prometheus安装和使用记录
Getting started | PrometheusConfiguration | PrometheusDownload | PrometheusDownload Grafana | Grafana Labs # prometheus<br>mkdir -m=777 -p /data/{download,app_logs,app/prometheus}
cd /data/download
wget https://github.com/prometheus/prometheus/releases/download/v2.45.0-rc.0/prometheus-2.45.0-rc.0.linux-amd64.tar.gz
tar xvfz prometheus-*.tar.gz<br>ln -s /data/download/prometheus-2.45.0-rc.0.linux-amd64/prometheus /usr/bin/prometheus<br>cp /data/download/prometheus-2.45.0-rc.0.linux-amd64/prometheus.yml /data/app/prometheus/prometheus.yml <br>prometheus --config.file=/data/app/prometheus/prometheus.yml --web.listen-address=:9090 --web.enable-lifecycle --storage.tsdb.path=/data/app/prometheus/data >>/data/app_logs/prometheus.log 2>&1 &<br><br># node_exporter 在需要监控的服务器里安装<br>mkdir -m=777 -p /data/{download,app_logs,app/prometheus}<br>cd /data/download<br>wget https://github.com/prometheus/node_exporter/releases/download/v1.6.0/node_exporter-1.6.0.linux-amd64.tar.gz<br><em>tar xvfz node_exporter*<br></em>ln -s /data/download/node_exporter-1.6.0.linux-amd64/node_exporter /usr/bin/node_exporter<br># 启动node_exporter,服务器暴露的端口是8080，同时服务器里有其他服务占用了8080端口，可以使用nginx将node_exporter获取指标的api暴露出去<br># location /metrics {<br># proxy_pass http://127.0.0.1:9000/metrics;<br># }<br>node_exporter --web.listen-address 127.0.0.1:9000 >>/data/app_logs/node_exporter.log 2>&1 &<br><em><em><em># 添加node_exporter之后，需要更新prometheus.xml添加targets，然后运行：curl -X PUT http://server_address:port/-/reload重新加载配置文件<br><br># </em></em></em><em>alert_manager可以和prometheus安装到同一台服务器<br></em>cd /data/download<br>wget https://github.com/prometheus/alertmanager/releases/download/v0.25.0/alertmanager-0.25.0.linux-amd64.tar.gz<br>tar xvfz alertmanager*<br>ln -s /data/download/alertmanager-0.25.0.linux-amd64/alertmanager /usr/bin/alertmanager<br>cp /data/download/alertmanager-0.25.0.linux-amd64/alertmanager.yml /data/app/prometheus/alertmanager.yml<br>alertmanager --config.file=/data/app/prometheus/alertmanager.yml --web.listen-address 127.0.0.1:9001 >>/data/app_logs/node_exporter.log 2>&1 &<br># 将alert_manager的地址添加到prometheus.yml里的alertmanagers的targets里，然后<em>运行：curl -X PUT http://server_address:port/-/reload重新加载配置文件</em>测试报警邮件功能：设置如果安装exporter的服务器内存占用率超过50%或者tcp timewait超过10的时候就发邮件（在实际工作中需要设置一个合适的条件）：
prometheus.yml里添加rule_files的路径：
# my global config
global:
scrape_interval: 15s # Set the scrape interval to every 15 seconds. Default is every 1 minute.
evaluation_interval: 15s # Evaluate rules every 15 seconds. The default is every 1 minute.
# scrape_timeout is set to the global default (10s).
# Alertmanager configuration
alerting:
alertmanagers:
- static_configs:
- targets:
- 127.0.0.1:9001
# Load rules once and periodically evaluate them according to the global 'evaluation_interval'.
rule_files:
# - "first_rules.yml"
# - "second_rules.yml"
- "/data/app/prometheus/alert.rules.yml"
# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
# The job name is added as a label `job=<job_name>` to any timeseries scraped from this config.
- job_name: "prometheus"
# metrics_path defaults to '/metrics'
# scheme defaults to 'http'.
scrape_interval: 5s
static_configs:
- targets: ["node1_ip:8080"]
- targets: ["node2_ip:8080"]
labels:
groups: 'container'alert.rules.yml里添加具体的rule，node_socket_TCP_tw这些具体的指标通过http://node_exporter_ip:port/metrics可以获取到
groups:
- name: tcp-alert-group
rules:
- alert: TcpTimeWait
expr: node_sockstat_TCP_tw > 10
for: 10m
labels:
severity: warning
annotations:
summary: tcp time wait more than 10
description: please check node_sockstat_TCP_tw metric
- alert: MemoryUse
expr: (node_memory_MemTotal_bytes-node_memory_MemFree_bytes-node_memory_Buffers_bytes-node_memory_Cached_bytes)/node_memory_MemTotal_bytes > 0.5
for: 10m
labels:
severity: warning
annotations:
summary: memory use more than 50% for 10 min
description: please check memory usealertmanager.yml里配置告警邮件的信息：
global:
resolve_timeout: 5m
smtp_smarthost: your_smpt_host:port
smtp_from: alertmanager@your_email_domain
smtp_require_tls: false
route:
group_by: ['alertname']
group_wait: 30s
group_interval: 5m
repeat_interval: 10m
receiver: 'email'
receivers:
- name: 'email'
email_configs:
- to: 'receiver_email'
send_resolved: trueyml文件一旦更新，需要重新加载配置：curl -X PUT http://server_address:port/-/reload
在Prometheus的界面可以看到添加的alert：
当alert的条件满足后，alertmanager就会发邮件
grafana的安装和启动：
# grafana可以和prometheus里安装到同一台服务器
yum install -y https://dl.grafana.com/enterprise/release/grafana-enterprise-10.0.0-1.x86_64.rpm
# grafana默认启动的端口号是3000，如果服务器没有暴露3000端口的话，需要修改grafana的配置文件
sed -i 's/3000/8080/g' /usr/share/grafana/conf/defaults.ini
grafana server >> /data/app_logs/grafana.log 2>&1 &
# grafana数据保存地址：/var/lib/grafana.dbgrafana启动之后就可以在浏览器上打开对应的地址，初次登录用户名和密码：admin/admin
Data sources里添加prometheus，grafana和prometheus启动在同一台服务器里的话，地址就可以用localhost
添加dashboard，在Explore里可以查询指标并且添加到dashboard
cpu使用率：avg(1-irate(node_cpu_seconds_total{mode="idle"})) by(instance)
内存使用率：(node_memory_MemTotal_bytes-node_memory_MemFree_bytes-node_memory_Buffers_bytes-node_memory_Cached_bytes)/node_memory_MemTotal_bytes
tcp连接数：node_sockstat_TCP_alloc
dashboard：
注意点：
1.prometheus启动的时候添加--web.enable-lifecycle才允许通过调用/-/reload接口重新加载配置文件<br>2.prometheus启动的时候指定一个固定的数据存放位置--storage.tsdb.path=/data/app/prometheus/data，如果数据存放位置不一致，启动后查不到历史数据，历史数据做备份的话，prometheus启动的服务器还可以变更<br>3.grafana的数据保存地址：/var/lib/grafana.db，定期做备份，服务器发生系统错误无法使用的时候，在新的服务器里同步/var/lib/grafana.db文件之后，启动grafana之前的配置不会丢失
来源:https://www.cnblogs.com/huizit1/p/17492003.html
免责声明：由于采集信息均来自互联网，如果侵犯了您的权益，请联系我们【E-Mail:cb@itdo.tech】 我们会及时删除侵权内容，谢谢合作！
页:
[1]