Skip to main content ホーム クリエイター johnnuwan eva_core prometheus-grafana
prometheus-grafana Monitoring & observabilité — Prometheus pour la collecte de métriques, Alertmanager pour les alertes, Grafana pour les dashboards, Loki pour les logs, bonnes pratiques SRE
インストールへ移動 Skills Marketplace コミュニティが作成したAIスキルを発見・探索
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/JohnNuwan/EVA_CORE --skill prometheus-grafanaコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか?SkillsMP が現在取得できるファイルをダウンロードできます。
Zipをダウンロード ダウンロード中... name prometheus-grafana description Monitoring & observabilité — Prometheus pour la collecte de métriques, Alertmanager pour les alertes, Grafana pour les dashboards, Loki pour les logs, bonnes pratiques SRE version 1.0.0 author EVA license Privée EVA category mlops metadata {"EVA":{"tags":["prometheus","grafana","monitoring","alerting","metrics","dashboards","loki","alertmanager","observabilité","sre"],"related_skills":["kubernetes-avance","docker-avance","ansible-automation","ci-cd-pipelines"]}}
Prometheus & Grafana — Monitoring & Observabilité
Vue d'ensemble
Prometheus (collecte de métriques) + Grafana (visualisation) + Alertmanager (notifications) forment la stack de monitoring standard dans l'écosystème cloud-native. Cette compétence couvre l'installation, la configuration, l'écriture de règles d'alerte, la création de dashboards, l'intégration Loki pour les logs, et les patterns SRE (Service Level Objectives).
Quand l'utiliser
Monitorer les métriques d'infrastructure (CPU, RAM, disque, réseau)
Surveiller des applications custom avec endpoints Prometheus
Créer des dashboards Grafana pour visualiser les KPIs
Mettre en place des alertes (Slack, PagerDuty, Telegram, email)
Centraliser les logs avec Loki et les corréler aux métriques
Définir des SLO/SLI/SLA et suivre les burn rates
1. Architecture Stack
Applications
│ (métriques HTTP /metrics)
▼
Prometheus Server ──► Alertmanager ──► Slack / Telegram / PagerDuty
│ ▲
│ scrape │
▼ │
TSDB (local) │
│ │
▼ (PromQL) │
Grafana ◄────────────────┘
│
▼
Loki (logs) ← Promtail (agents log)
2. Installation (Docker Compose)
version: "3.8"
networks:
monitoring:
driver: bridge
volumes:
prometheus_data:
grafana_data:
loki_data:
services:
prometheus:
image: prom/prometheus:v2.55
command:
- '--config.file=/etc/prometheus/prometheus.yml'
- '--storage.tsdb.path=/prometheus'
- '--storage.tsdb.retention.time=30d'
- '--web.console.libraries=/etc/prometheus/console_libraries'
- '--web.console.templates=/etc/prometheus/consoles'
- '--web.enable-lifecycle'
volumes:
- ./prometheus.yml:/etc/prometheus/prometheus.yml
- ./rules/:/etc/prometheus/rules/
- prometheus_data:/prometheus
ports:
- "9090:9090"
networks:
- monitoring
alertmanager:
image: prom/alertmanager:v0.27
command:
- '--config.file=/etc/alertmanager/alertmanager.yml'
- '--storage.path=/alertmanager'
3. Configuration Prometheus
global:
scrape_interval: 15s
evaluation_interval: 15s
external_labels:
cluster: 'eva'
env: 'production'
alerting:
alertmanagers:
- static_configs:
- targets: ['alertmanager:9093' ]
rule_files:
- '/etc/prometheus/rules/*.yml'
scrape_configs:
- job_name: 'prometheus'
static_configs:
- targets: ['localhost:9090' ]
- job_name: 'node'
static_configs:
- targets:
- 'node-exporter:9100'
- '192.168.1.10:9100'
- job_name: 'docker'
static_configs:
- targets: ['host.docker.internal:9323' ]
- job_name: 'cadvisor'
static_configs:
- targets: [ ]
[ ]
[ ]
[ ]
4. Règles d'Alerte
groups:
- name: infrastructure
interval: 30s
rules:
- alert: NodeDown
expr: up{job="node"} == 0
for: 5m
labels:
severity: critical
annotations:
summary: "Nœud {{ $labels.instance }} down"
description: "Le nœud {{ $labels.instance }} est injoignable depuis 5 minutes."
- alert: HighCPUUsage
expr: 100 - (avg by(instance)(rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100 ) > 90
for: 10m
labels:
severity: warning
annotations:
summary: "CPU > 90% sur {{ $labels.instance }} "
description: "CPU à {{ $value | humanizePercentage }} depuis 10 min."
5. Alertmanager
global:
resolve_timeout: 5m
slack_api_url: 'https://hooks.slack.com/services/T00/B00/xxx'
telegram_api_url: 'https://api.telegram.org'
route:
receiver: 'default'
group_by: [alertname , cluster , severity ]
group_wait: 30s
group_interval: 5m
repeat_interval: 4h
routes:
- match:
severity: critical
receiver: 'critical'
repeat_interval: 1h
- match:
severity: page
receiver: 'pagerduty'
receivers:
- name: 'default'
telegram_configs:
- bot_token: ${TELEGRAM_BOT_TOKEN}
chat_id: ${TELEGRAM_CHAT_ID}
parse_mode: HTML
message: |
<b>{{ .GroupLabels.alertname }}</b>
Sévérité : {{ .CommonLabels.severity }}
{{ range .Alerts }}
• {{ .Annotations.summary }}
{{ end }}
-
6. Dashboards Grafana (Datasources provisionnées)
apiVersion: 1
datasources:
- name: Prometheus
type: prometheus
access: proxy
url: http://prometheus:9090
isDefault: true
jsonData:
timeInterval: "15s"
- name: Loki
type: loki
access: proxy
url: http://loki:3100
- name: PostgreSQL
type: postgres
url: db:5432
database: eva
user: grafana
secureJsonData:
password: ${DB_PASSWORD}
jsonData:
sslmode: disable
Exemple de dashboard JSON (Node Exporter Full)
{
"title" : "Infrastructure - Nœud {{instance}}" ,
"tags" : [ "infrastructure" , "node-exporter" ] ,
"panels" : [
{
"title" : "CPU Usage" ,
"type" : "timeseries" ,
"datasource" : "Prometheus" ,
"targets" : [ {
"expr" : "100 - (avg by(cpu)(rate(node_cpu_seconds_total{mode=\"idle\", instance=~\"$instance\"}[5m])) * 100)" ,
"legendFormat" : "CPU {{cpu}}"
} ]
} ,
{
"title" : "Memory" ,
"type" : "gauge" ,
"targets" : [
7. Loki & Promtail (Logs)
auth_enabled: false
server:
http_listen_port: 3100
grpc_listen_port: 9095
common:
path_prefix: /loki
ring:
kvstore:
store: inmemory
replication_factor: 1
schema_config:
configs:
- from: 2024-01-01
store: tsdb
schema: v13
index:
prefix: index_
period: 24h
object_store: filesystem
storage_config:
tsdb_shipper:
active_index_directory: /loki/index
cache_location: /loki/cache
filesystem:
directory: /loki/chunks
compactor:
working_directory: /loki/compactor
retention_enabled: true
retention_delete_delay: 2h
limits_config:
retention_period: 90d
ingestion_rate_mb: 10
server:
http_listen_port: 9080
grpc_listen_port: 9095
positions:
filename: /tmp/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: docker
pipeline_stages:
- docker: {}
static_configs:
- targets: [localhost ]
labels:
job: docker
__path__: /var/lib/docker/containers/*/*-json.log
8. Pièges Courants
Rétention trop courte : 7 jours par défaut. En production, 30 jours minimum (--storage.tsdb.retention.time=30d).
Trop de métriques cardinales : Des labels comme request_id ou user_email explosent la cardinalité. Éviter les labels uniques.
Sondes Alertmanager sans groupement : 100 alertes par seconde pour un même incident submerge tout le monde. Grouper par alertname, cluster.
Dashboard sans variables : Un dashboard codé en dur pour un serveur ne sert à rien pour 50 serveurs. Ajouter des variables $instance, $job.
Promtail lit tout Docker : Sans filtrage, promtail ingère les logs de TOUS les conteneurs, même ceux non monitorés. Filtrer par label.
Absence de scrape targets : promtool check config prometheus.yml ne vérifie pas la résolution DNS. Tester avec curl http://target:9090/metrics.
9. SLO / SLI / SLA — Patterns SRE
groups:
- name: slo
rules:
- record: job:slo_errors_total:ratio_rate1h
expr: |
sum(rate(http_requests_total{status=~"5.."}[1h]))
/
sum(rate(http_requests_total[1h]))
- record: job:slo_burn_rate
expr: |
job:slo_errors_total:ratio_rate1h / (1 - 0.999)
- alert: SLOWarning
expr: job:slo_burn_rate > 1
for: 1h
labels:
severity: warning
annotations:
summary: "Burn rate SLO > 1x (atteindra le budget en 30j)"
10. Checklist Production
volumes:
-
./alertmanager.yml:/etc/alertmanager/alertmanager.yml
ports:
-
"9093:9093"
networks:
-
monitoring
grafana:
image:
grafana/grafana:11.2
environment:
-
GF_SECURITY_ADMIN_PASSWORD=${GRAFANA_PASSWORD:-admin}
-
GF_INSTALL_PLUGINS=grafana-piechart-panel
-
GF_AUTH_ANONYMOUS_ENABLED=false
volumes:
-
grafana_data:/var/lib/grafana
-
./grafana/dashboards/:/etc/grafana/provisioning/dashboards/
-
./grafana/datasources/:/etc/grafana/provisioning/datasources/
ports:
-
"3000:3000"
networks:
-
monitoring
depends_on:
-
prometheus
node-exporter:
image:
prom/node-exporter:v1.8
command:
-
'--path.rootfs=/host'
-
'--collector.filesystem.ignored-mount-points=^/(sys|proc|dev|host|etc)($$|/)'
pid:
host
volumes:
-
/:/host:ro,rslave
networks:
-
monitoring
loki:
image:
grafana/loki:3.1
command:
-config.file=/etc/loki/loki-config.yaml
volumes:
-
./loki-config.yaml:/etc/loki/loki-config.yaml
-
loki_data:/loki
ports:
-
"3100:3100"
networks:
-
monitoring
promtail:
image:
grafana/promtail:3.1
command:
-config.file=/etc/promtail/promtail-config.yaml
volumes:
-
./promtail-config.yaml:/etc/promtail/promtail-config.yaml
-
/var/log/:/var/log/:ro
-
/var/lib/docker/containers:/var/lib/docker/containers:ro
networks:
-
monitoring
'cadvisor:8080'
-
job_name:
'nginx'
static_configs:
-
targets:
'192.168.1.10:9113'
-
job_name:
'postgres'
static_configs:
-
targets:
'192.168.1.20:9187'
-
job_name:
'api-service'
metrics_path:
/metrics
static_configs:
-
targets:
'192.168.1.10:8000'
-
alert:
HighMemoryUsage
expr:
(1
-
(node_memory_MemAvailable_bytes
/
node_memory_MemTotal_bytes))
*
100
>
90
for:
5m
labels:
severity:
warning
annotations:
summary:
"Mémoire > 90% sur {{ $labels.instance }} "
-
alert:
DiskSpaceCritical
expr:
(node_filesystem_avail_bytes{mountpoint="/"}
/
node_filesystem_size_bytes{mountpoint="/"})
*
100
<
10
for:
5m
labels:
severity:
critical
annotations:
summary:
"Espace disque < 10% sur {{ $labels.instance }} "
description:
"{{ $value | humanizePercentage }} restant sur {{ $labels.mountpoint }} "
-
alert:
HighDiskIO
expr:
rate(node_disk_io_time_seconds_total[5m])
>
0.5
for:
15m
labels:
severity:
warning
annotations:
summary:
"I/O disque élevé sur {{ $labels.device }} "
description:
"Disque occupé à {{ $value | humanizePercentage }} "
-
name:
applications
interval:
15s
rules:
-
alert:
APIHighLatency
expr:
histogram_quantile(0.95,
rate(http_request_duration_seconds_bucket{job="api-service"}[5m]))
>
2
for:
5m
labels:
severity:
warning
annotations:
summary:
"API lente (p95 > 2s)"
description:
"Latence p95 à {{ $value }} s"
-
alert:
APIErrorRate
expr:
rate(http_requests_total{job="api-service",
status=~"5.."}[5m])
/
rate(http_requests_total{job="api-service"}[5m])
*
100
>
5
for:
5m
labels:
severity:
critical
annotations:
summary:
"Taux d'erreur API > 5%"
description:
"{{ $value | humanizePercentage }} d'erreurs 5xx"
-
alert:
ServiceDown
expr:
up{job="api-service"}
==
0
for:
1m
labels:
severity:
critical
annotations:
summary:
"Service {{ $labels.job }} down"
-
name:
sre
interval:
1m
rules:
-
alert:
HighBurnRate
expr:
|
(1 - (rate(http_requests_total{status=~"2..|3.."}[1h]) / rate(http_requests_total[1h]))) * 100 > 0.1
and on()
(1 - (rate(http_requests_total{status=~"2..|3.."}[1h]) / rate(http_requests_total[1h]))) * 100 > 5 * 0.1
for:
1h
labels:
severity:
page
annotations:
summary:
"Burn rate d'erreur > 5x le budget SLO"
name:
'critical'
slack_configs:
-
channel:
'#alerts-critical'
title:
'{{ .GroupLabels.alertname }}'
text:
'{{ range .Alerts }}{{ .Annotations.description }}\n{{ end }}'
color:
'danger'
-
name:
'pagerduty'
pagerduty_configs:
-
routing_key:
${PD_ROUTING_KEY}
severity:
critical
{
"expr"
:
"(1 - (node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes)) * 100"
}
]
,
"options"
:
{
"minViz"
:
0
,
"maxViz"
:
100
,
"thresholds"
:
[
{
"color"
:
"green"
,
"value"
:
null
}
,
{
"color"
:
"orange"
,
"value"
:
80
}
,
{
"color"
:
"red"
,
"value"
:
90
}
]
}
}
,
{
"title"
:
"Disk Usage"
,
"type"
:
"table"
,
"targets"
:
[
{
"expr"
:
"100 - (node_filesystem_avail_bytes{mountpoint!=\"\", instance=~\"$instance\"} / node_filesystem_size_bytes{mountpoint!=\"\"} * 100)"
,
"legendFormat"
:
"{{mountpoint}}"
}
]
}
]
}