6.6 KiB
| title |
|---|
| Monitoring Geth with InfluxDB and Grafana |
There are several ways to monitor the performance of a Geth node. Insights into a node's performance are useful for debugging, tuning and understanding what is really happening when Geth is running.
Prerequisites
To follow along with the instructions on this page it will be useful to have:
- a running Geth instance.
- basic working knowlegde of bash/terminal.
This video provides an excellent introduction to Geth monitoring.
Monitoring stack
An Ethereum client collects lots of data which can be read in the form of a chronological database. To make monitoring easier, this data can be fed into data visualisation software. There are many options available:
- Prometheus (pull model)
- InfluxDB (push model)
- Telegraf
- Grafana
- Datadog
- Chronograf
There's also Geth Prometheus Exporter, an option preconfigured with InfluxDB and Grafana. You can set it up easily using docker and Ethbian OS for RPi 4.
On this page, a Geth client will be configured to push data into a InfluxDB database and Grafana will be used to visualize the data.
Setting up InfluxDB
InfluxDB can be downloaded from the Influxdata release page. It can also be installed from a repository. For example for a Debian based Linux operating system:
curl -tlsv1.3 --proto =https -sL https://repos.influxdata.com/influxdb.key | sudo apt-key add
source /etc/lsb-release
echo "deb https://repos.influxdata.com/${DISTRIB_ID,,} ${DISTRIB_CODENAME} stable" | sudo tee /etc/apt/sources.list.d/influxdb.list
sudo apt update
sudo apt install influxdb -y
sudo systemctl enable influxdb
sudo systemctl start influxdb
sudo apt install influxdb-client
By default,InfluxDB it is reachable at localhost:8086. Before using the influx client, a new user with admin privileges
needs to be created. This user will serve for high level management, creating databases and users.
curl -XPOST "http://localhost:8086/query" --data-urlencode "q=CREATE USER username WITH PASSWORD 'password' WITH ALL PRIVILEGES"
Now the influx client can be used to enter InfluxDB shell with the new user.
influx -username 'username' -password 'password'
A database and user for geth metrics can be created by communicatign with it directly via its shell.
create database geth
create user geth with password choosepassword
Verify created entries with:
show databases
show users
Leave InfluxDB shell.
exit
InfluxDB is running and configured to store metrics from Geth.
Preparing Geth
After setting up database, metrics need to be enabled in Geth. Various options are available,
as documented in the METRICS AND STATS OPTIONS in geth --help and in our metrics page.
In this case Geth will be configured to push data into InfluxDB. Basic setup specifies the endpoint
where InfluxDB is reachable and authenticates the database.
geth --metrics --metrics.influxdb --metrics.influxdb.endpoint "http://0.0.0.0:8086" --metrics.influxdb.username "geth" --metrics.influxdb.password "chosenpassword"
These flags can be provided when Geth is started or saved to the configuration file.
Listing the metrics in the database verifies that Geth is pushing data correctly. In InfluxDB shell:
use geth
show measurements
Setting up Grafana
With the InfluxDB database setup and successfully receiving data from Geth, the next step is to install Grafana so that the data can be visualized. Instructions for specific operating systems are available on the Grafana downloads page.
Alternatively, the following code snippet shows how to download, install and run Grafana on a Debian based Linux system:
curl -tlsv1.3 --proto =https -sL https://packages.grafana.com/gpg.key | sudo apt-key add -
echo "deb https://packages.grafana.com/oss/deb stable main" | sudo tee -a /etc/apt/sources.list.d/grafana.list
sudo apt update
sudo apt install grafana
sudo systemctl enable grafana-server
sudo systemctl start grafana-server
When Grafana is up and running, it should be reachable at localhost:3000. A browser can be pointed to that URL
to access a visualization dashboard. The browser will prompt for login credentials (user: admin and password: admin).
When prompted, the default password should be changed and saved.
The browser first redirects to the Grafana home page to set up the source data. Click on the configuration icon in the left bar and select "Data sources".
There aren't any data sources yet, click on "Add data source" to define one.
Select "InfluxDB" and proceed.
Data source configuration is pretty straight forward if you are running tools on the same machine. You need to set the InfluxDB address and details for accessing the database. Refer to the picture below.
If everything is complete and InfluxDB is reachable, click on "Save and test" and wait for the confirmation to pop up.
Grafana is now set up to read data from InfluxDB. Now you need to create a dashboard which will interpret and display it. Dashboards properties are encoded in JSON files which can be created by anybody and easily imported. On the left bar, click on "Create and Import".
For a Geth monitoring dashboard, copy the ID of this dashboard and paste it in the "Import page" in Grafana. After saving the dashboard, it should look like this:
The dashboards can be customized further. Each panel can be edited, moved, removed or added. To learn more about how dashboards work, refer to Grafana's documentation.
Some users might also be interested in automatic alerting, which sets up alert notifications that are sent automatically when metrics reach certain values. Various communication channels are supported.