2026-03-07 14:30:36 +00:00
|
|
|
variable "tls_secret_name" {
|
2026-03-14 08:51:45 +00:00
|
|
|
type = string
|
2026-03-07 14:30:36 +00:00
|
|
|
sensitive = true
|
|
|
|
|
}
|
2026-02-22 13:56:34 +00:00
|
|
|
variable "public_ip" { type = string }
|
|
|
|
|
|
migrate 16 plan-time stacks: vault data source → ESO + kubernetes_secret
Replaced data "vault_kv_secret_v2" with:
1. ExternalSecret (ESO syncs Vault KV → K8s Secret)
2. data "kubernetes_secret" (reads ESO-created secret at plan time)
This removes the Vault provider dependency at plan time for these
stacks — they now only need K8s API access, not a Vault token.
Stacks: actualbudget, affine, audiobookshelf, calibre, changedetection,
coturn, freedify, freshrss, grampsweb, navidrome, novelapp, ollama,
owntracks, real-estate-crawler, servarr, ytdlp
2026-03-15 22:06:39 +00:00
|
|
|
resource "kubernetes_manifest" "external_secret" {
|
|
|
|
|
manifest = {
|
|
|
|
|
apiVersion = "external-secrets.io/v1beta1"
|
|
|
|
|
kind = "ExternalSecret"
|
|
|
|
|
metadata = {
|
|
|
|
|
name = "coturn-secrets"
|
|
|
|
|
namespace = "coturn"
|
|
|
|
|
}
|
|
|
|
|
spec = {
|
|
|
|
|
refreshInterval = "15m"
|
|
|
|
|
secretStoreRef = {
|
|
|
|
|
name = "vault-kv"
|
|
|
|
|
kind = "ClusterSecretStore"
|
|
|
|
|
}
|
|
|
|
|
target = {
|
|
|
|
|
name = "coturn-secrets"
|
|
|
|
|
}
|
|
|
|
|
dataFrom = [{
|
|
|
|
|
extract = {
|
|
|
|
|
key = "coturn"
|
|
|
|
|
}
|
|
|
|
|
}]
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
depends_on = [kubernetes_namespace.coturn]
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
data "kubernetes_secret" "eso_secrets" {
|
|
|
|
|
metadata {
|
|
|
|
|
name = "coturn-secrets"
|
|
|
|
|
namespace = kubernetes_namespace.coturn.metadata[0].name
|
|
|
|
|
}
|
|
|
|
|
depends_on = [kubernetes_manifest.external_secret]
|
2026-03-14 17:15:48 +00:00
|
|
|
}
|
2026-02-22 13:56:34 +00:00
|
|
|
|
2026-02-22 15:13:55 +00:00
|
|
|
locals {
|
|
|
|
|
turn_realm = "viktorbarzin.me"
|
|
|
|
|
turn_port = 3478
|
|
|
|
|
# Small relay range — 100 ports is plenty for a home lab (~50 concurrent streams)
|
|
|
|
|
min_port = 49152
|
|
|
|
|
max_port = 49252
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
resource "kubernetes_namespace" "coturn" {
|
|
|
|
|
metadata {
|
|
|
|
|
name = "coturn"
|
|
|
|
|
labels = {
|
|
|
|
|
tier = local.tiers.edge
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
module "tls_secret" {
|
|
|
|
|
source = "../../modules/kubernetes/setup_tls_secret"
|
|
|
|
|
namespace = kubernetes_namespace.coturn.metadata[0].name
|
2026-02-22 14:38:14 +00:00
|
|
|
tls_secret_name = var.tls_secret_name
|
2026-02-22 15:13:55 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
resource "kubernetes_config_map" "coturn_config" {
|
|
|
|
|
metadata {
|
|
|
|
|
name = "coturn-config"
|
|
|
|
|
namespace = kubernetes_namespace.coturn.metadata[0].name
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
data = {
|
|
|
|
|
"turnserver.conf" = <<-EOF
|
|
|
|
|
# TURN server configuration
|
|
|
|
|
listening-port=${local.turn_port}
|
|
|
|
|
fingerprint
|
|
|
|
|
lt-cred-mech
|
|
|
|
|
use-auth-secret
|
migrate 16 plan-time stacks: vault data source → ESO + kubernetes_secret
Replaced data "vault_kv_secret_v2" with:
1. ExternalSecret (ESO syncs Vault KV → K8s Secret)
2. data "kubernetes_secret" (reads ESO-created secret at plan time)
This removes the Vault provider dependency at plan time for these
stacks — they now only need K8s API access, not a Vault token.
Stacks: actualbudget, affine, audiobookshelf, calibre, changedetection,
coturn, freedify, freshrss, grampsweb, navidrome, novelapp, ollama,
owntracks, real-estate-crawler, servarr, ytdlp
2026-03-15 22:06:39 +00:00
|
|
|
static-auth-secret=${data.kubernetes_secret.eso_secrets.data["turn_secret"]}
|
2026-02-22 15:13:55 +00:00
|
|
|
realm=${local.turn_realm}
|
|
|
|
|
server-name=turn.${local.turn_realm}
|
|
|
|
|
|
|
|
|
|
# Network — use 0.0.0.0, coturn auto-detects pod IP
|
|
|
|
|
listening-ip=0.0.0.0
|
|
|
|
|
external-ip=${var.public_ip}
|
|
|
|
|
|
|
|
|
|
# Media relay port range (narrow — 100 ports)
|
|
|
|
|
min-port=${local.min_port}
|
|
|
|
|
max-port=${local.max_port}
|
|
|
|
|
|
|
|
|
|
# Logging
|
|
|
|
|
verbose
|
|
|
|
|
no-stdout-log
|
|
|
|
|
syslog
|
|
|
|
|
|
|
|
|
|
# Security
|
|
|
|
|
no-multicast-peers
|
|
|
|
|
no-cli
|
|
|
|
|
no-tlsv1
|
|
|
|
|
no-tlsv1_1
|
|
|
|
|
|
|
|
|
|
# Performance
|
|
|
|
|
total-quota=100
|
|
|
|
|
stale-nonce=600
|
|
|
|
|
max-bps=0
|
|
|
|
|
EOF
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
resource "kubernetes_deployment" "coturn" {
|
|
|
|
|
metadata {
|
|
|
|
|
name = "coturn"
|
|
|
|
|
namespace = kubernetes_namespace.coturn.metadata[0].name
|
|
|
|
|
labels = {
|
|
|
|
|
app = "coturn"
|
|
|
|
|
tier = local.tiers.edge
|
|
|
|
|
}
|
migrate 16 plan-time stacks: vault data source → ESO + kubernetes_secret
Replaced data "vault_kv_secret_v2" with:
1. ExternalSecret (ESO syncs Vault KV → K8s Secret)
2. data "kubernetes_secret" (reads ESO-created secret at plan time)
This removes the Vault provider dependency at plan time for these
stacks — they now only need K8s API access, not a Vault token.
Stacks: actualbudget, affine, audiobookshelf, calibre, changedetection,
coturn, freedify, freshrss, grampsweb, navidrome, novelapp, ollama,
owntracks, real-estate-crawler, servarr, ytdlp
2026-03-15 22:06:39 +00:00
|
|
|
annotations = {
|
|
|
|
|
"reloader.stakater.com/auto" = "true"
|
|
|
|
|
}
|
2026-02-22 15:13:55 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
spec {
|
|
|
|
|
replicas = 1
|
|
|
|
|
strategy {
|
|
|
|
|
type = "RollingUpdate"
|
|
|
|
|
rolling_update {
|
|
|
|
|
max_unavailable = 0
|
|
|
|
|
max_surge = 1
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
selector {
|
|
|
|
|
match_labels = {
|
|
|
|
|
app = "coturn"
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
template {
|
|
|
|
|
metadata {
|
|
|
|
|
labels = {
|
|
|
|
|
app = "coturn"
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
spec {
|
|
|
|
|
container {
|
|
|
|
|
name = "coturn"
|
|
|
|
|
image = "coturn/coturn:latest"
|
|
|
|
|
args = ["-c", "/etc/turnserver/turnserver.conf"]
|
|
|
|
|
|
|
|
|
|
# STUN/TURN signaling port
|
|
|
|
|
port {
|
|
|
|
|
name = "turn-udp"
|
|
|
|
|
container_port = local.turn_port
|
|
|
|
|
protocol = "UDP"
|
|
|
|
|
}
|
|
|
|
|
port {
|
|
|
|
|
name = "turn-tcp"
|
|
|
|
|
container_port = local.turn_port
|
|
|
|
|
protocol = "TCP"
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
volume_mount {
|
|
|
|
|
name = "config"
|
|
|
|
|
mount_path = "/etc/turnserver"
|
|
|
|
|
read_only = true
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
resources {
|
|
|
|
|
requests = {
|
[ci skip] right-size all pod resources based on VPA + live metrics audit
Full cluster resource audit: cross-referenced Goldilocks VPA recommendations,
live kubectl top metrics, and Terraform definitions for 100+ containers.
Critical fixes:
- dashy: CPU throttled at 98% (490m/500m) → 2 CPU limit
- stirling-pdf: CPU throttled at 99.7% (299m/300m) → 2 CPU limit
- traefik auth-proxy/bot-block-proxy: mem limit 32Mi → 128Mi
Added explicit resources to ~40 containers that had none:
- audiobookshelf, changedetection, cyberchef, dawarich, diun, echo,
excalidraw, freshrss, hackmd, isponsorblocktv, linkwarden, n8n,
navidrome, ntfy, owntracks, privatebin, send, shadowsocks, tandoor,
tor-proxy, wealthfolio, networking-toolbox, rybbit, mailserver,
cloudflared, pgadmin, phpmyadmin, crowdsec-web, xray, wireguard,
k8s-portal, tuya-bridge, ollama-ui, whisper, piper, immich-server,
immich-postgresql, osrm-foot
GPU containers: added CPU/mem alongside GPU limits:
- ollama: removed CPU/mem limits (models vary in size), keep GPU only
- frigate: req 500m/2Gi, lim 4/8Gi + GPU
- immich-ml: req 100m/1Gi, lim 2/4Gi + GPU
Right-sized ~25 over-provisioned containers:
- kms-web-page: 500m/512Mi → 50m/64Mi (was using 0m/10Mi)
- onlyoffice: CPU 8 → 2 (VPA upper 45m)
- realestate-crawler-api: CPU 2000m → 250m
- blog/travel-blog/webhook-handler: 500m → 100m
- coturn/health/plotting-book: reduced to match actual usage
Conservative methodology: limits = max(VPA upper * 2, live usage * 2)
2026-03-01 19:18:50 +00:00
|
|
|
cpu = "10m"
|
right-size memory: set requests=limits based on actual usage
- Set memory requests = limits across 56 stacks to prevent overcommit
- Right-sized limits based on actual pod usage (2x actual, rounded up)
- Scaled down trading-bot (replicas=0) to free memory
- Fixed OOMKilled services: forgejo, dawarich, health, meshcentral,
paperless-ngx, vault auto-unseal, rybbit, whisper, openclaw, clickhouse
- Added startup+liveness probes to calibre-web
- Bumped inotify limits on nodes 2,3 (max_user_instances 128->8192)
Post node2 OOM incident (2026-03-14). Previous kubelet config had no
kubeReserved/systemReserved set, allowing pods to starve the kernel.
2026-03-14 21:01:24 +00:00
|
|
|
memory = "64Mi"
|
2026-02-22 15:13:55 +00:00
|
|
|
}
|
|
|
|
|
limits = {
|
right-size memory: set requests=limits based on actual usage
- Set memory requests = limits across 56 stacks to prevent overcommit
- Right-sized limits based on actual pod usage (2x actual, rounded up)
- Scaled down trading-bot (replicas=0) to free memory
- Fixed OOMKilled services: forgejo, dawarich, health, meshcentral,
paperless-ngx, vault auto-unseal, rybbit, whisper, openclaw, clickhouse
- Added startup+liveness probes to calibre-web
- Bumped inotify limits on nodes 2,3 (max_user_instances 128->8192)
Post node2 OOM incident (2026-03-14). Previous kubelet config had no
kubeReserved/systemReserved set, allowing pods to starve the kernel.
2026-03-14 21:01:24 +00:00
|
|
|
memory = "64Mi"
|
2026-02-22 15:13:55 +00:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
volume {
|
|
|
|
|
name = "config"
|
|
|
|
|
config_map {
|
|
|
|
|
name = kubernetes_config_map.coturn_config.metadata[0].name
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
# LoadBalancer service with MetalLB — exposes STUN/TURN signaling + relay ports
|
|
|
|
|
resource "kubernetes_service" "coturn" {
|
|
|
|
|
metadata {
|
|
|
|
|
name = "coturn"
|
|
|
|
|
namespace = kubernetes_namespace.coturn.metadata[0].name
|
|
|
|
|
annotations = {
|
|
|
|
|
"metallb.universe.tf/loadBalancerIPs" = "10.0.20.200"
|
|
|
|
|
"metallb.universe.tf/allow-shared-ip" = "shared"
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
spec {
|
|
|
|
|
type = "LoadBalancer"
|
|
|
|
|
selector = {
|
|
|
|
|
app = "coturn"
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
# STUN/TURN signaling
|
|
|
|
|
port {
|
|
|
|
|
name = "turn-udp"
|
|
|
|
|
port = local.turn_port
|
|
|
|
|
target_port = local.turn_port
|
|
|
|
|
protocol = "UDP"
|
|
|
|
|
}
|
|
|
|
|
port {
|
|
|
|
|
name = "turn-tcp"
|
|
|
|
|
port = local.turn_port
|
|
|
|
|
target_port = local.turn_port
|
|
|
|
|
protocol = "TCP"
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
# Relay port range (49152-49252)
|
|
|
|
|
dynamic "port" {
|
|
|
|
|
for_each = range(local.min_port, local.max_port + 1)
|
|
|
|
|
content {
|
|
|
|
|
name = "relay-${port.value}"
|
|
|
|
|
port = port.value
|
|
|
|
|
target_port = port.value
|
|
|
|
|
protocol = "UDP"
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
2026-02-22 13:56:34 +00:00
|
|
|
}
|