> **Shared knowledge**: Read `AGENTS.md` at repo root for architecture, patterns, rules, and operations. This file adds Claude-specific features on top.
**ALL infrastructure changes MUST go through Terraform/Terragrunt.** Never use `kubectl apply/edit/patch/set`, `helm install/upgrade`, or any manual cluster mutation as the final state.
- **No exceptions for "quick fixes"** — even one-line changes must be in `.tf` files and applied via `scripts/tg apply`
- **kubectl is for read-only operations and temporary debugging only** (get, describe, logs, exec, port-forward)
- **If a resource isn't in Terraform yet**, evaluate whether it can be added before making manual changes. If manual change is unavoidable (e.g., emergency), document it immediately and create the Terraform resource in the same session
- **kubectl scale/patch during migrations is acceptable** as a transient step, but the final state must be in Terraform and applied via `scripts/tg apply`
- **Helm values live in Terraform** (templatefile or inline) — never `helm upgrade` directly
Violations cause state drift, which causes future applies to break or silently revert changes.
- **"remember X"**: Use `memory-tool store "content" --category facts --tags "tag1,tag2"` (via exec) for persistent cross-session memory. Also update this file + `AGENTS.md` (if shared knowledge), commit with `[ci skip]`. To recall: `memory-tool recall "query"`. To list: `memory-tool list`. To delete: `memory-tool delete <id>`. The native `memory_search` and `memory_get` tools are also available for searching indexed memory files. For **storing** new memories, always use the `memory-tool` CLI via exec.
- **Apply**: Authenticate via `vault login -method=oidc`, then use `scripts/tg` (preferred — handles state decrypt/encrypt) or `terragrunt` directly. `scripts/tg` adds `-auto-approve` for `--non-interactive` applies.
- **Private registry**: `registry.viktorbarzin.me` (htpasswd auth, credentials in Vault `secret/viktor`). Use `image: registry.viktorbarzin.me/<name>:<tag>` + `imagePullSecrets: [{name: registry-credentials}]`. Kyverno auto-syncs the secret to all namespaces. Build & push from registry VM (`10.0.20.10`). Containerd `hosts.toml` redirects pulls to LAN IP directly. Web UI at `docker.viktorbarzin.me` (Authentik-protected).
- **LinuxServer.io containers**: `DOCKER_MODS` runs apt-get on every start — bake slow mods into a custom image (`RUN /docker-mods || true` then `ENV DOCKER_MODS=`). Set `NO_CHOWN=true` to skip recursive chown that hangs on NFS mounts.
- **Node memory changes**: When changing VM memory on any k8s node, update kubelet `systemReserved`, `kubeReserved`, and eviction thresholds accordingly. Config: `/var/lib/kubelet/config.yaml`. Template: `stacks/infra/main.tf`. Current values: systemReserved=512Mi, kubeReserved=512Mi, evictionHard=500Mi, evictionSoft=1Gi.
- **Sealed Secrets**: User-managed secrets go in `sealed-*.yaml` files in the stack directory. Stacks pick them up via `kubernetes_manifest` + `fileset(path.module, "sealed-*.yaml")`. See AGENTS.md for full workflow.
- **CRITICAL — Update docs with every change**: When modifying infrastructure (Terraform, Vault, networking, storage, CI/CD, monitoring), you MUST update all affected documentation in the same commit. Check and update: `docs/architecture/*.md`, `docs/runbooks/*.md`, `.claude/CLAUDE.md`, `AGENTS.md`, `.claude/reference/service-catalog.md`. Stale docs cause incident response failures and onboarding confusion. If unsure which docs are affected, grep for the service/resource name across all doc files.
- **State is local** (`backend "local"`), encrypted with SOPS and committed as `.tfstate.enc` files.
- **Decrypt priority**: Vault Transit (primary, uses existing `vault login` session) → age key fallback (`~/.config/sops/age/keys.txt`, for bootstrap/DR).
- **Encrypt**: Always encrypts to both Vault Transit (`transit/keys/sops-state`) + age recipients.
- **Scripts**: `scripts/state-sync {encrypt|decrypt|commit} [stack]` — handles all state sync. `scripts/tg` auto-decrypts before and auto-encrypts+commits after mutating ops (apply/destroy/import).
- **Workflow**: `git pull` → `scripts/tg plan` → `scripts/tg apply` → `git push`. State sync is transparent.
- **Config**: `.sops.yaml` at repo root defines encryption rules. age public keys listed there.
- **Adding operator**: Generate age key (`age-keygen`), add pubkey to `.sops.yaml`, run `sops updatekeys` on all `.enc` files.
- **Two workstations**: Laptop (macOS) + DevVM (10.0.10.10, Linux). Both have age keys + Vault access. Keys backed up in Vault (`secret/viktor/sops_age_key_laptop`, `sops_age_key_devvm`).
## Secrets Management — Vault KV
- **Vault is the sole source of truth** for secrets.
- **`secret/viktor`** — go-to path for ALL personal secrets (135 keys). Contains every API key, token, password, SSH key, and config from the old terraform.tfvars. Check here first: `vault kv get -field=KEY secret/viktor`.
- **ESO (External Secrets Operator)**: `stacks/external-secrets/` — 43 ExternalSecrets + 9 DB-creds ExternalSecrets. API version `v1beta1`. Two ClusterSecretStores: `vault-kv` and `vault-database`.
- **Plan-time pattern**: Former plan-time stacks use `data "kubernetes_secret"` to read ESO-created K8s Secrets at plan time (no Vault dependency). First-apply gotcha: must `terragrunt apply -target=kubernetes_manifest.external_secret` first, then full apply. `count` on resources using secret values fails — remove conditional counts.
- **14 hybrid stacks** still keep `data "vault_kv_secret_v2"` for plan-time needs (job commands, Helm templatefile, module inputs). Platform has 48 plan-time refs — no migration possible without restructuring modules.
- **CI/CD (GHA + Woodpecker)**: Docker builds run on **GitHub Actions** (free on public repos). Woodpecker is **deploy-only** — receives image tag via API POST, runs `kubectl set image`. Woodpecker authenticates via K8s SA JWT → Vault K8s auth. Sync CronJob pushes `secret/ci/global` → Woodpecker API every 6h. Shell scripts in HCL heredocs: escape `$` → `$$`, `%{}` → `%%{}`.
- **Platform cannot depend on vault** (circular). Apply order: vault first, then platform. Platform has 48 vault refs, all in module inputs — no ESO migration possible.
- **Complex types** (maps/lists like `homepage_credentials`, `k8s_users`) stored as JSON strings in KV, decoded with `jsondecode()` in consuming stack `locals` blocks.
- **New stacks**: Add secret in Vault UI/CLI at `secret/<stack-name>`, add ExternalSecret + `data "kubernetes_secret"` for plan-time, `secret_key_ref` for env vars. Use `data "vault_kv_secret_v2"` only if `data "kubernetes_secret"` won't work (e.g., first-apply bootstrap).
- **LimitRange**: Tier-based defaults silently apply to pods with `resources: {}`. Always set explicit resources on containers needing more than defaults. Tier 3-edge and 4-aux now use Burstable QoS (request <limit)toreduceschedulerpressure.
- **Democratic-CSI sidecars**: Must set explicit resources (32-80Mi) in Helm values — 17 sidecars default to 256Mi each via LimitRange. `csiProxy` is a TOP-LEVEL chart key, not nested under controller/node.
- **ResourceQuota blocks rolling updates**: When quota is tight, scale to 0 then back to 1 instead of RollingUpdate. Or use Recreate strategy.
- **Kyverno ndots drift**: Kyverno injects dns_config on all pods. Add `lifecycle { ignore_changes = [spec[0].template[0].spec[0].dns_config] }` to kubernetes_deployment resources to prevent perpetual TF plan drift.
- **NVIDIA GPU operator resources**: dcgm-exporter and cuda-validator resources configurable via `dcgmExporter.resources` and `validator.resources` in nvidia values.yaml.
- **Quarterly right-sizing**: Check Goldilocks dashboard. Compare VPA upperBound to current request. Also check for under-provisioned (VPA upper > request x 0.8).
**`postgresql_host`** in `config.tfvars` is `pg-cluster-rw.dbaas.svc.cluster.local` (the CNPG primary). The legacy `postgresql.dbaas` service has no endpoints — never use it. This variable is shared by ~12 stacks.
| CrowdSec | Pin version, disable Metabase when not needed (CPU hog), LAPI scaled to 3, **DB on PostgreSQL** (migrated from MySQL), flush config: max_items=10000/max_age=7d/agents_autodelete=30d, DECISION_DURATION=168h in blocklist CronJob |
- Every new service gets Prometheus scrape config + Uptime Kuma monitor. External monitors auto-created for Cloudflare-proxied services by `external-monitor-sync` CronJob (10min, uptime-kuma ns).
- **External monitoring**: `[External] <service>` monitors in Uptime Kuma test full external path (DNS → Cloudflare → Tunnel → Traefik). Divergence metric `external_internal_divergence_count` → alert `ExternalAccessDivergence` (15min). Config: `stacks/uptime-kuma/`, targets from `cloudflare_proxied_names` in `config.tfvars`.
**Migration note**: CSI PV `volumeAttributes` are immutable — cannot update NFS server in place. New PV/PVC pairs required (convention: append `-host` to PV name).
**NFS CSI mount option requirements** (learned from [PM-2026-04-14]):
- **ALWAYS set `nfsvers=4`** in CSI mount options. NFSv3 is disabled on the PVE host (`vers3=n` in `/etc/nfs.conf`). Without this, mounts fail silently if kernel NFS client state is corrupt.
- **NEVER use `fsid=0`** in `/etc/exports` on `/srv/nfs`. `fsid=0` designates the NFSv4 pseudo-root, which breaks subdirectory path resolution for all CSI mounts. Only `fsid=1` (unique ID) is safe on `/srv/nfs-ssd`.
- **`/etc/exports` is git-managed** at `infra/scripts/pve-nfs-exports`. Deploy: `scp scripts/pve-nfs-exports root@192.168.1.127:/etc/exports && ssh root@192.168.1.127 exportfs -ra`
- **Critical services MUST NOT use NFS storage** — circular dependency risk. Alertmanager, Prometheus, and any monitoring that should alert about NFS must use `proxmox-lvm-encrypted`. Technitium DNS primary uses `proxmox-lvm-encrypted` (migrated 2026-04-14).
- **NFS PV template** (in `modules/kubernetes/nfs_volume/`): always include `mountOptions: ["nfsvers=4", "soft", "actimeo=5", "retrans=3", "timeo=30"]`
-`nfs-change-tracker.service` — Continuous inotifywait on `/srv/nfs` + `/srv/nfs-ssd`. Logs changed file paths to `/mnt/backup/.nfs-changes.log`. Consumed by offsite-sync-backup for incremental rsync (completes in seconds instead of 30+ minutes).
- **OpenClaw config is writable**: OpenClaw writes to `openclaw.json` at runtime (doctor --fix, plugin auto-enable). Never use subPath ConfigMap mounts for it — use an init container to copy into a writable volume. Needs 2Gi memory + `NODE_OPTIONS=--max-old-space-size=1536`.
- **Goldilocks VPA sets limits**: When increasing memory requests, always set explicit `limits` too — Goldilocks may have added a limit that blocks the change.
