⑥ prometheus+grafana监控Ceph集群
2020-05-07 本文已影响0人
Linux丶晨星
在新建的主机
192.168.0.5
上
一、安装grafana
1、清华源下载
[root@ceph-grafana-prometheus ~]# wget https://mirrors.tuna.tsinghua.edu.cn/grafana/yum/rpm/grafana-6.7.3-1.x86_64.rpm
2.安装grafana
rpm -ivh grafana-6.7.3-1.x86_64.rpm
或
yum localinstall grafana-6.7.3-1.x86_64.rpm
3.启动grafana并设为开机自启
systemctl start grafana-server.service
systemctl enable grafana-server.service
#grafana的默认端口为3000
二、安装prometheus
#1、下载安装包,下载地址
wget https://github.com/prometheus/prometheus/releases/download/v2.18.0/prometheus-2.18.0.linux-amd64.tar.gz
#2、解压压缩包
tar xf prometheus-2.14.0.linux-amd64.tar.gz
#3、将解压后的目录改名
mv prometheus-2.18.0.linux-amd64 /opt/prometheus
#4、查看prometheus版本
cd /opt/prometheus/
./prometheus --version
prometheus, version 2.18.0
#5、配置系统服务启动
vim /etc/systemd/system/prometheus.service
[Unit]
Description=Prometheus Monitoring System
Documentation=Prometheus Monitoring System
[Service]
ExecStart=/opt/prometheus/prometheus \
--config.file /opt/prometheus/prometheus.yml \
--web.listen-address=:9090
[Install]
WantedBy=multi-user.target
#6、加载系统服务
systemctl daemon-reload
#7、启动服务和添加开机自启动
systemctl restart prometheus
systemctl enable prometheus
#8、访问
http://192.168.0.5:9090
三、ceph mgr prometheus插件配置
在cephnode01上配置
#启动prometheus模块
ceph mgr module enable prometheus
#检查端口
netstat -nltp | grep mgr
ceph -s (mgr在cephnode03上)
#测试返回值
curl 192.168.0.8:9283/metrics
#访问
http://192.168.0.8:9283/metrics
四、配置prometheus
1、在 scrape_configs: 配置项下添加
#每个job_name要分开
vim prometheus.yml
- job_name: 'ceph_cluster'
honor_labels: true
scrape_interval: 5s
static_configs:
- targets: ['192.168.0.8:9283']
labels:
instance: ceph
2、重启prometheus服务
systemctl restart prometheus
3、检查prometheus服务器中是否添加成功
浏览器-》 http://x.x.x.x:9090 -》status -》Targets
五、配置grafana
1、浏览器登录 grafana 管理界面
http://192.168.0.5:3000/
2、添加data sources,点击configuration--》data sources
3、添加dashboard,点击HOME--》find dashboard on grafana.com
4、搜索ceph的dashboard
5、点击HOME--》Import dashboard, 选择合适的dashboard,记录编号
6、检测成功
7、也可以检测Cephfs文件系统