一、VictoriaMetrics安装

1. 单节点部署

bash
# 下载VictoriaMetrics
wget https://github.com/VictoriaMetrics/VictoriaMetrics/releases/download/v1.91.3/victoria-metrics-linux-amd64-v1.91.3.tar.gz
tar xzf victoria-metrics-linux-amd64-v1.91.3.tar.gz

# 创建存储目录
mkdir -p /data/victoria-metrics

# 启动服务
./victoria-metrics \
-storageDataPath=/data/victoria-metrics \
-retentionPeriod=3 \
-httpListenAddr=:8428

2. 集群部署配置

yaml
# docker-compose.yml
version:'3'
services:
  vmstorage:
    image: victoriametrics/vmstorage
    command:
-"--storageDataPath=/storage"
    volumes:
- storage-data:/storage

  vmselect:
    image: victoriametrics/vmselect
    depends_on:
- vmstorage
    command:
-"--storageNode=vmstorage:8400"

  vminsert:
    image: victoriametrics/vminsert
    depends_on:
- vmstorage
    command:
-"--storageNode=vmstorage:8400"

二、Grafana配置

1. Grafana安装

bash
# 添加Grafana源
wget -q -O - https://packages.grafana.com/gpg.key | apt-key add -
echo "deb https://packages.grafana.com/oss/deb stable main"| tee /etc/apt/sources.list.d/grafana.list

# 安装Grafana
apt update
apt install -y grafana

# 启动服务
systemctl enable grafana-server
systemctl start grafana-server

2. 数据源配置

json
{
"name":"VictoriaMetrics",
"type":"prometheus",
"url":"http://localhost:8428",
"access":"proxy",
"basicAuth":false,
"isDefault":true
}

三、数据采集配置

1. vmagent配置

yaml
# vmagent.yml
global:
  scrape_interval:15s
  evaluation_interval:15s

scrape_configs:
- job_name:'node'
    static_configs:
- targets:['localhost:9100']

- job_name:'mysql'
    static_configs:
- targets:['localhost:9104']

2. 服务发现配置

yaml
# service-discovery.yml
scrape_configs:
- job_name:'kubernetes-pods'
    kubernetes_sd_configs:
- role: pod
    relabel_configs:
- source_labels:[__meta_kubernetes_pod_annotation_prometheus_io_scrape]
        action: keep
        regex:true

四、告警配置

1. 告警规则

yaml
# rules.yml
groups:
- name: server_alerts
  rules:
- alert:HighCPUUsage
    expr:100-(avg by(instance)(irate(node_cpu_seconds_total{mode="idle"}[5m]))*100)>80
for:5m
    labels:
      severity: warning
    annotations:
      summary:"High CPU usage on {{$labels.instance}}"

2. 告警通知

json
// Grafana告警通道配置
{
"name":"webhook",
"type":"webhook",
"settings":{
"url":"http://alert-webhook:8080/notify",
"httpMethod":"POST"
}
}

五、Dashboard配置

1. 系统监控面板

json
{
"dashboard":{
"panels":[
{
"title":"CPU Usage",
"type":"graph",
"datasource":"VictoriaMetrics",
"targets":[
{
"expr":"100 - (avg by (instance) (irate(node_cpu_seconds_total{mode=\"idle\"}[5m])) * 100)"
}
]
}
]
}
}

2. 性能监控面板

json
{
"dashboard":{
"panels":[
{
"title":"Memory Usage",
"type":"gauge",
"datasource":"VictoriaMetrics",
"targets":[
{
"expr":"node_memory_MemUsed_bytes / node_memory_MemTotal_bytes * 100"
}
]
}
]
}
}

六、性能优化

1. VictoriaMetrics优化

bash
# VictoriaMetrics启动参数优化
./victoria-metrics \
-storageDataPath=/data/victoria-metrics \
-memory.allowedPercent=60 \
-search.maxUniqueTimeseries=1000000 \
-search.maxQueryDuration=30s

2. Grafana性能调优

ini
# grafana.ini
[server]
http_port =3000
root_url = https://grafana.example.com

[database]
type = mysql
host = localhost:3306
name = grafana
user = grafana
password = password

[session]
provider = redis
provider_config = addr=127.0.0.1:6379