Table of contents
Initialize the monitoring host
Monitoring host configuration monitoring monitored host
Initialize the monitoring host
Install go environment
Download the corresponding go version, unzip it, and set environment variables
export GOROOT=/usr/local/go export PATH=$PATH:$GOROOT/bin
Execute go version to ensure that go is installed successfully
Install Prometheus
Download the corresponding installation package Download | Prometheus
After downloading, unzip tar-zxvf prometheus-2.37.8.linux-amd64.tar.gz
start up
nohup ./prometheus --web.listen-address="0.0.0.0:9788" &
Visit your designated port to directly view the monitoring page that comes with Prometheus: such as my 192.168.55.191:9788
Install Grafana
Download the rpm package: https://dl.grafana.com/oss/release/grafana-6.4.3-1.x86_64.rpm
Local installation: yum localinstall grafana-6.4.3-1.x86_64.rpm
Start: systemctl start grafana-server.service
Visit: 192.168.55.191:3000 to open the grafana page, the default username and password are admin, the first login will allow you to change the password
Then click Add Data Source
Click Prometheus
After entering, fill in according to your own configuration, and then save
Select the report you need on the Dashboards page, or go directly to the official website to download and import Dashboards | Grafana Labs
Interlude: I imported a json from the official website, but the following error occurred, the reason is that my grafana is too low for this monitoring version, so I used rpm -Uvh grafana-7.2.1-1.x86_64.rpm to upgrade the rpm (upgrade Remember to stop the previous service), after upgrading, you can view the version grafana-server -v, and then restart: systemctl start grafana-server.service
Initialize the monitored host
Download node_exporter at https://github.com/prometheus/node_exporter/
Unzip and start (nohup ./node_exporter --web.listen-address=:9788 &)
Visit the local startup port verification, for example, I visit the local port 9788
Monitoring host configuration monitoring monitored host
Modify Prometheus configuration
According to the official case of monitoring yourself, continue to add the services to be monitored.
- job_name: "prometheus"
static_configs:
- targets: ["localhost:9788"]
Then restart Prometheus
after class
Gracefully shut down and reload Prometheus : --web.enable-lifecycle Description Enable shutdown and reload via HTTP request. You can then close the program with an HTTP request, for example:
curl -X POST http://localhost:9788/-/quitThe grafana configuration file installed in this way is in /etc/grafana/grafana.ini