Grafana 可觀測性平台完整指南

Fri, 03 Apr 2026 00:00:00 +0000

專案簡介

Grafana 是最受歡迎的開源可觀測性平台，提供美觀的儀表板來視覺化監控資料。支援 Prometheus、InfluxDB、Elasticsearch 等數十種資料源。

GitHub Stars: 72K+

主要功能

儀表板 - 豐富的視覺化元件
多資料源 - 整合 100+ 資料來源
告警系統 - 多管道通知
日誌查詢 - Loki 整合
分散式追蹤 - Tempo 整合

快速部署

Docker

1
2
3
4
5


docker run -d \
 -p 3000:3000 \
 --name grafana \
 -v grafana-storage:/var/lib/grafana \
 grafana/grafana-oss

Docker Compose（完整堆疊）

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31


version: '3.8'
services:
 grafana:
 image: grafana/grafana-oss
 ports:
 - "3000:3000"
 environment:
 - GF_SECURITY_ADMIN_PASSWORD=admin
 volumes:
 - grafana-data:/var/lib/grafana

 prometheus:
 image: prom/prometheus
 ports:
 - "9090:9090"
 volumes:
 - ./prometheus.yml:/etc/prometheus/prometheus.yml

 loki:
 image: grafana/loki
 ports:
 - "3100:3100"

 promtail:
 image: grafana/promtail
 volumes:
 - /var/log:/var/log
 - ./promtail-config.yml:/etc/promtail/config.yml

volumes:
 grafana-data:

訪問 http://localhost:3000（admin/admin）

資料源設定

Prometheus

Configuration → Data Sources → Add
選擇 Prometheus
URL: http://prometheus:9090
Save & Test

Loki（日誌）

Configuration → Data Sources → Add
選擇 Loki
URL: http://loki:3100

MySQL

 1
 2
 3
 4
 5
 6
 7
 8
 9
10


# datasources/mysql.yaml
apiVersion: 1
datasources:
 - name: MySQL
 type: mysql
 url: mysql:3306
 database: mydb
 user: grafana
 secureJsonData:
 password: secret

建立儀表板

Panel 類型

類型	用途
Time series	時間序列資料
Stat	單一數值
Gauge	量表顯示
Bar chart	長條圖
Table	表格資料
Heatmap	熱力圖
Logs	日誌顯示

PromQL 範例

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11


# CPU 使用率
100 - (avg by(instance) (rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100)

# 記憶體使用率
(1 - (node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes)) * 100

# HTTP 請求率
sum(rate(http_requests_total[5m])) by (status)

# P95 延遲
histogram_quantile(0.95, sum(rate(http_request_duration_seconds_bucket[5m])) by (le))

LogQL 範例（Loki）

1
2
3
4
5
6
7
8


# 查詢特定應用日誌
{app="nginx"} |= "error"

# JSON 解析
{app="api"} | json | response_code >= 500

# 統計錯誤數
sum(rate({app="api"} |= "error" [5m])) by (level)

告警設定

告警規則

Alerting → Alert Rules → Create
設定條件

1
2
3
4
5
6
7
8


# 範例：CPU 高使用率告警
alert: HighCPUUsage
expr: 100 - (avg by(instance) (rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100) > 80
for: 5m
labels:
 severity: warning
annotations:
 summary: "High CPU usage on {{ $labels.instance }}"

通知管道

支援多種通知方式：

Email
Slack
PagerDuty
Webhook
Microsoft Teams
Discord

Slack 設定

Alerting → Contact points → New
選擇 Slack
輸入 Webhook URL

儀表板即程式碼

JSON 匯出

Dashboard Settings → JSON Model → Copy

Provisioning

1
2
3
4
5
6
7
8


# provisioning/dashboards/default.yaml
apiVersion: 1
providers:
 - name: 'default'
 folder: ''
 type: file
 options:
 path: /var/lib/grafana/dashboards

Terraform

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12


resource "grafana_dashboard" "metrics" {
 config_json = file("dashboard.json")
}

resource "grafana_alert_rule" "cpu" {
 name = "High CPU"
 folder_id = grafana_folder.alerts.id
 rule_group {
 name = "cpu_alerts"
 interval = "1m"
 }
}

Grafana Loki

Promtail 設定

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24


# promtail-config.yml
server:
 http_listen_port: 9080

positions:
 filename: /tmp/positions.yaml

clients:
 - url: http://loki:3100/loki/api/v1/push

scrape_configs:
 - job_name: containers
 static_configs:
 - targets:
 - localhost
 labels:
 job: containerlogs
 __path__: /var/lib/docker/containers/*/*.log
 pipeline_stages:
 - json:
 expressions:
 log: log
 stream: stream
 time: time

查詢日誌

1
2
3
4
5


# 最近錯誤
{job="containerlogs"} |= "error" | limit 100

# 按級別統計
sum by (level) (count_over_time({app="api"}[1h]))

效能優化

資料保留

1
2
3
4
5
6
7
8
9


# prometheus.yml
global:
 scrape_interval: 15s
 evaluation_interval: 15s

storage:
 tsdb:
 retention.time: 15d
 retention.size: 50GB

快取設定

1
2
3
4


# grafana.ini
[caching]
enabled = true
ttl = 300

安全設定

認證

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16


# grafana.ini
[auth]
disable_login_form = false

[auth.ldap]
enabled = true
config_file = /etc/grafana/ldap.toml

[auth.generic_oauth]
enabled = true
name = Keycloak
client_id = grafana
client_secret = secret
auth_url = https://keycloak/auth
token_url = https://keycloak/token
api_url = https://keycloak/userinfo

HTTPS

1
2
3
4


[server]
protocol = https
cert_file = /etc/grafana/cert.pem
cert_key = /etc/grafana/key.pem

grafana on Astroicers Blog