dev #20250613-01

2025-06-13 17:07:45 +02:00 · 2025-06-13 17:07:45 +02:00 · b3eefd65a2
commit b3eefd65a2
parent 232a151b85
9 changed files with 855 additions and 24 deletions
--- a/.env.dist
+++ b/.env.dist
@ -47,4 +47,9 @@ LOKI_GEN_LOG_DIR=<path_to_your_loki_data_dir>
 #------------------------------------------------
 # promtail
 # conf dir promatal.yaml (ex: /home/docker/promtail)
-PROMTAIL_CONF_DIR=<path_to_your_promtail_conf_dir>
+PROMTAIL_CONF_DIR=<path_to_your_promtail_conf_dir>
+#------------------------------------------------
+# flog
+# flog docker image
+#FLOG_DOCKER_IMG=mingrammer/flog # (amd64)
+#FLOG_DOCKER_IMG=gmouchet/flog # (arm64)
--- a/FROM-AI.md
+++ b/FROM-AI.md
@ -555,4 +555,626 @@ scrape_configs:
          __path__: /var/log/*log
 ✅ Démarrage

-docker-compose up -d
+docker-compose up -d
+
+
+PROBLEME CPU/MEM
+Lorsque vous voyez une utilisation à 100% du CPU et de la mémoire avec Loki et Promtail, il y a plusieurs causes fréquentes. Voici comment analyser et résoudre ce problème.
+🔍 Analyse des causes possibles
+1. Trop de logs générés par flog
+
+    Problème : flog est configuré avec --number=10 et --delay=100ms, ce qui signifie potentiellement 10 logs tous les 100ms, soit 100 logs/seconde.
+
+    Impact : Cela peut engorger promtail, qui tente de parser et pousser un très grand nombre d’entrées à loki.
+
+2. Promtail consomme trop pour parser les logs
+
+    Le bloc :
+
+    pipeline_stages:
+      - json:
+          expressions:
+            http_method: 'method'
+            http_status: "status"
+      - labels:
+          http_method:
+          http_status:
+
+    est très coûteux si les logs sont volumineux et mal structurés (ou si les champs ne sont pas toujours présents).
+
+3. Pas de limits sur les ressources
+
+    Aucun conteneur ne semble avoir de limites de ressources Docker, donc ils peuvent saturer votre système.
+
+✅ Recommandations concrètes
+🧹 A. Réduire la charge générée par flog
+
+Modifiez la configuration pour réduire le débit :
+
+command:
+  - --loop
+  - --format=json
+  - --number=1       # Réduit à 1 log/sec
+  - --delay=1000ms   # Une entrée par seconde
+
+⚙️ B. Ajoutez des limites de ressources aux conteneurs lourds
+
+Exemple pour promtail :
+
+  promtail:
+    ...
+    deploy:
+      resources:
+        limits:
+          cpus: '0.50'
+          memory: 256M
+        reservations:
+          cpus: '0.25'
+          memory: 128M
+
+Idem pour loki, flog et éventuellement grafana.
+🔍 C. Surveillez le volume de logs et les erreurs
+
+Ajoutez un volume pour visualiser les erreurs Promtail :
+
+    volumes:
+      - /var/log:/var/log
+      - ${PROMTAIL_CONF_DIR}/config.yml:/etc/promtail/config.yml
+      - /tmp/promtail-log:/tmp/promtail-log
+
+Puis dans Promtail :
+
+positions:
+  filename: /tmp/promtail-log/positions.yaml
+
+🧪 D. Testez sans pipeline_stages pour vérifier l’impact
+
+Pour voir si le parsing JSON est le coupable, commentez temporairement :
+
+# pipeline_stages:
+#   - json:
+#       expressions:
+#         http_method: 'method'
+#         http_status: "status"
+#   - labels:
+#       http_method:
+#       http_status:
+
+📊 Outils de surveillance
+
+Lancez docker stats pour identifier en temps réel quel conteneur consomme le plus de ressources.
+✅ En résumé
+Action	Objectif
+Réduire le débit de flog	Diminuer la charge Promtail
+Limiter les ressources Promtail/Loki	Empêcher de saturer la machine
+Désactiver temporairement pipeline_stages	Tester l’impact du parsing JSON
+Analyser les logs Promtail	Voir s’il y a des erreurs
+
+Souhaitez-vous que je vous fournisse un fichier docker-compose.yaml corrigé avec ces ajustements ?
+Vous avez dit :
+oui
+ChatGPT a dit :
+
+Voici un fichier docker-compose.yaml corrigé avec :
+
+✅ Débit réduit du générateur de logs (flog)
+✅ Limites de ressources pour loki, promtail, flog
+✅ Répertoire pour logs de position de promtail
+✅ Parsing JSON désactivé temporairement dans promtail.yml (à réactiver si stable)
+✅ docker-compose.yaml modifié :
+
+networks:
+  loki:
+    name: loki-net
+
+services:
+  init:
+    image: busybox
+    user: root
+    entrypoint: >
+      /bin/sh -c "
+      chown 472:root /var/lib/grafana;
+      chown 101:root /etc/nginx/conf.d/default.conf;
+      chown 10001:10001 /loki -R;
+      chown nobody:nobody /prometheus -R;
+      exit 0;"
+    volumes:
+      - '${GF_DATA_DIR}:/var/lib/grafana'
+      - '${PRX_NGINX_CONF_DIR}/default.conf:/etc/nginx/conf.d/default.conf'
+      - '${PROM_DATA_DIR}:/prometheus'
+      - '${LOKI_DATA_DIR}:/loki'
+    networks:
+      - loki
+
+  proxy:
+    image: nginxinc/nginx-unprivileged
+    volumes:
+      - ${PRX_NGINX_CONF_DIR}/default.conf:/etc/nginx/conf.d/default.conf
+      - ${PRX_CERTS_DIR}:/etc/nginx/certs
+    ports:
+      - "443:8443"
+    networks:
+      - loki
+    restart: always
+
+  minio:
+    image: minio/minio:RELEASE.2025-03-12T18-04-18Z
+    restart: always
+    volumes:
+      - ${MINIO_DATA_ROOT_DIR}:/data
+    ports:
+      - "9000:9000"
+      - "9001:9001"
+    environment:
+      MINIO_ROOT_USER: ${MINIO_ACCESS_KEY}
+      MINIO_ROOT_PASSWORD: ${MINIO_SECRET_KEY}
+      MINIO_BROWSER_REDIRECT_URL: ${MINIO_REDIRECT_URL}
+    command: server /data --console-address ":9001" 
+    networks:
+      - loki
+
+  minio-cli:
+    image: minio/mc
+    depends_on:
+      - minio
+    entrypoint: >
+      /bin/sh -c "tail -f /dev/null"
+    volumes:
+      - ${MINIO_CONF_CLI_DIR}/.mc:/root/.mc  
+    restart: unless-stopped
+    networks:
+      - loki  
+
+  grafana:
+    image: grafana/grafana-enterprise
+    ports:
+      - "3000:3000"
+    volumes:
+      - '${GF_DATA_DIR}:/var/lib/grafana'
+    environment:
+      - GF_SECURITY_ADMIN_USER=${GF_ADMIN_USER}
+      - GF_SECURITY_ADMIN_PASSWORD=${GF_ADMIN_PASS}
+      - GF_SERVER_ROOT_URL=${GF_ROOT_URL}
+    restart: unless-stopped
+    networks:
+      - loki
+
+  prometheus:
+    image: prom/prometheus
+    volumes:
+      - ${PROM_CONF_DIR}/prometheus.yml:/etc/prometheus/prometheus.yml
+      - ${PROM_DATA_DIR}:/prometheus 
+    ports:
+      - 9090:9090
+    restart: unless-stopped
+    networks:
+      - loki
+
+  node-exporter:
+    image: prom/node-exporter
+    networks:
+      - loki
+    ports:
+      - 9100:9100
+    restart: unless-stopped
+
+  loki:
+    image: grafana/loki:latest
+    ports:
+      - 3100:3100
+    volumes:
+      - ${LOKI_CONF_DIR}/local-config.yaml:/etc/loki/local-config.yaml
+      - ${LOKI_DATA_DIR}:/loki
+    restart: unless-stopped
+    networks:
+      - loki
+    deploy:
+      resources:
+        limits:
+          cpus: '1.0'
+          memory: 512M
+        reservations:
+          cpus: '0.5'
+          memory: 256M
+
+  promtail:
+    image: grafana/promtail:latest
+    volumes:
+      - ${PROMTAIL_CONF_DIR}/config.yml:/etc/promtail/config.yml
+      - ${LOKI_GEN_LOG_DIR}:/tmp/
+      - /var/log/:/var/log/
+      - /tmp/promtail-log:/tmp/promtail-log
+    restart: unless-stopped
+    networks:
+      - loki
+    deploy:
+      resources:
+        limits:
+          cpus: '0.5'
+          memory: 256M
+        reservations:
+          cpus: '0.25'
+          memory: 128M
+
+  log-generator:
+    image: mingrammer/flog
+    command:
+      - --loop
+      - --format=json
+      - --number=1         # Réduit à 1 ligne/sec
+      - --delay=1000ms     # Pause 1s entre chaque log
+      - --output=/tmp/loki-gen-log/generated-logs.txt
+      - --overwrite
+      - --type=log
+    volumes:
+      - ${LOKI_GEN_LOG_DIR}:/tmp/
+    deploy:
+      resources:
+        limits:
+          cpus: '0.25'
+          memory: 128M
+        reservations:
+          cpus: '0.1'
+          memory: 64M
+
+✅ promtail.yml modifié (pipeline JSON désactivé) :
+
+server:
+  http_listen_port: 9080
+  grpc_listen_port: 0
+
+positions:
+  filename: /tmp/promtail-log/positions.yaml
+
+clients:
+  - url: http://loki:3100/loki/api/v1/push
+
+scrape_configs:
+- job_name: system
+  static_configs:
+    - targets:
+        - localhost
+      labels:
+        job: varlogs
+        __path__: /var/log/*log
+
+- job_name: generated-logs
+  static_configs:
+    - targets:
+        - localhost
+      labels:
+        job: generated-logs
+        __path__: /tmp/loki-gen-log/generated-logs.txt
+
+# Désactivé temporairement pour test de perf
+#  pipeline_stages:
+#    - json:
+#        expressions:
+#          http_method: 'method'
+#          http_status: "status"
+#    - labels:
+#        http_method:
+#        http_status:
+
+Souhaitez-vous également un script Bash pour automatiser le démarrage avec vérification de la charge ?
+Vous avez dit :
+oui
+ChatGPT a dit :
+
+Voici un script Bash pour :
+
+    Lancer vos conteneurs avec docker compose up
+
+    Attendre que les conteneurs soient “healthy”
+
+    Surveiller brièvement l’usage CPU/mémoire avec docker stats
+
+    Alerter si un conteneur dépasse les seuils définis
+
+✅ start-and-monitor.sh
+
+#!/bin/bash
+
+set -e
+
+COMPOSE_FILE="docker-compose.yaml"
+MAX_CPU=80     # Seuil CPU en %
+MAX_MEM_MB=500 # Seuil mémoire en Mo
+
+echo "🔄 Lancement des conteneurs avec Docker Compose..."
+docker compose -f "$COMPOSE_FILE" up -d
+
+echo "⏳ Attente du démarrage des conteneurs..."
+sleep 10
+
+echo "📦 Conteneurs lancés :"
+docker compose ps
+
+echo "⏱ Surveillance des ressources pendant 30s..."
+docker stats --no-stream --format "table {{.Name}}\t{{.CPUPerc}}\t{{.MemUsage}}" > /tmp/docker_stats_snapshot.txt
+
+echo "📊 Résultat :"
+cat /tmp/docker_stats_snapshot.txt
+
+echo ""
+echo "🚨 Vérification des seuils (CPU > ${MAX_CPU}% ou MEM > ${MAX_MEM_MB}MB)..."
+while read -r line; do
+  NAME=$(echo $line | awk '{print $1}')
+  CPU=$(echo $line | awk '{print $2}' | sed 's/%//')
+  MEM_RAW=$(echo $line | awk '{print $3}')
+  MEM_UNIT=$(echo $MEM_RAW | grep -o '[MG]B')
+  MEM_VAL=$(echo $MEM_RAW | grep -o '[0-9.]*')
+
+  if [[ $MEM_UNIT == "GB" ]]; then
+    MEM_MB=$(echo "$MEM_VAL * 1024" | bc)
+  else
+    MEM_MB=$MEM_VAL
+  fi
+
+  CPU_EXCEED=$(echo "$CPU > $MAX_CPU" | bc)
+  MEM_EXCEED=$(echo "$MEM_MB > $MAX_MEM_MB" | bc)
+
+  if [[ "$CPU_EXCEED" -eq 1 || "$MEM_EXCEED" -eq 1 ]]; then
+    echo "❌ $NAME dépasse les limites (CPU=${CPU}%, MEM=${MEM_MB}MB)"
+  fi
+done < <(tail -n +2 /tmp/docker_stats_snapshot.txt) # skip header
+
+echo "✅ Surveillance terminée."
+
+🔧 Instructions
+
+    Enregistrez ce script sous start-and-monitor.sh
+
+    Rendez-le exécutable :
+
+chmod +x start-and-monitor.sh
+
+    Exécutez-le :
+
+./start-and-monitor.sh
+
+Souhaitez-vous que je vous génère aussi un dashboard Grafana Loki/Promtail prêt à importer ?
+Vous avez dit :
+oui
+ChatGPT a dit :
+
+Voici un dashboard Grafana Loki/Promtail prêt à importer, qui vous permettra de :
+
+    Visualiser les logs par job (varlogs, generated-logs)
+
+    Compter les logs par http_status
+
+    Afficher les volumes de logs en temps réel
+
+    Rechercher les erreurs (level=error)
+
+✅ 1. Dashboard JSON à importer dans Grafana
+
+Enregistrez le contenu suivant dans un fichier nommé :
+
+grafana-loki-dashboard.json
+
+{
+  "uid": "loki-log-monitor",
+  "title": "Loki & Promtail Logs",
+  "timezone": "browser",
+  "schemaVersion": 37,
+  "version": 1,
+  "refresh": "5s",
+  "panels": [
+    {
+      "type": "logs",
+      "title": "🔍 Logs récents - All Jobs",
+      "datasource": "Loki",
+      "targets": [
+        {
+          "expr": "{job=~\"varlogs|generated-logs\"}",
+          "refId": "A"
+        }
+      ],
+      "gridPos": { "x": 0, "y": 0, "w": 24, "h": 10 }
+    },
+    {
+      "type": "graph",
+      "title": "📊 Volume de logs par job (lignes/sec)",
+      "datasource": "Loki",
+      "targets": [
+        {
+          "expr": "rate({job=~\"varlogs|generated-logs\"}[1m])",
+          "refId": "B"
+        }
+      ],
+      "yaxes": [
+        {
+          "label": "Logs/sec",
+          "format": "short"
+        },
+        {
+          "format": "short"
+        }
+      ],
+      "gridPos": { "x": 0, "y": 10, "w": 12, "h": 8 }
+    },
+    {
+      "type": "stat",
+      "title": "🚨 Erreurs détectées (status >= 500)",
+      "datasource": "Loki",
+      "targets": [
+        {
+          "expr": "count_over_time({job=\"generated-logs\", http_status=~\"5..\"}[5m])",
+          "refId": "C"
+        }
+      ],
+      "gridPos": { "x": 12, "y": 10, "w": 12, "h": 8 }
+    },
+    {
+      "type": "barchart",
+      "title": "📌 Repartition par http_status",
+      "datasource": "Loki",
+      "targets": [
+        {
+          "expr": "count_over_time({job=\"generated-logs\"}[1m]) by (http_status)",
+          "refId": "D"
+        }
+      ],
+      "gridPos": { "x": 0, "y": 18, "w": 24, "h": 8 }
+    }
+  ]
+}
+
+📥 2. Importer dans Grafana
+
+    Ouvrez Grafana : http://localhost:3000
+
+    Allez dans Dashboards > New > Import
+
+    Cliquez sur Upload JSON file et sélectionnez grafana-loki-dashboard.json
+
+    Choisissez la source de données Loki
+
+    Cliquez sur Import
+
+🧪 Résultat attendu
+
+Vous verrez :
+
+    Tous les logs Promtail en live
+
+    Stats agrégées par job
+
+    Détection d’erreurs HTTP 5xx
+
+    Histogramme par http_status
+
+Souhaitez-vous également un dashboard Prometheus pour surveiller l’utilisation CPU/MEM de vos conteneurs ?
+Vous avez dit :
+oui
+ChatGPT a dit :
+
+Parfait ! Voici un dashboard Grafana Prometheus prêt à importer, pour surveiller :
+
+    🧠 Mémoire utilisée par conteneur Docker
+
+    🔥 CPU par conteneur
+
+    ⚠️ Alertes sur consommation élevée
+
+    📊 Résumé graphique par nom de conteneur (container_label_com_docker_compose_service)
+
+✅ 1. Dashboard JSON – Prometheus Docker Containers
+
+Enregistrez ce fichier sous :
+
+grafana-prometheus-containers.json
+
+{
+  "uid": "docker-containers",
+  "title": "🧠 Docker - CPU / Memory Usage",
+  "version": 1,
+  "schemaVersion": 37,
+  "refresh": "10s",
+  "panels": [
+    {
+      "type": "graph",
+      "title": "🔥 CPU Usage (%) by Container",
+      "datasource": "Prometheus",
+      "targets": [
+        {
+          "expr": "rate(container_cpu_usage_seconds_total{container_label_com_docker_compose_service!=\"\", image!=\"\"}[1m]) * 100",
+          "legendFormat": "{{ container_label_com_docker_compose_service }}",
+          "refId": "A"
+        }
+      ],
+      "yaxes": [
+        {
+          "format": "percent",
+          "label": "CPU %",
+          "min": 0,
+          "max": 100
+        },
+        {
+          "format": "short"
+        }
+      ],
+      "gridPos": { "x": 0, "y": 0, "w": 24, "h": 8 }
+    },
+    {
+      "type": "graph",
+      "title": "🧠 Memory Usage (MB) by Container",
+      "datasource": "Prometheus",
+      "targets": [
+        {
+          "expr": "container_memory_usage_bytes{container_label_com_docker_compose_service!=\"\", image!=\"\"} / 1024 / 1024",
+          "legendFormat": "{{ container_label_com_docker_compose_service }}",
+          "refId": "B"
+        }
+      ],
+      "yaxes": [
+        {
+          "format": "megabytes",
+          "label": "Memory (MB)",
+          "min": 0
+        },
+        {
+          "format": "short"
+        }
+      ],
+      "gridPos": { "x": 0, "y": 8, "w": 24, "h": 8 }
+    },
+    {
+      "type": "stat",
+      "title": "🚨 Containers with >80% CPU",
+      "datasource": "Prometheus",
+      "targets": [
+        {
+          "expr": "count(count_over_time(rate(container_cpu_usage_seconds_total[1m]) * 100 > 80)[1m:1s])",
+          "refId": "C"
+        }
+      ],
+      "gridPos": { "x": 0, "y": 16, "w": 12, "h": 4 }
+    },
+    {
+      "type": "stat",
+      "title": "⚠️ Containers with >500MB RAM",
+      "datasource": "Prometheus",
+      "targets": [
+        {
+          "expr": "count(count_over_time(container_memory_usage_bytes / 1024 / 1024 > 500)[1m:1s])",
+          "refId": "D"
+        }
+      ],
+      "gridPos": { "x": 12, "y": 16, "w": 12, "h": 4 }
+    }
+  ]
+}
+
+📥 2. Import dans Grafana
+
+    Accédez à Grafana (http://localhost:3000)
+
+    Menu Dashboards > Import
+
+    Cliquez sur Upload JSON file
+
+    Sélectionnez grafana-prometheus-containers.json
+
+    Choisissez la source Prometheus
+
+    Cliquez sur Import
+
+✅ Prérequis Prometheus
+
+Assurez-vous que prometheus.yml scrape bien les métriques Docker via cAdvisor ou le Node Exporter.
+
+Exemple pour Node Exporter :
+
+scrape_configs:
+  - job_name: 'node'
+    static_configs:
+      - targets: ['node-exporter:9100']
+
+Pour cAdvisor (optionnel mais utile) :
+
+  - job_name: 'cadvisor'
+    static_configs:
+      - targets: ['cadvisor:8080']
--- a/README.md
+++ b/README.md
@ -195,6 +195,7 @@ docker exec -it gmo-loki-proxy-1 /bin/bash -c "curl -H \"Accept: application/jso
  - 1860 - Node Exporter Full
  - 3662 - Prometheus 2.0 Overview
  - 13502 - MinIO Dashboard
+  - 13639 Loki - Log analysis
   
 ### [1.0.0] - 2025-06-08
 #### Added
--- a/config/config.yml
+++ b/config/config.yml
@ -23,7 +23,7 @@ scrape_configs:
      - localhost
    labels:
      job: generated-logs
-      __path__: /tmp/loki-gen-log/generated-logs.txt
+      __path__: /tmp/generated-logs.txt
  pipeline_stages:
    - json:
        expressions:
--- a/config/default.conf
+++ b/config/default.conf
@ -1,3 +1,8 @@
+map $http_upgrade $connection_upgrade {  
+    default upgrade;
+    ''      close;
+}
+
 server {
    listen 8080 default_server;
    server_name _;
@ -26,11 +31,23 @@ server {
  }

  #------- GRAFANA -------
-  location /grafana/ {
-    proxy_pass http://grafana:3000/;
+  location /grafana {
    proxy_set_header Host $host;
+    proxy_pass http://grafana:3000/;
+    rewrite ^/grafana/(.*) /$1 break;
  }

+    # Proxy Grafana Live WebSocket connections.
+  location /grafana/api/live/ {
+    proxy_http_version 1.1;
+    proxy_set_header Upgrade $http_upgrade;
+    proxy_set_header Connection $connection_upgrade;
+    proxy_set_header Host $host;
+    proxy_pass http://grafana:3000;
+    rewrite  ^/grafana/(.*)  /$1 break;
+  }
+  
+
  #------- MINIO UI -------
  location /minio/ui/ {
    rewrite ^/minio/ui/(.*) /$1 break;
--- a/config/local-config.yaml
+++ b/config/local-config.yaml
@ -128,13 +128,25 @@ storage_config:
    insecure: true
    s3forcepathstyle: true

-compactor:
-  working_directory: /loki/compactor
-  #shared_store: s3
+#compactor:
+# working_directory: /loki/compactor
+# #shared_store: s3
+
+#limits_config:
+#  reject_old_samples: true
+#  reject_old_samples_max_age: 168h

 limits_config:
+  max_cache_freshness_per_query: '10m'
  reject_old_samples: true
-  reject_old_samples_max_age: 168h
+  reject_old_samples_max_age: 30m
+  ingestion_rate_mb: 10
+  ingestion_burst_size_mb: 20
+  # parallelize queries in 15min intervals
+  split_queries_by_interval: 15m
+  volume_enabled: true
+
+

 #chunk_store_config:
 #  max_look_back_period: 0s
@ -143,6 +155,7 @@ table_manager:
  retention_deletes_enabled: true
  retention_period: 336h  # 14 jours

+
 ruler:
  storage:
    type: local
@ -154,3 +167,24 @@ ruler:
    kvstore:
      store: memberlist
  enable_api: true
+
+query_range:
+  # make queries more cache-able by aligning them with their step intervals
+  align_queries_with_step: true
+  max_retries: 5
+  parallelise_shardable_queries: true
+  cache_results: true
+
+frontend:
+  log_queries_longer_than: 5s
+  compress_responses: true
+  max_outstanding_per_tenant: 2048
+
+query_scheduler:
+  max_outstanding_requests_per_tenant: 1024
+
+querier:
+  query_ingesters_within: 2h
+
+compactor:
+  working_directory: /tmp/compactor
--- a/docker-compose.yaml
+++ b/docker-compose.yaml
@ -24,13 +24,13 @@ services:
      - loki

  #------- container with tools dns (nslookup, tcptraceroute, etc). Use just for debug
-#  dns-tools:
-#    image: jonlabelle/network-tools
-#    command: tail -f /dev/null
-#    stdin_open: true
-#    tty: true
-#    networks:
-#      - loki
+  dns-tools:
+    image: jonlabelle/network-tools
+    command: tail -f /dev/null
+    stdin_open: true
+    tty: true
+    networks:
+      - loki

  #-------proxy (nginx)
  proxy:
@ -134,33 +134,57 @@ services:
    restart: unless-stopped
    networks:
      - loki
+    deploy:
+      resources:
+        limits:
+          cpus: '0.50'
+          memory: 256M
+        reservations:
+          cpus: '0.25'
+          memory: 128M

  #------- promtail
  promtail:
    image: grafana/promtail:latest
    volumes:
      - ${PROMTAIL_CONF_DIR}/config.yml:/etc/promtail/config.yml
-      - ${LOKI_GEN_LOG_DIR}:/tmp/
+      - ${LOKI_GEN_LOG_DIR}:/tmp
      - /var/log/:/var/log/
    restart: unless-stopped
    networks:
      - loki
+    deploy:
+      resources:
+        limits:
+          cpus: '0.50'
+          memory: 256M
+        reservations:
+          cpus: '0.25'
+          memory: 128M

  #------- flog generated log
  log-generator:
-    image: mingrammer/flog
-    #image: gmouchet/flog
+    image: ${FLOG_DOCKER_IMG}
    command:
      - --loop
      - --format=json
-      - --number=10 # number of log lines to generate per second
-      - --delay=100ms # delay between log lines
+      - --number=5 # number of log lines to generate per second (delay=1000ms)
+      - --delay=1000ms # delay between log lines
      #- --output=/var/log/generated-logs.txt
-      - --output=/tmp/loki-gen-log/generated-logs.txt
+      - --output=/tmp/generated-logs.txt
      - --overwrite
      - --type=log
+      #- --bytes=50000
    volumes:
      - ${LOKI_GEN_LOG_DIR}:/tmp/
+    deploy:
+      resources:
+        limits:
+          cpus: '0.50'
+          memory: 256M
+        reservations:
+          cpus: '0.25'
+          memory: 128M
 #volumes:
 #  grafana-storage:
 #  prometheus:
--- a/manage.sh
+++ b/manage.sh
@ -0,0 +1,113 @@
+#!/bin/bash
+############################################################
+# Decription: Script to manage stack docker
+#             
+# Author: Gilles Mouchet (gilles.mouchet@gmail.com)
+# Creation Date: 12-06-2025
+# Version: 1.0
+# Install: If applicable - Install instruction or 
+#                          see README.md
+# Usage: If applicable - usage or see README.md
+# Changelog:
+#    V1.0 - dd-Mmm-2025 - GMo
+#      Added
+#      - Creation of script from scratch
+#
+############################################################ 
+
+#-----------------------------------------------------------------
+# Doesn't change anything from here
+#-----------------------------------------------------------------
+version="v1.0-rc1"
+
+progName="./$(/bin/basename $0)"
+
+containers=(minio grafana proxy titi)
+#-----------------------------------------------------------------
+# Functions
+#-----------------------------------------------------------------
+# display help
+function print_usage {
+# print
+/bin/cat << EOF
+
+Usage: $progName [options]
+
+Options:
+   create                   Create container
+   start                    Start stack '${COMPOSE_PROJECT_NAME}'
+   stop                     Stop stack '${COMPOSE_PROJECT_NAME}'
+   down                     Stop stack '${COMPOSE_PROJECT_NAME}' and delete containers
+   restart                  Restart stack '${COMPOSE_PROJECT_NAME}' 
+   version,-v,--version     Display script version
+   help,-h,--help           Display this help
+
+Examples:
+   To start stack '${COMPOSE_PROJECT_NAME}'
+       $progName start
+   To stop stack '${COMPOSE_PROJECT_NAME}'
+       $progName stop
+
+   To create a new stack
+       $progName create
+       $progName start
+EOF
+}
+# read .env if exist
+if [ ! -f .env ]; then
+	echo "file '.env' doesn't exist!"
+    echo "See README.md"
+    exit
+fi
+. .env
+
+#if [ ! -z "$1" ]; then
+#print_usage
+#         exit 0
+#fi         
+
+   case "$1" in
+      create)
+          docker compose create        
+      ;;
+       start)
+         for cont_item in "${containers[@]}"; do
+             echo "Start container $cont_item"
+             docker compose up -d $cont_item --remove-orphans 
+             echo $?
+         done
+      ;;
+      down)
+          docker compose down
+      ;;
+      -h|--help|help)
+          print_usage
+          exit 0
+      ;;
+      -v|--version|version)
+          echo $version
+      ;;
+      *)
+         #echo "${progName}: invalid option -- '$1'!"
+         #echo -e "Try '$progName help' for more information.\n"
+          print_usage
+          exit 0
+      ;;
+   esac
+   
+#!/bin/bash
+
+# Couleurs
+VERT="\e[32m"
+ROUGE="\e[31m"
+NORMAL="\e[0m"
+
+## Message initial
+#echo "Lancement du conteneur..."
+#
+## Exécution de la commande
+#if docker compose start toto > /dev/null 2>&1; then
+    #echo -e "Lancement du conteneur : ${VERT}OK${NORMAL}"
+#else
+    #echo -e "Lancement du conteneur : ${ROUGE}notOK${NORMAL}"
+#fi
--- a/preprare.sh
+++ b/preprare.sh
@ -1,5 +1,8 @@
 #!/bin/bash

+containers=(minio grafana proxy)
+containers_no_start=(dns-tools minio-cli)
+
 if [ ! -f .env ]; then
 	echo "file '.env' doesn't exist!"
    echo "See README.md"
@ -29,7 +32,7 @@ COMPOSE_PROJECT_NAME = ${COMPOSE_PROJECT_NAME}
    LOKI_GEN_LOG_DIR = ${LOKI_GEN_LOG_DIR}
   PROMTAIL_CONF_DIR = ${PROMTAIL_CONF_DIR}
 EOF
-
+echo "create folders"
 sudo mkdir -p ${PRX_CERTS_DIR}
 sudo mkdir -p ${PRX_NGINX_CONF_DIR}
 sudo mkdir -p ${MINIO_DATA_ROOT_DIR}
@ -41,10 +44,22 @@ sudo mkdir -p ${LOKI_CONF_DIR}
 sudo mkdir -p ${LOKI_DATA_DIR}
 sudo mkdir -p ${LOKI_GEN_LOG_DIR}
 sudo mkdir -p ${PROMTAIL_CONF_DIR}
+echo "down"
 docker compose down
+echo "copy config file"
 sudo cp ./config/default.conf ${PRX_NGINX_CONF_DIR}/.
 sudo cp ./config/prometheus.yml ${PROM_CONF_DIR}/. 
 sudo cp ./config/local-config.yaml ${LOKI_CONF_DIR}/. 
 sudo cp ./config/config.yml ${PROMTAIL_CONF_DIR}/.
-docker compose up -d --remove-orphans
+#docker compose up -d --remove-orphans
+echo "create container"
+docker compose create

+for cont_item in "${containers[@]}"; do
+    echo "Start container $cont_item"
+    docker compose up -d $cont_item --remove-orphans
+done
+
+
+#docker compose stop minio-client
+#docker compose stop log-generator