarden → vaultwarden-data-encrypted dbaas → datadir-mysql-cluster-0, pg-cluster-{1,2}, dbaas-pgadmin-encrypted mailserver → mailserver-data-encrypted, roundcubemail-{enigma,html}-encrypted nextcloud → nextcloud-data-encrypted forgejo → forgejo-data-encrypted matrix → matrix-data-encrypted n8n → n8n-data-encrypted affine → affine-data-encrypted health → health-uploads-encrypted hackmd → hackmd-data-encrypted redis → redis-data-redis-node-{0,1} headscale → headscale-data-encrypted frigate → frigate-config-encrypted meshcentral → meshcentral-{data,files}-encrypted ``` Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-15 20:15:30 +00:00
claim_name = kubernetes_persistent_volume_claim.data_encrypted.metadata[0].name
}
}
# container {
# image = "simcu/headscale-ui:0.1.4"
# name = "headscale-ui"
# port {
# container_port = 80
# }
# }
container {
image = "ghcr.io/gurucomputing/headscale-ui@sha256:015f5ba04bcbd5ee03178540a1dbbfc97b6896d7411032e3bf33c2f3e08f8b6f"
# image = "ghcr.io/tale/headplane:0.3.2"
name = "headscale-ui"
resources {
requests = {
cpu = "25m"
memory = "128Mi"
}
limits = {
memory = "128Mi"
}
}
port {
container_port = 8081
# container_port = 3000
}
env {
name = "HTTP_PORT"
value = "8081"
}
# env {
# name = "HTTPS_PORT"
# value = "8082"
# }
env {
name = "HEADSCALE_URL"
value = "http://localhost:8080"
}
env {
name = "COOKIE_SECRET"
value = var.ui_cookie_secret
}
env {
name = "ROOT_API_KEY"
value = var.ui_api_key
}
}
dns_config {
option {
name = "ndots"
value = "2"
}
}
}
}
}
[infra] Sweep dns_config ignore_changes across all pod-owning resources [ci skip] ## Context Wave 3A (commit c9d221d5) added the `# KYVERNO_LIFECYCLE_V1` marker to the 27 pre-existing `ignore_changes = [...dns_config]` sites so they could be grepped and audited. It did NOT address pod-owning resources that were simply missing the suppression entirely. Post-Wave-3A sampling (2026-04-18) found that navidrome, f1-stream, frigate, servarr, monitoring, crowdsec, and many other stacks showed perpetual `dns_config` drift every plan because their `kubernetes_deployment` / `kubernetes_stateful_set` / `kubernetes_cron_job_v1` resources had no `lifecycle {}` block at all. Root cause (same as Wave 3A): Kyverno's admission webhook stamps `dns_config { option { name = "ndots"; value = "2" } }` on every pod's `spec.template.spec.dns_config` to prevent NxDomain search-domain flooding (see `k8s-ndots-search-domain-nxdomain-flood` skill). Without `ignore_changes` on every Terraform-managed pod-owner, Terraform repeatedly tries to strip the injected field. ## This change Extends the Wave 3A convention by sweeping EVERY `kubernetes_deployment`, `kubernetes_stateful_set`, `kubernetes_daemon_set`, `kubernetes_cron_job_v1`, `kubernetes_job_v1` (+ their `_v1` variants) in the repo and ensuring each carries the right `ignore_changes` path: - **kubernetes_deployment / stateful_set / daemon_set / job_v1**: `spec[0].template[0].spec[0].dns_config` - **kubernetes_cron_job_v1**: `spec[0].job_template[0].spec[0].template[0].spec[0].dns_config` (extra `job_template[0]` nesting — the CronJob's PodTemplateSpec is one level deeper) Each injection / extension is tagged `# KYVERNO_LIFECYCLE_V1: Kyverno admission webhook mutates dns_config with ndots=2` inline so the suppression is discoverable via `rg 'KYVERNO_LIFECYCLE_V1' stacks/`. Two insertion paths are handled by a Python pass (`/tmp/add_dns_config_ignore.py`): 1. **No existing `lifecycle {}`**: inject a brand-new block just before the resource's closing `}`. 108 new blocks on 93 files. 2. **Existing `lifecycle {}` (usually for `DRIFT_WORKAROUND: CI owns image tag` from Wave 4, commit a62b43d1)**: extend its `ignore_changes` list with the dns_config path. Handles both inline (`= [x]`) and multiline (`= [\n x,\n]`) forms; ensures the last pre-existing list item carries a trailing comma so the extended list is valid HCL. 34 extensions. The script skips anything already mentioning `dns_config` inside an `ignore_changes`, so re-running is a no-op. ## Scale - 142 total lifecycle injections/extensions - 93 `.tf` files touched - 108 brand-new `lifecycle {}` blocks + 34 extensions of existing ones - Every Tier 0 and Tier 1 stack with a pod-owning resource is covered - Together with Wave 3A's 27 pre-existing markers → **169 greppable `KYVERNO_LIFECYCLE_V1` dns_config sites across the repo** ## What is NOT in this change - `stacks/trading-bot/main.tf` — entirely commented-out block (`/* … */`). Python script touched the file, reverted manually. - `_template/main.tf.example` skeleton — kept minimal on purpose; any future stack created from it should either inherit the Wave 3A one-line form or add its own on first `kubernetes_deployment`. - `terraform fmt` fixes to pre-existing alignment issues in meshcentral, nvidia/modules/nvidia, vault — unrelated to this commit. Left for a separate fmt-only pass. - Non-pod resources (`kubernetes_service`, `kubernetes_secret`, `kubernetes_manifest`, etc.) — they don't own pods so they don't get Kyverno dns_config mutation. ## Verification Random sample post-commit: ``` $ cd stacks/navidrome && ../../scripts/tg plan → No changes. $ cd stacks/f1-stream && ../../scripts/tg plan → No changes. $ cd stacks/frigate && ../../scripts/tg plan → No changes. $ rg -c 'KYVERNO_LIFECYCLE_V1' stacks/ --include='*.tf' --include='*.tf.example' \ | awk -F: '{s+=$2} END {print s}' 169 ``` ## Reproduce locally 1. `git pull` 2. `rg 'KYVERNO_LIFECYCLE_V1' stacks/ | wc -l` → 169+ 3. `cd stacks/navidrome && ../../scripts/tg plan` → expect 0 drift on the deployment's dns_config field. Refs: code-seq (Wave 3B dns_config class closed; kubernetes_manifest annotation class handled separately in 8d94688d for tls_secret) Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-18 21:19:48 +00:00
lifecycle {
# KYVERNO_LIFECYCLE_V1: Kyverno admission webhook mutates dns_config with ndots=2
ignore_changes = [spec[0].template[0].spec[0].dns_config]
}
}
resource "kubernetes_service" "headscale" {
metadata {
name = "headscale"
namespace = kubernetes_namespace.headscale.metadata[0].name
labels = {
"app" = "headscale"
}
annotations = {
"prometheus.io/scrape" = "true"
"prometheus.io/port" = "9090"
}
# annotations = {
# "metallb.universe.tf/allow-shared-ip" : "shared"
# }
}
spec {
# type = "LoadBalancer"
# external_traffic_policy = "Cluster"
selector = {
app = "headscale"
}
port {
name = "headscale"
port = "8080"
protocol = "TCP"
}
port {
name = "headscale-ui"
port = "80"
target_port = 8081
# target_port = 3000
protocol = "TCP"
}
port {
name = "metrics"
port = "9090"
protocol = "TCP"
}
}
}
module "ingress" {
ingress_factory: replace `protected` bool with `auth` enum + audit pass across 100 stacks Phase 3+4 of default-deny ingress plan. Replaces the `protected = bool` (default false → unprotected) variable in `modules/kubernetes/ingress_factory` with `auth = string` enum (default "required" → fail-closed). Touches every ingress_factory caller so the audit decision is recorded explicitly in code. ingress_factory (Phase 3): - `auth = "required"`: standard Authentik forward-auth (the legacy `protected = true` semantic). - `auth = "public"`: forward-auth via the new `authentik-forward-auth-public` middleware → dedicated public outpost → guest auto-bind. Logged-in users keep their real identity. - `auth = "none"`: no Authentik middleware. For Anubis-fronted content, native client APIs (Git, /v2/, WebDAV), webhook receivers, the Authentik outpost itself. - `effective_anti_ai` default flips ON only when `auth = "none"` (auth-gated ingresses don't need anti-AI noise; the auth flow already discourages bots). Audit pass (Phase 4) across 96 ingress_factory call sites: - 49 explicit `protected = true` → `auth = "required"` - 8 explicit `protected = false` → `auth = "none"` (5) or `auth = "public"` (3) - 64 previously-default (no protected line) → `auth = "required"` ADDED, then reviewed individually: * 9 Anubis-fronted (blog, www, kms, travel, f1, cyberchef, jsoncrack, homepage, wrongmove UI, privatebin) → `auth = "none"` * 22 native-client / programmatic surfaces (Forgejo Git+/v2/, webhook handler, claude-memory MCP, Nextcloud WebDAV, Matrix, Vault CLI/OIDC, xray VPN, ntfy, woodpecker webhooks, n8n triggers, ntfy push, dawarich location ingestion, immich frame kiosk, headscale CP, send anonymous drops, rybbit beacon, vaultwarden API, Authentik UI itself + outposts) → `auth = "none"` * Remaining ~33 → `auth = "required"` confirmed (admin tools, internal UIs, services without app-level auth) - Smoke-test promotions to `auth = "public"`: fire-planner public UI, k8s-portal API, insta2spotify callback. Three call sites in wrapper modules (`stacks/freedify/factory/`, `stacks/reverse-proxy/modules/reverse_proxy/`) keep their internal `protected` bool — they translate to `auth` internally, out of scope for this rename. Behavior change: previously-default ingresses now fail closed (require Authentik login) unless explicitly flipped to `auth = "none"` or `auth = "public"`. This is the audit goal — no more accidentally-unprotected surfaces. Sites that were intentionally public (Anubis content, native APIs, webhooks) are now explicitly recorded as `auth = "none"`. Drive-by: `modules/create-vm/main.tf` picked up cosmetic alignment via `terraform fmt -recursive` during the audit. Behavior-neutral. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
2026-05-10 18:53:49 +00:00
source = "../../../../modules/kubernetes/ingress_factory"
# Headscale is the Tailscale control plane — native Tailscale clients
# register, exchange keys, and pull DERP maps from headscale.viktorbarzin.me.
# Forward-auth would break every Tailscale client. Headscale has its own
# OIDC + preauth-key auth at the app layer; the web admin UI lives on a
# separate /web ingress that remains auth=required.
auth = "none"
[infra] Auto-create Cloudflare DNS records from ingress_factory ## Context Deploying new services required manually adding hostnames to cloudflare_proxied_names/cloudflare_non_proxied_names in config.tfvars — a separate file from the service stack. This was frequently forgotten, leaving services unreachable externally. ## This change: - Add `dns_type` parameter to `ingress_factory` and `reverse_proxy/factory` modules. Setting `dns_type = "proxied"` or `"non-proxied"` auto-creates the Cloudflare DNS record (CNAME to tunnel or A/AAAA to public IP). - Simplify cloudflared tunnel from 100 per-hostname rules to wildcard `*.viktorbarzin.me → Traefik`. Traefik still handles host-based routing. - Add global Cloudflare provider via terragrunt.hcl (separate cloudflare_provider.tf with Vault-sourced API key). - Migrate 118 hostnames from centralized config.tfvars to per-service dns_type. 17 hostnames remain centrally managed (Helm ingresses, special cases). - Update docs, AGENTS.md, CLAUDE.md, dns.md runbook. ``` BEFORE AFTER config.tfvars (manual list) stacks/<svc>/main.tf | module "ingress" { v dns_type = "proxied" stacks/cloudflared/ } for_each = list | cloudflare_record auto-creates tunnel per-hostname cloudflare_record + annotation ``` ## What is NOT in this change: - Uptime Kuma monitor migration (still reads from config.tfvars) - 17 remaining centrally-managed hostnames (Helm, special cases) - Removal of allow_overwrite (keep until migration confirmed stable) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-16 13:45:04 +00:00
dns_type = "non-proxied"
namespace = kubernetes_namespace.headscale.metadata[0].name
name = "headscale"
port = 8080
tls_secret_name = var.tls_secret_name
extra_annotations = {
"gethomepage.dev/enabled" = "true"
"gethomepage.dev/name" = "Headscale"
"gethomepage.dev/description" = "VPN mesh network"
"gethomepage.dev/icon" = "headscale.png"
"gethomepage.dev/group" = "Identity & Security"
"gethomepage.dev/pod-selector" = ""
}
}
# Dedicated IngressRoute for DERP — bypasses CrowdSec, rate limiting, anti-AI,
# and error pages middlewares that interfere with the Upgrade: DERP protocol.
resource "kubernetes_manifest" "derp_ingress_route" {
manifest = {
apiVersion = "traefik.io/v1alpha1"
kind = "IngressRoute"
metadata = {
name = "headscale-derp"
namespace = kubernetes_namespace.headscale.metadata[0].name
}
spec = {
entryPoints = ["websecure"]
routes = [{
match = "Host(`headscale.viktorbarzin.me`) && PathPrefix(`/derp`)"
kind = "Rule"
services = [{
name = kubernetes_service.headscale.metadata[0].name
port = 8080
}]
# Minimal middleware — retry + rate-limit. No CrowdSec/anti-AI (DERP is a relay protocol)
middlewares = [
{
name = "retry"
namespace = "traefik"
},
{
name = "rate-limit"
namespace = "traefik"
},
]
}]
tls = {
secretName = var.tls_secret_name
}
}
}
}
module "ingress-ui" {
source = "../../../../modules/kubernetes/ingress_factory"
ingress_factory: replace `protected` bool with `auth` enum + audit pass across 100 stacks Phase 3+4 of default-deny ingress plan. Replaces the `protected = bool` (default false → unprotected) variable in `modules/kubernetes/ingress_factory` with `auth = string` enum (default "required" → fail-closed). Touches every ingress_factory caller so the audit decision is recorded explicitly in code. ingress_factory (Phase 3): - `auth = "required"`: standard Authentik forward-auth (the legacy `protected = true` semantic). - `auth = "public"`: forward-auth via the new `authentik-forward-auth-public` middleware → dedicated public outpost → guest auto-bind. Logged-in users keep their real identity. - `auth = "none"`: no Authentik middleware. For Anubis-fronted content, native client APIs (Git, /v2/, WebDAV), webhook receivers, the Authentik outpost itself. - `effective_anti_ai` default flips ON only when `auth = "none"` (auth-gated ingresses don't need anti-AI noise; the auth flow already discourages bots). Audit pass (Phase 4) across 96 ingress_factory call sites: - 49 explicit `protected = true` → `auth = "required"` - 8 explicit `protected = false` → `auth = "none"` (5) or `auth = "public"` (3) - 64 previously-default (no protected line) → `auth = "required"` ADDED, then reviewed individually: * 9 Anubis-fronted (blog, www, kms, travel, f1, cyberchef, jsoncrack, homepage, wrongmove UI, privatebin) → `auth = "none"` * 22 native-client / programmatic surfaces (Forgejo Git+/v2/, webhook handler, claude-memory MCP, Nextcloud WebDAV, Matrix, Vault CLI/OIDC, xray VPN, ntfy, woodpecker webhooks, n8n triggers, ntfy push, dawarich location ingestion, immich frame kiosk, headscale CP, send anonymous drops, rybbit beacon, vaultwarden API, Authentik UI itself + outposts) → `auth = "none"` * Remaining ~33 → `auth = "required"` confirmed (admin tools, internal UIs, services without app-level auth) - Smoke-test promotions to `auth = "public"`: fire-planner public UI, k8s-portal API, insta2spotify callback. Three call sites in wrapper modules (`stacks/freedify/factory/`, `stacks/reverse-proxy/modules/reverse_proxy/`) keep their internal `protected` bool — they translate to `auth` internally, out of scope for this rename. Behavior change: previously-default ingresses now fail closed (require Authentik login) unless explicitly flipped to `auth = "none"` or `auth = "public"`. This is the audit goal — no more accidentally-unprotected surfaces. Sites that were intentionally public (Anubis content, native APIs, webhooks) are now explicitly recorded as `auth = "none"`. Drive-by: `modules/create-vm/main.tf` picked up cosmetic alignment via `terraform fmt -recursive` during the audit. Behavior-neutral. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
2026-05-10 18:53:49 +00:00
auth = "required"
namespace = kubernetes_namespace.headscale.metadata[0].name
name = "headscale-ui"
host = "headscale"
service_name = "headscale"
port = 80
ingress_path = ["/web"]
tls_secret_name = var.tls_secret_name
}
resource "kubernetes_service" "headscale-server" {
metadata {
name = "headscale-server"
namespace = kubernetes_namespace.headscale.metadata[0].name
labels = {
"app" = "headscale"
}
annotations = {
"metallb.io/loadBalancerIPs" = "10.0.20.200"
"metallb.io/allow-shared-ip" = "shared"
}
}
spec {
type = "LoadBalancer"
external_traffic_policy = "Cluster"
selector = {
app = "headscale"
}
# port {
# name = "headscale-tcp"
# port = "41641"
# protocol = "TCP"
# }
port {
name = "headscale-udp"
port = "41641"
protocol = "UDP"
}
port {
name = "stun"
port = "3479"
protocol = "UDP"
}
}
}
resource "kubernetes_config_map" "headscale-config" {
metadata {
name = "headscale-config"
namespace = kubernetes_namespace.headscale.metadata[0].name
annotations = {
"reloader.stakater.com/match" = "true"
}
}
data = {
"config.yaml" = var.headscale_config
"acl.yaml" = var.headscale_acl
"derp.yaml" = var.headscale_derp_map
}
}
# Backup CronJob — sqlite3 .backup from proxmox-lvm to NFS for cloud sync pickup
# Uses pod_affinity to co-locate with headscale pod (required for RWO PVC access)
resource "kubernetes_cron_job_v1" "headscale_backup" {
metadata {
name = "headscale-backup"
namespace = kubernetes_namespace.headscale.metadata[0].name
}
spec {
concurrency_policy = "Replace"
schedule = "0 */6 * * *"
successful_jobs_history_limit = 1
failed_jobs_history_limit = 3
job_template {
metadata {}
spec {
backoff_limit = 3
ttl_seconds_after_finished = 10
template {
metadata {}
spec {
affinity {
pod_affinity {
required_during_scheduling_ignored_during_execution {
label_selector {
match_labels = {
app = "headscale"
}
}
topology_key = "kubernetes.io/hostname"
}
}
}
container {
name = "backup"
image = "docker.io/library/alpine"
command = ["/bin/sh", "-c", <<-EOT
set -euxo pipefail
apk add --no-cache sqlite
now=$(date +"%Y_%m_%d_%H_%M")
mkdir -p /backup
sqlite3 /data/db.sqlite ".backup /backup/db.sqlite.bak"
echo "Backup completed at $(date)"
EOT
]
volume_mount {
name = "data"
mount_path = "/data"
read_only = true
}
volume_mount {
name = "backup"
mount_path = "/backup"
}
}
volume {
name = "data"
persistent_volume_claim {
feat(storage): migrate all sensitive services to proxmox-lvm-encrypted Reconcile Terraform with cluster state after manual encrypted PVC migrations and complete the remaining unfinished migrations. All services storing sensitive data now use LUKS2-encrypted block storage via the Proxmox CSI plugin. ## Context Only Technitium DNS was using encrypted storage in Terraform. Many services had been manually migrated to encrypted PVCs in the cluster, but Terraform was never updated — creating dangerous state drift where a `tg apply` could recreate unencrypted PVCs. ## This change Phase 0 — Infrastructure: - Add `proxmox-lvm-encrypted` StorageClass to Helm values (extraParameters) - Add ExternalSecret for LUKS encryption passphrase to Terraform - Fix CSI node plugin memory: `node.plugin.resources` (not `node.resources`) with 1280Mi limit for LUKS2 Argon2id key derivation Phase 1 — TF state reconciliation (zero downtime): - Health, Matrix, N8N, Forgejo, Vaultwarden, Mailserver: state rm + import - Redis, DBAAS MySQL, DBAAS PostgreSQL: Helm/CNPG value updates Phase 2 — Data migration (encrypted PVCs existed but unused): - Headscale, Frigate, MeshCentral: rsync + switchover - Nextcloud (20Gi): rsync + chart_values update Phase 3 — New encrypted PVCs: - Roundcube HTML, HackMD, Affine, DBAAS pgadmin: create + rsync + switchover Phase 4 — Cleanup: - Deleted 5 orphaned unencrypted PVCs ## Services migrated (18 PVCs across 14 namespaces) ``` vaultwarden → vaultwarden-data-encrypted dbaas → datadir-mysql-cluster-0, pg-cluster-{1,2}, dbaas-pgadmin-encrypted mailserver → mailserver-data-encrypted, roundcubemail-{enigma,html}-encrypted nextcloud → nextcloud-data-encrypted forgejo → forgejo-data-encrypted matrix → matrix-data-encrypted n8n → n8n-data-encrypted affine → affine-data-encrypted health → health-uploads-encrypted hackmd → hackmd-data-encrypted redis → redis-data-redis-node-{0,1} headscale → headscale-data-encrypted frigate → frigate-config-encrypted meshcentral → meshcentral-{data,files}-encrypted ``` Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-15 20:15:30 +00:00
claim_name = kubernetes_persistent_volume_claim.data_encrypted.metadata[0].name
}
}
volume {
name = "backup"
persistent_volume_claim {
claim_name = module.nfs_data_host.claim_name
}
}
restart_policy = "OnFailure"
}
}
}
}
}
[infra] Sweep dns_config ignore_changes across all pod-owning resources [ci skip] ## Context Wave 3A (commit c9d221d5) added the `# KYVERNO_LIFECYCLE_V1` marker to the 27 pre-existing `ignore_changes = [...dns_config]` sites so they could be grepped and audited. It did NOT address pod-owning resources that were simply missing the suppression entirely. Post-Wave-3A sampling (2026-04-18) found that navidrome, f1-stream, frigate, servarr, monitoring, crowdsec, and many other stacks showed perpetual `dns_config` drift every plan because their `kubernetes_deployment` / `kubernetes_stateful_set` / `kubernetes_cron_job_v1` resources had no `lifecycle {}` block at all. Root cause (same as Wave 3A): Kyverno's admission webhook stamps `dns_config { option { name = "ndots"; value = "2" } }` on every pod's `spec.template.spec.dns_config` to prevent NxDomain search-domain flooding (see `k8s-ndots-search-domain-nxdomain-flood` skill). Without `ignore_changes` on every Terraform-managed pod-owner, Terraform repeatedly tries to strip the injected field. ## This change Extends the Wave 3A convention by sweeping EVERY `kubernetes_deployment`, `kubernetes_stateful_set`, `kubernetes_daemon_set`, `kubernetes_cron_job_v1`, `kubernetes_job_v1` (+ their `_v1` variants) in the repo and ensuring each carries the right `ignore_changes` path: - **kubernetes_deployment / stateful_set / daemon_set / job_v1**: `spec[0].template[0].spec[0].dns_config` - **kubernetes_cron_job_v1**: `spec[0].job_template[0].spec[0].template[0].spec[0].dns_config` (extra `job_template[0]` nesting — the CronJob's PodTemplateSpec is one level deeper) Each injection / extension is tagged `# KYVERNO_LIFECYCLE_V1: Kyverno admission webhook mutates dns_config with ndots=2` inline so the suppression is discoverable via `rg 'KYVERNO_LIFECYCLE_V1' stacks/`. Two insertion paths are handled by a Python pass (`/tmp/add_dns_config_ignore.py`): 1. **No existing `lifecycle {}`**: inject a brand-new block just before the resource's closing `}`. 108 new blocks on 93 files. 2. **Existing `lifecycle {}` (usually for `DRIFT_WORKAROUND: CI owns image tag` from Wave 4, commit a62b43d1)**: extend its `ignore_changes` list with the dns_config path. Handles both inline (`= [x]`) and multiline (`= [\n x,\n]`) forms; ensures the last pre-existing list item carries a trailing comma so the extended list is valid HCL. 34 extensions. The script skips anything already mentioning `dns_config` inside an `ignore_changes`, so re-running is a no-op. ## Scale - 142 total lifecycle injections/extensions - 93 `.tf` files touched - 108 brand-new `lifecycle {}` blocks + 34 extensions of existing ones - Every Tier 0 and Tier 1 stack with a pod-owning resource is covered - Together with Wave 3A's 27 pre-existing markers → **169 greppable `KYVERNO_LIFECYCLE_V1` dns_config sites across the repo** ## What is NOT in this change - `stacks/trading-bot/main.tf` — entirely commented-out block (`/* … */`). Python script touched the file, reverted manually. - `_template/main.tf.example` skeleton — kept minimal on purpose; any future stack created from it should either inherit the Wave 3A one-line form or add its own on first `kubernetes_deployment`. - `terraform fmt` fixes to pre-existing alignment issues in meshcentral, nvidia/modules/nvidia, vault — unrelated to this commit. Left for a separate fmt-only pass. - Non-pod resources (`kubernetes_service`, `kubernetes_secret`, `kubernetes_manifest`, etc.) — they don't own pods so they don't get Kyverno dns_config mutation. ## Verification Random sample post-commit: ``` $ cd stacks/navidrome && ../../scripts/tg plan → No changes. $ cd stacks/f1-stream && ../../scripts/tg plan → No changes. $ cd stacks/frigate && ../../scripts/tg plan → No changes. $ rg -c 'KYVERNO_LIFECYCLE_V1' stacks/ --include='*.tf' --include='*.tf.example' \ | awk -F: '{s+=$2} END {print s}' 169 ``` ## Reproduce locally 1. `git pull` 2. `rg 'KYVERNO_LIFECYCLE_V1' stacks/ | wc -l` → 169+ 3. `cd stacks/navidrome && ../../scripts/tg plan` → expect 0 drift on the deployment's dns_config field. Refs: code-seq (Wave 3B dns_config class closed; kubernetes_manifest annotation class handled separately in 8d94688d for tls_secret) Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-18 21:19:48 +00:00
lifecycle {
# KYVERNO_LIFECYCLE_V1: Kyverno admission webhook mutates dns_config with ndots=2
ignore_changes = [spec[0].job_template[0].spec[0].template[0].spec[0].dns_config]
}
}
# Grafana dashboard
resource "kubernetes_config_map" "grafana_headscale_dashboard" {
metadata {
name = "grafana-headscale-dashboard"
namespace = "monitoring"
labels = {
grafana_dashboard = "1"
}
annotations = {
grafana_folder = "Networking"
}
}
data = {
"headscale.json" = file("${path.module}/dashboards/headscale.json")
}
}