prometheus+grafana监控以及企业微信告警(单机二进制部署)
一.下载部署包,更改其中两个包名称,放到/data下
1.安装包以及解压步骤
grafana-enterprise-8.4.4-1.x86_64.rpm
prometheus-2.34.0.tar.gz
node_exporter-1.3.1.linux-amd64.tar.gz
alertmanager-0.23.0.linux-amd64.tar.gz
tar -xzvf prometheus-2.34.0.tar.gz
mv prometheus-2.34.0.linux-amd64 /data/prometheus
tar -xzvf node_exporter-1.3.1.linux-amd64.tar.gz
mv node_exporter-1.3.1.linux-amd64 /data/node_exporter
tar -xzvf alertmanager-0.23.0.linux-amd64.tar.gz
mv alertmanager-0.23.0.linux-amd64 /data/alertmanager
yum install ./grafana-enterprise-8.4.4-1.x86_64.rpm -y
二、添加各个服务的service
# 普罗米修斯
cat <<EOF | sudo tee /lib/systemd/system/prometheus.service
[Unit]
Description=Prometheus
After=network.target
[Service]
Type=simple
User=root
WorkingDirectory=/data/prometheus
ExecStart=/data/prometheus/prometheus --config.file=/data/prometheus/prometheus.yml
Restart=on-failure
LimitNOFILE=65536
[Install]
WantedBy=multi-user.target
EOF
# node_exporter
cat <<EOF | sudo tee /lib/systemd/system/node_exporter.service
[Unit]
Description=Node Exporter
After=network.target
Wants=network-online.target
[Service]
Type=simple
User=root
ExecStart=/data/node_exporter/node_exporter
Restart=on-failure
LimitNOFILE=65536
[Install]
WantedBy=multi-user.target
EOF
# alertmanager
cat <<EOF | sudo tee /lib/systemd/system/alertmanager.service
[Unit]
Description=alertmanager
After=alertmanager.target
[Service]
ExecStart=/data/alertmanager/alertmanager --config.file=/data/alertmanager/alertmanager.yml
Restart=on-failure
LimitNOFILE=65536
[Install]
WantedBy=multi-user.target
EOF
三.启动
systemctl enable node_exporter
systemctl enable prometheus
systemctl enable grafana-server
systemctl enable alertmanager
systemctl start node_exporter
systemctl start prometheus
systemctl start grafana-server
/usr/share/grafana/bin/grafana-cli plugins install grafana-piechart-panel
四.监控linux主机和windows主机
监控linux
1.在每台被控linux主机安装node_exporter
2.在普罗米yml文件里添加所要监控的主机
监控windows
1.在每台被控windows主机安装wmi_exporter。wmi_exporter下载地址:https://github.com/martinlindhe/wmi_exporter/releases
下载后,双击即可完成安装。完成安装后,会自动创建一个开机自启的服务。
2.在普罗米yml文件里添加配置
五.配置grafana,添加数据源
2.根据需求添加模板(linux模板12633,windows12422)
六.企业微信监控告警
1.在企业微信创建群聊,添加机器人会有webhook地址。
2.在普罗米yml文件里开启alertmanager服务端口和自定义规则。
3.在上述规则目录编写规则yml文件
创建目录:
mkdir /data/rules
进入目录:
cd /data/rules
编写规则文件:
1.vim host_rules.yml
groups:
# 报警组组名称
- name: alters
#报警组规则
rules:
#告警名称,需唯一
- alert: cpu使用率大于75%
#promQL表达式
expr: sum(avg without (cpu)(irate(node_cpu_seconds_total{mode!='idle'}[5m]))) by (instance) > 0.75
#满足此表达式持续时间超过for规定的时间才会触发此报警
for: 1m
labels:
#严重级别
severity: warning
annotations:
#发出的告警标题
summary: "实例 {{ $labels.instance }} CPU 使用率过高"
#发出的告警内容
description: "实例{{ $labels.instance }} CPU 使用率超过 75% (当前值为: {{ $value }})"
- alert: 内存使用率大于90%
expr: (node_memory_MemTotal_bytes - node_memory_MemAvailable_bytes)/node_memory_MemTotal_bytes > 0.90
for: 1m
labels:
severity: warning
annotations:
summary: "实例 {{ $labels.instance }} 内存使用率过高"
description: "实例 {{ $labels.instance }} 内存使用率 90% (当前值为: {{ $value }})"
2.vim node_alters.yml
groups:
- name: example
rules:
- alert: 服务器宕机
expr: avg by (instance) (up{}) == 0
for: 10s
labels:
severity: emergency
annotations:
description: "实例 {{ $labels.instance }}服务器已宕机,请进行检查."
summary: "{{ $labels.instance }} 服务器宕机"
- alert: 磁盘使用率大于80%
expr: 100 - (node_filesystem_free_bytes{mountpoint="/",fstype=~"ext4|xfs"} / node_filesystem_size_bytes{fstype=~"ext4|xfs"} * 100) > 80
for: 2m
labels:
severity: warning
annotations:
description: "{{ $labels.instance }} : {{ $labels.job }} :{{ $labels.mountpoint }} 这个分区使用大于百分之80% (当前值:{{ $value }})"
summary: "Instance {{ $labels.instance }} :{{ $labels.mountpoint }} 分区使用率过高"
4.访问普罗米页面,看是否添加成功
5.在普罗米服务端添加python自动化脚本
(1)安装python3环境
(2)把脚本app.py放在/usr/bin/目录下,脚本内容如下:
cd /usr/bin
vim app.py
import json
import os
from flask import request, Flask
from dingtalkchatbot.chatbot import DingtalkChatbot
def Open(s):
Path = './temp.json'
if os.path.exists(Path):
os.remove(Path)
f = open(Path, 'w')
print(s, file = f)
f.close()
def GetData():
PostData = request.get_data()
Data = json.loads(PostData)
JsonData = json.dumps(Data, ensure_ascii=False, indent=4)
return Data
app = Flask(__name__)
app.config['JSON_AS_ASCII'] = False
@app.route('/webhook/test/', methods=['POST'])
def IssueCreate():
# Open(GetData())
status = GetData()['status']
alertname = GetData()['alerts'][0]['labels']['alertname']
serverity = GetData()['alerts'][0]['labels']['severity']
instance = GetData()['alerts'][0]['labels']['instance']
start_time = GetData()['alerts'][0]['startsAt']
message = str(
'##' + 'Prometheus告警:%s' + '##' + '\n'
'状态:%s' %status + '\n'
'告警信息:%s' %alertname + '\n'
'告警级别:%s' %serverity + '\n'
'IP地址:%s' %instance + '\n'
'开始时间:%s' %start_time + '\n'
)
print(message)
send_message(message)
return "OK", 200
def send_message(message):
webhook = '机器人地址'
xiaoding = DingtalkChatbot(webhook)
xiaoding.send_text(msg=message, is_at_all=True)
if __name__ == '__main__':
app.run(debug = False, host = '0.0.0.0', port = 8888)
(3)配置脚本自启动文件
vim /usr/lib/systemd/system/app.service
[Unit]
Description=QR Code Reader Service
After=multi-user.target
Conflicts=getty@tty1.service
[Service]
Type=simple
ExecStart=/usr/bin/python3 /usr/bin/app.py
StandardInput=tty-force
[Install]
WantedBy=multi-user.target
(4)在alertmanager的yml文件修改
global:
resolve_timeout: 5m 处理超时时间
route:
group_by: ['alertname']
group_wait: 10s 等待时间
group_interval: 10s 相同的Gourp之间发送告警通知的时间间隔
repeat_interval: 5m 重复报警的间隔时长
receiver: 'web.hook'
receivers:
- name: 'web.hook'
webhook_configs:
- url: 'http://本机ip:8888/webhook/test/'
inhibit_rules:
- source_match:
severity: 'critical'
target_match:
severity: 'warning'
equal: ['alertname', 'dev', 'instance']
(5)重启服务alertmanager。看企业微信是否可以收到告警信息。