prometheus installation and configuration

1, Binary package installation

We can go to Prometheus binary installation package Download page Download the corresponding installation package according to your own operating system. Let's take Amazon Linux 2 as a demonstration.

We will download the latest version v2.17.1 according to the current time.

1.1 download Prometheus Server

wget https://github.com/prometheus/prometheus/releases/download/v2.17.1/prometheus-2.17.1.linux-amd64.tar.gz

tar xf prometheus-2.17.1.linux-amd64.tar.gz
mv prometheus-2.17.1.linux-amd64/prometheus /usr/local/bin/

1.2 configuration startup file

Because it's a production environment, we configure the startup file for it.

vim /etc/systemd/system/prometheus.service

[Unit]
Description=Prometheus Server
After=network.target
Documentation=https://prometheus.io/docs/introduction/overview/

[Service]
Type=simple
WorkingDirectory=/home/data/prometheus/
ExecStart=/usr/local/bin/prometheus \
  --config.file=/etc/prometheus/prometheus.yml \
  --web.read-timeout=5m \
  --web.max-connections=512 \
  --storage.tsdb.retention=15d \
  --storage.tsdb.path=/home/data/prometheus \
  --query.timeout=2m

Restart=on-failure

[Install]
WantedBy=multi-user.target

1.3 configuration file

Prometheus monitors the target node (hereinafter referred to as the "sampling target") by collecting metrics on the HTTP port of the target node. Because Prometheus exposes his data in the same way, he can also collect and check his health.

Let's move the configuration file to the standard directory / etc/prometheus/

mkdir /etc/prometheus
mv prometheus-2.17.1.linux-amd64/prometheus.yml /etc/prometheus

The initial configuration file is relatively simple. As follows, a job is provided to monitor the status of prometheus server.

global:
  scrape_interval:     15s 
  evaluation_interval: 15s

alerting:
  alertmanagers:
  - static_configs:
    - targets:

rule_files:

scrape_configs:
  - job_name: 'prometheus'
    static_configs:
      - targets: ['localhost:9090']

1.4 start up

systemctl daemon-reload
systemctl start prometheus
systemctl enable prometheus

1.5 view

After startup, the prometheus service will listen on port 9090. We can use IP + Port to view the simple image interface of prometheus:

It can be seen that Prometheus binary installation is very convenient and independent, with its own query web interface.
In the production environment, we can add Prometheus to the init configuration or use supervisor as the service self startup.

2, Docker installation

First, make sure you have installed the latest version of Docker. If not, please click Here.

I'll use the Mac version of Docker as a demonstration.

2.1 installation

docker run \
    -p 9090:9090 \
    -v /etc/prometheus/prometheus.yml:/etc/prometheus/prometheus.yml \
    prom/prometheus

2.2 Docker management prometheus

Run docker ps to view all services:

CONTAINER ID        IMAGE                           COMMAND                  CREATED             STATUS              PORTS                      NAMES
e9ebc2435387        quay.io/prometheus/prometheus   "/bin/prometheus -..."   26 minutes ago      Up 26 minutes       0.0.0.0:9090->9090/tcp   prometheus

Run docker start prometheus to start the service

Run docker states prometheus to view prometheus status

Run docker stop prometheus to stop the service

3, Node? Export installation

node_exporter Mainly used for * NIX system monitoring, written in Golang.

3.1 binary installation

Download address: https://prometheus.io/download/

wget https://github.com/prometheus/node_exporter/releases/download/v0.18.1/node_exporter-0.18.1.linux-amd64.tar.gz

tar xf node_exporter-0.18.1.linux-amd64.tar.gz

mv node_exporter-0.18.1.linux-amd64/node_exporter /usr/local/bin/

3.2 configuration startup file

vim /etc/systemd/system/node_export.service

[Unit]
Description=Node Export
After=network.target
Documentation=https://prometheus.io/docs/guides/node-exporter/

[Service]
Type=simple
WorkingDirectory=/tmp/
ExecStart=/usr/local/bin/node_exporter 

Restart=on-failure

[Install]
WantedBy=multi-user.target

systemctl daemon-reload
systemctl start node_export
systemctl enable node_export

3.3. Monitoring

We manually add prometheus monitoring, modify its configuration file, and add the following content at the end:

  - job_name: 'node_export'
    static_configs:
      - targets: 
        - localhost:9100
        - web1:9100
        - web2:9100
        - web3:9100
        - web4:9100

The prometheus server needs to be able to resolve these addresses, and we can modify its / etc/hosts file.

3.4 Docker installation

docker run -d \
  --net="host" \
  --pid="host" \
  -v "/:/host:ro,rslave" \
  quay.io/prometheus/node-exporter \
  --path.rootfs=/host

4, grafana rpm installation

4.1 download

Download address: https://grafana.com/grafana/download

I recommend using rpm package for installation, so that many dependencies can be solved automatically, and the startup script is also configured.

wget https://dl.grafana.com/oss/release/grafana-6.7.2-1.x86_64.rpm
sudo yum install grafana-6.7.2-1.x86_64.rpm

4.2 startup

systemctl enable grafana-server
systemctl start grafana-server

4.3 visit

Access IP + Port. The default startup port of grafana is 3000. The initial account and password are all admin. The following is the interface after login. The screenshot here is that the data source has been configured.