mirror of
https://github.com/Alkatrazz24/Funk-lab.git
synced 2026-07-08 19:14:43 +02:00
feat(postfix): relay SMTP Gmail + suppression règles Conan nftables
- postfix_relay: switch Brevo → Gmail SMTP (smtp.gmail.com:587, App Password) - smtp_generic_maps: réécriture expéditeur root@lab.local → aliyesilkaya93@gmail.com - nftables: suppression DNAT/NAT Conan Exiles (ports 7777/7778/27015) - gateway handler: reloaded → restarted (flush complet des règles nft) - hermes_agent: pip → uv pour installation qdrant-client (venv sans pip binaire) - CLAUDE.md: corrections funk-cluster, hermes-switch, RAID5, postfix - docs: admin/infra/email.md — procédure complète relay Gmail Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
parent
24c0bf9e12
commit
8da92cb721
13 changed files with 552 additions and 145 deletions
290
CLAUDE.md
290
CLAUDE.md
|
|
@ -6,15 +6,19 @@ This file provides guidance to Claude Code (claude.ai/code) when working with co
|
|||
|
||||
**Funk** — homelab IA composé d'un cluster Kubernetes (Talos) pour les services applicatifs, d'un hôte d'inférence LLM dédié (gpu-01 / llama-server ROCm), et d'un agent autonome (Hermes Agent) sur storage-01. Stack IaC complète : Ansible + talhelper + ArgoCD.
|
||||
|
||||
Référence complète : `Funk/Funk-infra/HOMELAB.md`
|
||||
|
||||
**État actuel (2026-05-13)** :
|
||||
- ✅ Hermes Agent opérationnel sur storage-01 (llama-server GPU via LiteLLM)
|
||||
**État actuel (2026-05-17)** :
|
||||
- ✅ Hermes Agent opérationnel — multi-profils (funk-ai/system/monitor/brain)
|
||||
- ✅ RAG opérationnel — Qdrant 284 chunks, `rag-ingest` / `rag-query`, skill funk-ai
|
||||
- ✅ Monitoring complet — Prometheus + Grafana + AlertManager → webhook Hermes
|
||||
- ✅ storage-01 installé — `192.168.10.1` (LAN cluster), `192.168.1.200` (WAN)
|
||||
- ✅ gpu-01 installé — llama-server GPU (:1234) + CPU (:1236/:1237) opérationnels
|
||||
- ✅ LiteLLM sur storage-01 — proxy multi-modèles + ask-agent
|
||||
- ✅ Cluster k8s Talos opérationnel (MetalLB + Traefik + wildcard DNS lab.local)
|
||||
- ✅ postfix_relay — SMTP sortant via Brevo (aliyesilkaya93@gmail.com)
|
||||
- ⏳ WOL compute nodes non configuré (BIOS) — allumage manuel requis
|
||||
- ⏳ ArgoCD bootstrap + workloads IA (Open WebUI, n8n)
|
||||
|
||||
Journal de progression : `PROGRESS.md` → `progress/YYYY-MM-DD.md`
|
||||
|
||||
---
|
||||
|
||||
|
|
@ -22,42 +26,16 @@ Référence complète : `Funk/Funk-infra/HOMELAB.md`
|
|||
|
||||
| Machine | IP | OS | Rôle |
|
||||
|---|---|---|---|
|
||||
| storage-01 | 192.168.10.1 | AlmaLinux 9.7 | Passerelle + bastion admin + données + Hermes Agent — **hors cluster** |
|
||||
| compute-01 | 192.168.10.11 | Talos v1.13+ | k8s control-plane |
|
||||
| compute-02 | 192.168.10.12 | Talos v1.13+ | k8s worker |
|
||||
| compute-03 | 192.168.10.13 | Talos v1.13+ | k8s worker |
|
||||
| storage-01 | 192.168.10.1 / 192.168.1.200 | AlmaLinux 9.7 | Passerelle + bastion admin + données + Hermes Agent — **hors cluster** |
|
||||
| compute-01 | 192.168.10.11 | Talos v1.13 | k8s control-plane |
|
||||
| compute-02 | 192.168.10.12 | Talos v1.13 | k8s worker |
|
||||
| compute-03 | 192.168.10.13 | Talos v1.13 | k8s worker |
|
||||
| gpu-01 | 192.168.10.20 | AlmaLinux 9.7 | llama-server dédié (RX 6700XT, ROCm 7.x) — **hors cluster** |
|
||||
|
||||
**Points clés d'architecture :**
|
||||
- `storage-01` et `gpu-01` sont **hors du cluster k8s** — ils sont consommés via des `ExternalName` services ou `Endpoints` manuels
|
||||
- Le cluster n'a que 3 nœuds (compute-01/02/03), pas de GPU dans k8s
|
||||
- `storage-01` et `gpu-01` sont **hors du cluster k8s** — consommés via `ExternalName` services ou `Endpoints` manuels
|
||||
- Single control-plane assumé : si compute-01 tombe, `kubectl apply` ne fonctionne plus mais les workloads continuent
|
||||
|
||||
---
|
||||
|
||||
## Structure du repo lab-infra (cible)
|
||||
|
||||
```
|
||||
lab-infra/
|
||||
├── ansible/ # gère storage-01 + gpu-01 (AlmaLinux)
|
||||
│ ├── inventory.yml
|
||||
│ ├── playbooks/site.yml
|
||||
│ ├── group_vars/
|
||||
│ └── roles/ # common, gateway, dnsmasq, nfs_server, postgresql,
|
||||
│ # qdrant, minio, rocm, llama_server, litellm, hermes_agent, hermes_user
|
||||
├── talos/ # nœuds Talos via talhelper
|
||||
│ ├── talconfig.yaml # source de vérité (lue par talhelper)
|
||||
│ ├── talsecret.sops.yaml
|
||||
│ ├── talenv.yaml
|
||||
│ ├── patches/
|
||||
│ └── clusterconfig/ # généré → au .gitignore, ne pas committer
|
||||
└── k8s/ # workloads k8s (ArgoCD)
|
||||
├── argocd-bootstrap/
|
||||
├── apps-of-apps/root.yaml
|
||||
├── infra/ # metallb, traefik, cert-manager, longhorn, nfs-provisioner
|
||||
└── apps/ # litellm, open-webui, n8n, monitoring,
|
||||
# external-services (ExternalName vers storage-01/gpu-01)
|
||||
```
|
||||
- Modèles LLM stockés sur NFS RAID5 de storage-01 (`/srv/data/models`), montés sur gpu-01 (`/mnt/models`)
|
||||
|
||||
---
|
||||
|
||||
|
|
@ -65,22 +43,67 @@ lab-infra/
|
|||
|
||||
| Outil | Cible | Déclenchement |
|
||||
|---|---|---|
|
||||
| **Ansible** | storage-01, gpu-01 (AlmaLinux) | Push manuel : `ansible-playbook` |
|
||||
| **Ansible** | storage-01, gpu-01 (AlmaLinux) | Push manuel : `make apply-storage` / `make apply-gpu` |
|
||||
| **talhelper** | compute-01/02/03 (Talos) | Push manuel : `talosctl apply-config` |
|
||||
| **ArgoCD** | Workloads k8s | Pull automatique depuis Git |
|
||||
|
||||
### Ansible
|
||||
---
|
||||
|
||||
## Ansible
|
||||
|
||||
### Commandes via Makefile (depuis la racine du repo)
|
||||
|
||||
```bash
|
||||
cd ansible/
|
||||
ansible-playbook -i inventory.yml playbooks/site.yml --check # dry-run
|
||||
ansible-playbook -i inventory.yml playbooks/site.yml # apply
|
||||
ansible-playbook -i inventory.yml playbooks/site.yml --tags rocm # un rôle
|
||||
make install # pip install + ansible-galaxy (à faire une fois)
|
||||
make ping # vérifier la connectivité vers tous les hôtes
|
||||
make check # dry-run site.yml (storage-01 + gpu-01)
|
||||
make apply # appliquer site.yml complet
|
||||
make apply-storage # appliquer uniquement storage-01.yml
|
||||
make apply-gpu # appliquer uniquement gpu-01.yml
|
||||
make vault-view # lire le vault chiffré
|
||||
make vault-edit # éditer le vault (ouvre $EDITOR)
|
||||
make vault-encrypt # (ré)chiffrer vault.yml
|
||||
```
|
||||
|
||||
Rôles : `common`, `gateway`, `dnsmasq`, `nfs_server`, `nfs_client`, `postgresql`, `qdrant`, `minio`, `rocm`, `llama_server`, `litellm`, `hermes_agent`, `hermes_user`. Tous idempotents. Secrets via Ansible Vault.
|
||||
### Commandes directes (depuis `ansible/`)
|
||||
|
||||
### talhelper (nœuds Talos)
|
||||
```bash
|
||||
ansible-playbook -i inventory.yml playbooks/storage-01.yml --tags litellm
|
||||
ansible-playbook -i inventory.yml playbooks/gpu-01.yml --tags llama_server
|
||||
```
|
||||
|
||||
Les tags correspondent aux noms de rôles. Tous les rôles sont idempotents.
|
||||
|
||||
### Inventaire et connectivité
|
||||
|
||||
- `storage-01` : accès direct SSH via `192.168.1.200` (WAN)
|
||||
- `gpu-01` : accès via ProxyJump sur `storage-01` (configuré dans `inventory.yml`)
|
||||
- Vault password dans `.vault_pass` (ignoré par git, chargé automatiquement via `ansible.cfg`)
|
||||
|
||||
### Rôles — storage-01
|
||||
|
||||
`hermes_user`, `common`, `gateway` (nftables + NAT), `dnsmasq`, `nfs_server`, `qdrant`, `postgresql`, `minio`, `litellm`, `hermes_agent`, `rag`, `node_exporter`, `alertmanager_webhook`, `postfix_relay`
|
||||
|
||||
### Rôles — gpu-01
|
||||
|
||||
`hermes_user`, `common`, `rocm`, `nfs_client`, `llama_server`, `node_exporter`
|
||||
|
||||
> `lm_studio` est présent dans `ansible/roles/` mais n'est plus utilisé dans aucun playbook.
|
||||
|
||||
### Caveat premier déploiement — hermes_agent
|
||||
|
||||
Le rôle `hermes_agent` installe le binaire via `curl | bash` seulement si le binaire est absent. La configuration (`.env`, `config.yaml`) est ensuite conditionnelle à `hermes_binary.stat.exists`. Sur un hôte vierge, il faut **deux runs consécutifs** : le premier installe, le second configure.
|
||||
|
||||
### Secrets Ansible Vault
|
||||
|
||||
Variables chiffrées dans `ansible/group_vars/all/vault.yml` :
|
||||
- `vault_anthropic_api_key` — clé Anthropic pour le profil brain
|
||||
- `vault_pg_hermes_password`, `vault_pg_litellm_password`
|
||||
- `vault_postfix_relay_password` — SMTP Brevo
|
||||
|
||||
---
|
||||
|
||||
## talhelper (nœuds Talos)
|
||||
|
||||
```bash
|
||||
cd talos/
|
||||
|
|
@ -90,102 +113,151 @@ talhelper gencommand bootstrap --node compute-01 | bash # bootstrap (1ère fo
|
|||
talhelper gencommand kubeconfig | bash # récupérer kubeconfig
|
||||
```
|
||||
|
||||
`talos/clusterconfig/` est généré — ne pas committer (mettre au `.gitignore`).
|
||||
`talos/clusterconfig/` est généré — gitignored, ne pas committer.
|
||||
`talos/talsecret.sops.yaml` chiffré avec SOPS+age (clé dans `.config/sops/`, ignorée par git).
|
||||
|
||||
### ArgoCD (workloads k8s)
|
||||
---
|
||||
|
||||
Pattern "App of Apps" : modifier un manifest → `git commit` + `git push` → ArgoCD applique automatiquement (poll toutes les ~3 min).
|
||||
## ArgoCD (workloads k8s)
|
||||
|
||||
Pattern "App of Apps" : modifier un manifest → `git commit` + `git push` → ArgoCD poll automatiquement (~3 min).
|
||||
|
||||
```bash
|
||||
vim k8s/apps/open-webui/deployment.yaml
|
||||
git add k8s/apps/open-webui/
|
||||
git commit -m "open-webui: bump to v0.4.0"
|
||||
git push
|
||||
```
|
||||
|
||||
ArgoCD utilise le repo `git@github.com:Alkatrazz24/Funk-lab.git` branche `main`. Les deux namespaces gérés actuellement : `infra` (Traefik, MetalLB, NFS-provisioner) et `monitoring` (kube-prometheus-stack 85.0.2).
|
||||
|
||||
---
|
||||
|
||||
## Réseaux internes Kubernetes
|
||||
## Stack IA — Hermes Agent
|
||||
|
||||
- CNI (Flannel) — Pods : `10.42.0.0/16`
|
||||
### Profils
|
||||
|
||||
| Profil | Modèle | Port | Usage |
|
||||
|---|---|---|---|
|
||||
| `funk-ai` | Qwen3-8B Q4_K_M | GPU :1234 | Agent principal (~35 tok/s) |
|
||||
| `system` | Qwen3-1.7B Q4_K_M | CPU :1236 | Calculs, formatage, résumés |
|
||||
| `monitor` | Qwen3-1.7B Q4_K_M | CPU :1237 | Analyse logs, métriques, état services |
|
||||
| `brain` | Claude Sonnet 4.6 API | Anthropic | Raisonnement complexe — facturé |
|
||||
|
||||
Hermes écoute sur le port `8080` (gateway). Dashboard web sur le port `9119` (accessible uniquement depuis `192.168.1.10`).
|
||||
|
||||
### Switcher le modèle de Hermes à chaud
|
||||
|
||||
```bash
|
||||
hermes-switch status # modèle actuel de hermes-default
|
||||
hermes-switch qwen # → Qwen3-8B sur gpu-01 (local, gratuit)
|
||||
hermes-switch claude # → Claude Sonnet 4.6 via API Anthropic (facturé)
|
||||
```
|
||||
|
||||
`hermes-switch` modifie `/etc/litellm/config.yaml` puis redémarre `litellm`. L'alias `hermes-default` dans LiteLLM est ce que Hermes consomme — changer cet alias ne nécessite pas de reconfigurer Hermes.
|
||||
|
||||
### ask-agent (délégation inter-agents)
|
||||
|
||||
Binaire dans `/usr/local/bin/ask-agent` sur storage-01. Appel LiteLLM direct (:4000), réponse en ~2-5s.
|
||||
|
||||
**Règle** : ask-agent reçoit du **texte**, jamais des commandes shell. Toujours 2 étapes :
|
||||
```bash
|
||||
# Étape 1 — collecter les données
|
||||
journalctl -u litellm -n 20 --no-pager --output=cat
|
||||
# Étape 2 — passer le texte à ask-agent
|
||||
ask-agent monitor "logs litellm : <résultat étape 1> — erreurs ?"
|
||||
```
|
||||
|
||||
### RAG — Documentation Funk
|
||||
|
||||
Qdrant collection `funk-docs` indexée depuis `admin/` (284 chunks, score min 0.60).
|
||||
|
||||
```bash
|
||||
rag-query "comment relancer dnsmasq ?"
|
||||
rag-query "nftables port cluster configuration"
|
||||
rag-query "hermes profils litellm modèles" --top 3
|
||||
```
|
||||
|
||||
Scripts installés sur storage-01 : `/usr/local/bin/rag-query`, `/usr/local/bin/rag-ingest`
|
||||
|
||||
Après modification de `admin/`, re-déployer via `make apply-storage --tags rag` (exécute `rag-ingest`). Le script `admin/index_knowledge.py` est l'ingesteur Python brut.
|
||||
|
||||
### Hermes Skills — `hermes-skills/`
|
||||
|
||||
Versionnés dans ce repo, déployés par le rôle `hermes_agent` vers `~hermes/`.
|
||||
|
||||
```
|
||||
hermes-skills/
|
||||
├── funk/
|
||||
│ ├── agent-delegation/SKILL.md # usage de ask-agent (2 étapes, agents dispo)
|
||||
│ └── rag-docs/SKILL.md # usage de rag-query
|
||||
└── souls/ # prompts système par profil
|
||||
├── funk-ai.md # profil principal (agent sur storage-01)
|
||||
├── system.md # profil calculs/formatage
|
||||
├── monitor.md # profil analyse logs
|
||||
└── brain.md # profil Claude API
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## Base de connaissance — `admin/`
|
||||
|
||||
Documentation admin versionnée, indexée dans le RAG :
|
||||
|
||||
```
|
||||
admin/
|
||||
├── ia/ # hermes, litellm, llama_server, rocm, rag, alertmanager-webhook
|
||||
├── infra/ # dnsmasq, nfs, réseau, ssh, email
|
||||
├── install/ # procédures installation storage-01, gpu-01, k8s
|
||||
├── k8s/ # argocd, talos, monitoring
|
||||
└── ops/ # ansible, cluster, système
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## Réseaux
|
||||
|
||||
- CNI Flannel — Pods : `10.42.0.0/16`
|
||||
- Services : `10.43.0.0/16`
|
||||
- MetalLB pool (IP virtuelles Traefik/services) : `192.168.10.200-230`
|
||||
- MetalLB pool : `192.168.10.200-230` (IP virtuelles Traefik/services)
|
||||
- DNS wildcard `*.lab.local` via dnsmasq → Traefik
|
||||
|
||||
nftables géré par le rôle `gateway` sur storage-01. Modifier `ansible/roles/gateway/templates/nftables.conf.j2` puis `make apply-storage --tags gateway`.
|
||||
|
||||
---
|
||||
|
||||
## Contrainte GPU critique
|
||||
|
||||
La RX 6700XT (gfx1031) n'est **pas officiellement supportée** par ROCm. Toujours s'assurer que la variable est présente dans le service systemd `llama-server` :
|
||||
La RX 6700XT (gfx1031) n'est **pas officiellement supportée** par ROCm. Variable obligatoire dans tous les services GPU :
|
||||
|
||||
```
|
||||
HSA_OVERRIDE_GFX_VERSION=10.3.0
|
||||
```
|
||||
|
||||
ROCm tourne directement sur l'hôte AlmaLinux (pas dans des conteneurs). L'inférence est assurée par llama-server (llama.cpp compilé ROCm 7.x).
|
||||
|
||||
CLI `lms` disponible sur gpu-01 uniquement pour télécharger des modèles (`lms get <url-hf>`, `lms ls`).
|
||||
Gérée automatiquement par le rôle `llama_server`. ROCm tourne directement sur l'hôte AlmaLinux.
|
||||
|
||||
---
|
||||
|
||||
## Deux setups IA
|
||||
|
||||
### Setup 1 — Poste perso (9950X3D) — dev local
|
||||
|
||||
| Composant | Endpoint |
|
||||
|---|---|
|
||||
| Ollama | `http://127.0.0.1:11434` |
|
||||
| LM Studio (optionnel) | `http://127.0.0.1:1234/v1` |
|
||||
| Hermes Agent | `base_url: http://127.0.0.1:1234/v1` ou Ollama |
|
||||
|
||||
### Setup 2 — Cluster Funk (production)
|
||||
|
||||
| Composant | Endpoint |
|
||||
|---|---|
|
||||
| llama-server GPU (gpu-01) | `http://192.168.10.20:1234/v1` |
|
||||
| llama-server CPU system (gpu-01) | `http://192.168.10.20:1236/v1` |
|
||||
| llama-server CPU monitor (gpu-01) | `http://192.168.10.20:1237/v1` |
|
||||
| LiteLLM proxy (storage-01) | `http://127.0.0.1:4000/v1` (master_key: `lm-studio`) |
|
||||
| Hermes Agent | service systemd sur storage-01, via LiteLLM |
|
||||
## Gestion du cluster
|
||||
|
||||
```bash
|
||||
# Vérifier l'état complet depuis le poste perso
|
||||
funk-status
|
||||
funk-cluster status # état de tous les services (storage-01 + gpu-01 + k8s)
|
||||
funk-cluster start # WOL gpu-01 + démarrage services storage-01
|
||||
funk-cluster start --k8s # idem + nœuds k8s (WOL → drain-aware boot)
|
||||
funk-cluster stop # arrêt propre gpu-01 + services
|
||||
funk-cluster stop --k8s # idem + drain k8s → NFS check → talosctl shutdown
|
||||
```
|
||||
|
||||
**etcd backup** (single control-plane → critique) :
|
||||
```bash
|
||||
talosctl etcd snapshot /path/to/backup.db --nodes compute-01
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## Conventions
|
||||
|
||||
- **Domaine** : `lab.local` — wildcard DNS via dnsmasq sur storage-01
|
||||
- **Ingress** : `<service>.lab.local` → IP MetalLB de Traefik (`192.168.10.200-230`)
|
||||
- **llama-server endpoint** : `http://192.168.10.20:1234/v1` (GPU) — accès via LiteLLM sur storage-01
|
||||
- **Namespaces k8s** : `argocd`, `infra`, `storage`, `external`, `ai`, `workflows`
|
||||
- **Git** : branches `main` (prod) / `feature/*` — commits en conventional commits (`feat:`, `fix:`, `chore:`, `docs:`)
|
||||
- **Secrets** : Ansible Vault (Ansible), SOPS+age (Talos via talhelper), Sealed Secrets (k8s)
|
||||
|
||||
## Opérations cluster importantes
|
||||
|
||||
**etcd backup** (single control-plane → critique) :
|
||||
```bash
|
||||
talosctl etcd snapshot /path/to/backup.db --nodes compute-01
|
||||
```
|
||||
À planifier en cron Ansible vers le RAID5 de storage-01.
|
||||
|
||||
**Hermes Agent** — données persistantes dans `~/.hermes/` → doit résider sur le RAID5, pas sur le SSD OS.
|
||||
|
||||
**Ce qui reste toujours manuel** (GitOps ne peut pas bootstrapper lui-même) :
|
||||
- Installation OS depuis USB (AlmaLinux ou Talos)
|
||||
- Configuration IP/SSH minimale pour le premier Ansible
|
||||
- Premier déploiement ArgoCD via Helm
|
||||
|
||||
---
|
||||
|
||||
## RAM serrée sur compute-02/03 (8 GB)
|
||||
|
||||
Toujours définir `resources.requests` et `resources.limits` sur les pods déployés sur ces nœuds — sans quoi l'OOM-killer peut tuer kubelet.
|
||||
|
||||
Réservations à configurer dans talhelper :
|
||||
```
|
||||
system-reserved: cpu=500m,memory=1Gi
|
||||
kube-reserved: cpu=500m,memory=1Gi
|
||||
```
|
||||
- **Git** : branches `main` (prod) / `feature/*` — conventional commits (`feat:`, `fix:`, `chore:`, `docs:`)
|
||||
- **Secrets** : Ansible Vault (Ansible), SOPS+age (Talos), Sealed Secrets (k8s)
|
||||
- **Namespaces k8s** : `argocd`, `infra`, `monitoring`, `external`
|
||||
- **Hermes data** : `~hermes/` doit résider sur le RAID5 `/srv/data`, pas sur le SSD OS
|
||||
- **RAM compute-02/03** : toujours définir `resources.requests` + `resources.limits` (8 GB RAM, OOM-killer actif)
|
||||
|
|
|
|||
172
admin/infra/email.md
Normal file
172
admin/infra/email.md
Normal file
|
|
@ -0,0 +1,172 @@
|
|||
# Email — Relay SMTP via Postfix → Gmail
|
||||
|
||||
## Architecture
|
||||
|
||||
```
|
||||
Hermes / AlertManager / cron / kubectl
|
||||
│
|
||||
▼
|
||||
Postfix (storage-01 :25)
|
||||
smtp_generic_maps → réécrit l'expéditeur en aliyesilkaya93@gmail.com
|
||||
│ SASL + TLS (App Password)
|
||||
▼
|
||||
smtp.gmail.com:587
|
||||
│
|
||||
▼
|
||||
aliyesilkaya93@gmail.com
|
||||
```
|
||||
|
||||
Postfix est configuré en **relay-only** sur storage-01 : aucune livraison locale, tous les mails transitent vers Gmail SMTP. Les réseaux autorisés à relayer : `127.0.0.0/8`, `192.168.10.0/24`, pods k8s (`10.42.0.0/16`), services k8s (`10.43.0.0/16`).
|
||||
|
||||
**Pourquoi Gmail et pas Brevo ?** Brevo exige de posséder un domaine vérifié (SPF/DKIM) pour relayer. Gmail SMTP avec App Password ne nécessite aucun domaine.
|
||||
|
||||
---
|
||||
|
||||
## Prérequis — Google App Password
|
||||
|
||||
La vérification en 2 étapes Google doit être activée sur le compte.
|
||||
|
||||
1. `myaccount.google.com` → Sécurité → Validation en 2 étapes → Activer
|
||||
2. `myaccount.google.com/apppasswords` → Nouveau → nom `funk-postfix` → Créer
|
||||
3. Google génère un code 16 caractères (ex: `abcd efgh ijkl mnop`) — c'est le mot de passe SMTP
|
||||
|
||||
---
|
||||
|
||||
## Installation
|
||||
|
||||
### 1. Ajouter le App Password dans le vault
|
||||
|
||||
```bash
|
||||
make vault-edit
|
||||
```
|
||||
|
||||
Modifier ou ajouter :
|
||||
```yaml
|
||||
vault_postfix_relay_password: "abcd efgh ijkl mnop" # App Password Google (avec espaces, OK)
|
||||
```
|
||||
|
||||
```bash
|
||||
make vault-encrypt
|
||||
```
|
||||
|
||||
### 2. Vérifier les vars gateway
|
||||
|
||||
Dans `ansible/group_vars/gateway/vars.yml` :
|
||||
```yaml
|
||||
postfix_relay_host: smtp.gmail.com
|
||||
postfix_relay_port: 587
|
||||
postfix_relay_user: "aliyesilkaya93@gmail.com"
|
||||
postfix_sender_email: "aliyesilkaya93@gmail.com"
|
||||
```
|
||||
|
||||
### 3. Déployer
|
||||
|
||||
```bash
|
||||
make apply-storage
|
||||
# ou cibler uniquement le rôle :
|
||||
ansible-playbook -i inventory.yml playbooks/storage-01.yml --tags postfix_relay
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## Vérification
|
||||
|
||||
### Test depuis storage-01
|
||||
|
||||
```bash
|
||||
echo "Test Funk Lab - $(date)" | sudo mail -s "[Funk] Test postfix" aliyesilkaya93@gmail.com
|
||||
```
|
||||
|
||||
### Suivre les logs en temps réel
|
||||
|
||||
```bash
|
||||
sudo journalctl -u postfix -f --no-pager --output=cat
|
||||
```
|
||||
|
||||
Résultat attendu : `status=sent (250 2.0.0 OK ... - gsmtp)`
|
||||
|
||||
### Vérifier / vider la file d'attente
|
||||
|
||||
```bash
|
||||
sudo mailq # lister les mails en attente
|
||||
sudo postqueue -f # forcer le flush
|
||||
sudo postsuper -d ALL # vider la queue (mails échoués)
|
||||
```
|
||||
|
||||
### Vérifier la réécriture expéditeur
|
||||
|
||||
```bash
|
||||
sudo postconf smtp_generic_maps
|
||||
sudo cat /etc/postfix/generic
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## Utilisation depuis les services
|
||||
|
||||
### Depuis storage-01 (shell / cron / Hermes)
|
||||
|
||||
```bash
|
||||
echo "Corps du mail" | sudo mail -s "[Funk] Sujet" aliyesilkaya93@gmail.com
|
||||
```
|
||||
|
||||
Via Python (dans un skill Hermes) :
|
||||
```python
|
||||
import smtplib
|
||||
from email.message import EmailMessage
|
||||
|
||||
msg = EmailMessage()
|
||||
msg['Subject'] = '[Funk] Alerte'
|
||||
msg['From'] = 'aliyesilkaya93@gmail.com'
|
||||
msg['To'] = 'aliyesilkaya93@gmail.com'
|
||||
msg.set_content('Contenu du mail')
|
||||
|
||||
with smtplib.SMTP('127.0.0.1', 25) as s:
|
||||
s.send_message(msg)
|
||||
```
|
||||
|
||||
### Depuis AlertManager (k8s)
|
||||
|
||||
Dans `k8s/infra/monitoring/values.yaml` :
|
||||
|
||||
```yaml
|
||||
alertmanager:
|
||||
config:
|
||||
receivers:
|
||||
- name: email
|
||||
email_configs:
|
||||
- to: aliyesilkaya93@gmail.com
|
||||
from: aliyesilkaya93@gmail.com
|
||||
smarthost: "192.168.10.1:25"
|
||||
require_tls: false # TLS géré par Postfix vers Gmail, pas sur le LAN interne
|
||||
route:
|
||||
receiver: email
|
||||
```
|
||||
|
||||
### Depuis n'importe quel pod k8s
|
||||
|
||||
SMTP relay accessible sur `192.168.10.1:25` depuis tout le cluster.
|
||||
|
||||
---
|
||||
|
||||
## Dépannage
|
||||
|
||||
| Symptôme | Cause probable | Action |
|
||||
|---|---|---|
|
||||
| `Connection refused` sur :25 | postfix non démarré | `sudo systemctl status postfix` |
|
||||
| `Relay access denied` | IP source hors mynetworks | `sudo postconf mynetworks` |
|
||||
| `Authentication failed` | App Password révoqué ou 2FA désactivé | Régénérer dans `myaccount.google.com/apppasswords` |
|
||||
| Mail jamais reçu | Réécriture expéditeur manquante | Vérifier `/etc/postfix/generic` et `postconf smtp_generic_maps` |
|
||||
| Mail en spam | Expéditeur non reconnu | Normal si From ≠ aliyesilkaya93@gmail.com |
|
||||
| Queue bloquée | Ancien mail échoué en retry | `sudo postsuper -d ALL` puis retester |
|
||||
|
||||
---
|
||||
|
||||
## Configuration Ansible
|
||||
|
||||
- **Rôle** : `ansible/roles/postfix_relay/`
|
||||
- **Vars** : `ansible/group_vars/gateway/vars.yml` → `postfix_relay_*`, `postfix_sender_email`
|
||||
- **Secret** : `vault_postfix_relay_password` dans `ansible/group_vars/all/vault.yml`
|
||||
- **Templates** : `main.cf.j2`, `sasl_passwd.j2`
|
||||
- **Sender rewrite** : `/etc/postfix/generic` (déployé par le rôle, postmap automatique)
|
||||
- **nftables** : port 25 ouvert automatiquement depuis `192.168.10.0/24`
|
||||
|
|
@ -1,30 +1,34 @@
|
|||
$ANSIBLE_VAULT;1.1;AES256
|
||||
33306234383266653631633563663630353937323132323336356161626533373832313530393064
|
||||
3130383235333530336531636665623737663261316332370a313739353336306130643461653635
|
||||
35353961333236636263613761393935373762336633396337346661373062636131373632366339
|
||||
6562383163316162390a393865383736623132653531643064663338393164366166303532633362
|
||||
38343131656238653562323632303362303566363635316132383335326462316432336330366230
|
||||
63663732356366356234646664646539316531613131313932313237663365643230343063613763
|
||||
63633061663833353337626433666363363336646136376334303166613764323634323463393464
|
||||
37393533333437386666653663323331666130656161646437663734363161653534666661383339
|
||||
38623934643636643361643766356664393335353034373532313262666537373330353762336361
|
||||
64343465333534313239323432656135626237333434653034333364663061383965653838663763
|
||||
63653832313934326163623831653839613330346238643630663539623436373738323837386162
|
||||
66326466623764346666613266663530363664343037643766653562613034363239346639326664
|
||||
30346337663530623732333264383839396534383762623338313437633266656665366538633564
|
||||
63363633343731636234623330393838313237633064333939663236616365326130323735623339
|
||||
34386635623033636166633536633661643133653263646333326363613362303461376432373466
|
||||
36613538393334626266396531616366383931636665633862373331343365623962643362306133
|
||||
66363730626639336232626430636239643261666661396664396661393635306466363338306662
|
||||
35343432346461306330623536363330653262323863343162623965633164653035373432316232
|
||||
36303630306465393837373033656130636261316535323464613563656235643136663038366536
|
||||
36343062623932346161616134653562346666333266393563623938376535323637376265353164
|
||||
65346663373331393561653030356163633038376234663231336332653161636662323537306633
|
||||
38636661613535336463616662363033313961613434323033316230653865356233636138343336
|
||||
35323438656564613931656538646561663432363234653134323933626236646237326636353833
|
||||
36383838666263626632613462646561393035663435373736653861383063623862613131653332
|
||||
66393731623630376463623533336133353036386234633361653237626335323432643332313036
|
||||
31363032323763656530313230346466343537663664623931623933616166333135326437666231
|
||||
30363135643632633962613631656339616665356463653162653832366336663165646264636638
|
||||
66663866303733633037313131353033393263346138643139626539326431383762646565613437
|
||||
32333462353765653931613431366639323434356365376636646237303166323762
|
||||
65373732613635326664316362333337616633646539666662353839353665636135373262623230
|
||||
6433633033393837323330323730356331313031366664380a376533393237613334663635646636
|
||||
30663138346337643935636236646337653762383137306634353437313865356331376630656630
|
||||
6233323836333137330a346530376665623064346462633437656139376433666436333835333835
|
||||
64393738393238633132373831326436343630373261303466376537303731333164353731646434
|
||||
34366437346335653538353763353631623639613865336431653964613163653033653062366232
|
||||
37353132633463396661613864633065306564353337383035303037626435623634646530323464
|
||||
63346230383463323534653434663632303030333030636636306234353566343039363433643938
|
||||
65393063633962376133353264376261386436386661633933613135653537663039346630373033
|
||||
32646534653163303239646131306131666632623337363836623033656562353334313366633934
|
||||
61623436346431623263633536653236393235393539616133376135656164323336636664333566
|
||||
36616230613564626338333934386336616631313739366330303431333836373935356465613365
|
||||
61336662636535333566313763383738363363393535323434336234303130643632646364373433
|
||||
37343635623165343965613263616535373036616337613234656231313230623330383566623633
|
||||
63393263383039613639316135633133636364613532633237313934313136336661376635383435
|
||||
65633065653539363066633832396437613364303634656334386266636461393663623462323030
|
||||
37653138323037323965336165626531663463383334663937666536663935303230313235383038
|
||||
66353162306233313034326431383438333031393735363264386232333034313664366630366262
|
||||
36643164663837363132653165636635643162303331386262653130303438383038646361363564
|
||||
62643962613061326332643862623538646262333264656335323235353733663563336639633262
|
||||
62616161306235656666353738313234316462373065373330623134323836633833343032623832
|
||||
35363663653463616265303535623066623134356233663638333565623034323362323163656633
|
||||
32316363313865653633636665333561303866376339366163623331396430303436346633323563
|
||||
34633266306435633535306439323638336532656335663864646434343233366535396335643763
|
||||
36363662323861643638656630383738356331343834303736393337336435646663333134306532
|
||||
35383039623432353161633036323430333036613239356439376666323866653331376362613338
|
||||
32356661336131373334333132333039623665383065346439626531653364346434643734313237
|
||||
34343966313934613734646465663231396435333135356165366435386366336338343466363331
|
||||
36666461363735636637613364656366346263353338333539353662643039383338396437316633
|
||||
38343434386533613439316632313963346563373131333965396637313937656265346133373635
|
||||
33643338316433323139636262353636616231636564383332343862656661636463663938636131
|
||||
38386665663635356666336261633761653362373534363161623166643830373737363730613361
|
||||
6262
|
||||
|
|
|
|||
|
|
@ -3,3 +3,10 @@ wan_interface: enp4s0 # NIC1 — port 2 switch — VLAN 1 (192.168.1.20
|
|||
lan_interface: enp6s0f3u2c2 # NIC2 — port 7 switch — VLAN 10 (192.168.10.1)
|
||||
wan_ip: 192.168.1.200
|
||||
lan_ip: 192.168.10.1
|
||||
|
||||
# Postfix relay — SMTP sortant via provider externe
|
||||
postfix_relay_host: smtp.gmail.com
|
||||
postfix_relay_port: 587
|
||||
postfix_relay_user: "aliyesilkaya93@gmail.com"
|
||||
postfix_sender_email: "aliyesilkaya93@gmail.com" # expéditeur = compte Gmail (DMARC OK)
|
||||
postfix_relay_password: "{{ vault_postfix_relay_password }}" # dans vault.yml
|
||||
|
|
|
|||
|
|
@ -15,3 +15,4 @@
|
|||
- { role: rag, tags: [rag] }
|
||||
- { role: node_exporter, tags: [node_exporter] }
|
||||
- { role: alertmanager_webhook, tags: [alertmanager_webhook] }
|
||||
- { role: postfix_relay, tags: [postfix_relay] }
|
||||
|
|
|
|||
|
|
@ -2,4 +2,4 @@
|
|||
- name: Reload nftables
|
||||
ansible.builtin.systemd:
|
||||
name: nftables
|
||||
state: reloaded
|
||||
state: restarted
|
||||
|
|
|
|||
|
|
@ -69,6 +69,10 @@ table inet filter {
|
|||
}
|
||||
|
||||
table ip nat {
|
||||
chain prerouting {
|
||||
type nat hook prerouting priority -100;
|
||||
}
|
||||
|
||||
chain postrouting {
|
||||
type nat hook postrouting priority 100;
|
||||
|
||||
|
|
|
|||
|
|
@ -76,14 +76,17 @@
|
|||
|
||||
# --- Packages Python pour les skills ---
|
||||
|
||||
- name: Install Python packages for Hermes skills
|
||||
ansible.builtin.pip:
|
||||
name: "qdrant-client>=1.12.0"
|
||||
executable: "{{ hermes_home }}/.hermes/hermes-agent/venv/bin/pip"
|
||||
state: present
|
||||
- name: Install Python packages for Hermes skills (via uv)
|
||||
ansible.builtin.command:
|
||||
cmd: >
|
||||
{{ hermes_home }}/.local/bin/uv pip install qdrant-client>=1.12.0
|
||||
--python {{ hermes_home }}/.hermes/hermes-agent/venv/bin/python
|
||||
environment:
|
||||
UV_CACHE_DIR: /tmp/uv-cache
|
||||
become: true
|
||||
become_user: "{{ hermes_user }}"
|
||||
become_user: root
|
||||
when: hermes_binary.stat.exists
|
||||
changed_when: false
|
||||
|
||||
# --- Configuration ---
|
||||
|
||||
|
|
|
|||
22
ansible/roles/postfix_relay/defaults/main.yml
Normal file
22
ansible/roles/postfix_relay/defaults/main.yml
Normal file
|
|
@ -0,0 +1,22 @@
|
|||
---
|
||||
postfix_relay_host: smtp-relay.brevo.com
|
||||
postfix_relay_port: 587
|
||||
postfix_relay_user: "" # à définir dans group_vars/gateway/vars.yml
|
||||
postfix_relay_password: "" # depuis vault (vault_postfix_relay_password)
|
||||
|
||||
postfix_myhostname: "storage-01.lab.local"
|
||||
postfix_myorigin: "$myhostname"
|
||||
|
||||
# Adresse expéditeur réécrite pour tous les mails sortants
|
||||
postfix_sender_email: "" # à définir dans group_vars/gateway/vars.yml
|
||||
|
||||
# Réseaux autorisés à relayer — ajuster si besoin
|
||||
postfix_mynetworks:
|
||||
- "127.0.0.0/8"
|
||||
- "[::1]/128"
|
||||
- "192.168.10.0/24" # LAN cluster
|
||||
- "10.42.0.0/16" # pods k8s
|
||||
- "10.43.0.0/16" # services k8s
|
||||
|
||||
# Ouvrir le port 25 depuis le LAN cluster dans nftables
|
||||
postfix_open_nftables: true
|
||||
10
ansible/roles/postfix_relay/handlers/main.yml
Normal file
10
ansible/roles/postfix_relay/handlers/main.yml
Normal file
|
|
@ -0,0 +1,10 @@
|
|||
---
|
||||
- name: Restart postfix
|
||||
ansible.builtin.systemd:
|
||||
name: postfix
|
||||
state: restarted
|
||||
|
||||
- name: Reload nftables
|
||||
ansible.builtin.systemd:
|
||||
name: nftables
|
||||
state: reloaded
|
||||
72
ansible/roles/postfix_relay/tasks/main.yml
Normal file
72
ansible/roles/postfix_relay/tasks/main.yml
Normal file
|
|
@ -0,0 +1,72 @@
|
|||
---
|
||||
- name: Install postfix and SASL auth library
|
||||
ansible.builtin.dnf:
|
||||
name:
|
||||
- postfix
|
||||
- cyrus-sasl-plain
|
||||
- s-nail
|
||||
state: present
|
||||
|
||||
- name: Deploy main.cf
|
||||
ansible.builtin.template:
|
||||
src: main.cf.j2
|
||||
dest: /etc/postfix/main.cf
|
||||
owner: root
|
||||
group: root
|
||||
mode: '0644'
|
||||
notify: Restart postfix
|
||||
|
||||
- name: Deploy sasl_passwd
|
||||
ansible.builtin.template:
|
||||
src: sasl_passwd.j2
|
||||
dest: /etc/postfix/sasl_passwd
|
||||
owner: root
|
||||
group: root
|
||||
mode: '0600'
|
||||
register: sasl_passwd_result
|
||||
|
||||
- name: Run postmap on sasl_passwd
|
||||
ansible.builtin.command: postmap /etc/postfix/sasl_passwd
|
||||
when: sasl_passwd_result.changed
|
||||
changed_when: true
|
||||
notify: Restart postfix
|
||||
|
||||
- name: Deploy generic sender rewrite map
|
||||
ansible.builtin.copy:
|
||||
dest: /etc/postfix/generic
|
||||
content: |
|
||||
# Réécriture expéditeur — géré par Ansible
|
||||
@{{ postfix_myhostname }} {{ postfix_sender_email }}
|
||||
root {{ postfix_sender_email }}
|
||||
owner: root
|
||||
group: root
|
||||
mode: '0644'
|
||||
register: generic_map_result
|
||||
when: postfix_sender_email | length > 0
|
||||
|
||||
- name: Run postmap on generic map
|
||||
ansible.builtin.command: postmap /etc/postfix/generic
|
||||
when: generic_map_result is changed
|
||||
changed_when: true
|
||||
notify: Restart postfix
|
||||
|
||||
- name: Reset postfix failed state if any
|
||||
ansible.builtin.command: systemctl reset-failed postfix.service
|
||||
changed_when: false
|
||||
failed_when: false
|
||||
|
||||
- name: Enable and start postfix
|
||||
ansible.builtin.systemd:
|
||||
name: postfix
|
||||
enabled: true
|
||||
state: started
|
||||
daemon_reload: true
|
||||
|
||||
- name: Open port 25 from cluster LAN in nftables
|
||||
ansible.builtin.lineinfile:
|
||||
path: /etc/nftables.conf
|
||||
insertafter: '# alertmanager-webhook'
|
||||
line: " tcp dport 25 ip saddr 192.168.10.0/24 accept comment \"postfix-relay\""
|
||||
state: present
|
||||
when: postfix_open_nftables
|
||||
notify: Reload nftables
|
||||
38
ansible/roles/postfix_relay/templates/main.cf.j2
Normal file
38
ansible/roles/postfix_relay/templates/main.cf.j2
Normal file
|
|
@ -0,0 +1,38 @@
|
|||
# Postfix — relay-only — géré par Ansible (rôle postfix_relay)
|
||||
# Ne pas éditer manuellement
|
||||
|
||||
compatibility_level = 2
|
||||
myhostname = {{ postfix_myhostname }}
|
||||
myorigin = {{ postfix_myorigin }}
|
||||
|
||||
# Relay SMTP externe
|
||||
relayhost = [{{ postfix_relay_host }}]:{{ postfix_relay_port }}
|
||||
|
||||
# Écouter sur toutes les interfaces pour permettre au cluster de relayer
|
||||
inet_interfaces = all
|
||||
inet_protocols = ipv4
|
||||
|
||||
# Réseaux autorisés à relayer (pas un open relay)
|
||||
mynetworks = {{ postfix_mynetworks | join(' ') }}
|
||||
|
||||
# Désactiver la livraison locale
|
||||
mydestination =
|
||||
local_recipient_maps =
|
||||
local_transport = error:local delivery disabled
|
||||
|
||||
# TLS sortant
|
||||
smtp_tls_security_level = encrypt
|
||||
smtp_tls_CAfile = /etc/ssl/certs/ca-bundle.crt
|
||||
smtp_tls_loglevel = 1
|
||||
|
||||
# Authentification SASL vers le relay
|
||||
smtp_sasl_auth_enable = yes
|
||||
smtp_sasl_password_maps = hash:/etc/postfix/sasl_passwd
|
||||
smtp_sasl_security_options = noanonymous
|
||||
smtp_sasl_tls_security_options = noanonymous
|
||||
|
||||
# Réécriture expéditeur — remplace root@lab.local par une adresse réelle
|
||||
{% if postfix_sender_email %}
|
||||
smtp_generic_maps = hash:/etc/postfix/generic
|
||||
{% endif %}
|
||||
|
||||
2
ansible/roles/postfix_relay/templates/sasl_passwd.j2
Normal file
2
ansible/roles/postfix_relay/templates/sasl_passwd.j2
Normal file
|
|
@ -0,0 +1,2 @@
|
|||
# Géré par Ansible — ne pas éditer manuellement
|
||||
[{{ postfix_relay_host }}]:{{ postfix_relay_port }} {{ postfix_relay_user }}:{{ postfix_relay_password }}
|
||||
Loading…
Add table
Add a link
Reference in a new issue