Playing Glances 3.0 with Grafana

Library stack

  • Grafana v5.3.2
  • Glances v3.0.2
  • InfluxDb v1.0.2

Install & conf Glances 3.0

Install Glances from PIP:

sudo apt-get install python python-pip python-dev influxdb-client
sudo pip install glances influxdb

If you need additional support for your services you have to install optional dependencies. In my case I also installed docker, hddtemp 

sudo pip install docker hddtemp

Login into InfluxDB and create database:

influx -username LOGIN -password PASS
CREATE DATABASE glances
CREATE RETENTION POLICY "one_year_only" ON "glances" DURATION 365d REPLICATION 1 DEFAULT

Create configuration file for glances:

nano /usr/src/glances.conf
[influxdb]
host=127.0.0.1
port=8086
user=USER
password=PASSWORD
db=glances
prefix=localhost

Create systemd service

Add glances to Systemd as new service. Create new file and copy/past example (source):

sudo nano /etc/systemd/system/glances.service
[Unit]
Description=Glances
After=network.target

[Service]
ExecStart=/usr/local/bin/glances --quiet -t 15 --export influxdb -C /usr/src/glances.conf
Restart=on-failure
RestartSec=30s
TimeoutSec=30s

[Install]
WantedBy=multi-user.target

Parameter -t says how often data will be refreshed. Default is 3s. I changed it, because 3s resolution is useless.

sudo systemctl enable glances.service
sudo systemctl start glances.service

After few second we can see new data stored in InfluxDB 

> use glances
Using database glances
> show series
key
localhost.cpu
localhost.diskio
localhost.docker
localhost.fs
localhost.ip
localhost.load
localhost.mem
localhost.memswap
localhost.network
localhost.percpu
localhost.processcount
localhost.sensors
localhost.system
localhost.uptime

Grafana config

Create data source

Grafana

Source configuration

Import dashboard

You can import predefined dashboard from community or prepare yours own.

Grafana

Dashboard

1 thought on “Playing Glances 3.0 with Grafana

Leave a Reply to fitri Cancel reply

Your email address will not be published. Required fields are marked *