Monitoring your validator with Grafana and Prometheus

PreviousNext Steps NextMobile App Node Monitoring by beaconcha.in

Last updated 3 months ago

Monitoring your validator with Grafana and Prometheus

Prometheus is a monitoring platform that collects metrics from monitored targets by scraping metrics HTTP endpoints on these targets.

Grafana is a dashboard used to visualize the collected data.

1. Install Prometheus and Node Exporter

sudo apt-get install -y prometheus prometheus-node-exporter

2. Install Grafana

sudo apt-get install -y apt-transport-https
sudo apt-get install -y software-properties-common wget
sudo wget -q -O /usr/share/keyrings/grafana.key https://apt.grafana.com/gpg.key

echo "deb [signed-by=/usr/share/keyrings/grafana.key] https://apt.grafana.com stable main" | sudo tee -a /etc/apt/sources.list.d/grafana.list
sudo apt-get update && sudo apt-get install -y grafana

3. Enable services so they start automatically

sudo systemctl enable grafana-server prometheus prometheus-node-exporter

4. Create the prometheus.yml config file

Remove the default prometheus.yml configuration file and edit a new one.

sudo rm /etc/prometheus/prometheus.yml
sudo nano /etc/prometheus/prometheus.yml

Choose the tab for your consensus client. Paste the following configuration into the file.

global:
  scrape_interval:     15s # By default, scrape targets every 15 seconds.

  # Attach these labels to any time series or alerts when communicating with
  # external systems (federation, remote storage, Alertmanager).
  external_labels:
    monitor: 'codelab-monitor'

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
   - job_name: 'node_exporter'
     static_configs:
       - targets: ['localhost:9100']
   - job_name: 'lighthouse'
     metrics_path: /metrics    
     static_configs:
       - targets: ['localhost:8008']
   - job_name: 'lighthouse_validator'
     metrics_path: /metrics
     static_configs:
       - targets: ['localhost:8009']

global:
  scrape_interval:     15s # By default, scrape targets every 15 seconds.

  # Attach these labels to any time series or alerts when communicating with
  # external systems (federation, remote storage, Alertmanager).
  external_labels:
    monitor: 'codelab-monitor'

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
   - job_name: 'node_exporter'
     static_configs:
       - targets: ['localhost:9100']
   - job_name: 'Nimbus'
     metrics_path: /metrics
     static_configs:
       - targets: ['localhost:8008']
   - job_name: 'Nimbus_Validator'
     metrics_path: /metrics
     static_configs:
       - targets: ['localhost:8009']

global:
  scrape_interval:     15s # By default, scrape targets every 15 seconds.

  # Attach these labels to any time series or alerts when communicating with
  # external systems (federation, remote storage, Alertmanager).
  external_labels:
    monitor: 'codelab-monitor'

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
   - job_name: 'node_exporter'
     static_configs:
       - targets: ['localhost:9100']
   - job_name: 'Teku'
     metrics_path: /metrics
     static_configs:
       - targets: ['localhost:8008']
   - job_name: 'Teku_Validator'
     metrics_path: /metrics
     static_configs:
       - targets: ['localhost:8009']

global:
  scrape_interval:     15s # By default, scrape targets every 15 seconds.

  # Attach these labels to any time series or alerts when communicating with
  # external systems (federation, remote storage, Alertmanager).
  external_labels:
    monitor: 'codelab-monitor'

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
   - job_name: 'node_exporter'
     static_configs:
       - targets: ['localhost:9100']
   - job_name: 'validator'
     static_configs:
       - targets: ['localhost:8009']
   - job_name: 'Prysm'
     static_configs:
       - targets: ['localhost:8008']

global:
  scrape_interval:     15s # By default, scrape targets every 15 seconds.

  # Attach these labels to any time series or alerts when communicating with
  # external systems (federation, remote storage, Alertmanager).
  external_labels:
    monitor: 'codelab-monitor'

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
   - job_name: 'node_exporter'
     static_configs:
       - targets: ['localhost:9100']
   - job_name: 'beacon'
     metrics_path: /metrics    
     static_configs:
       - targets: ['localhost:8008']
   - job_name: 'validator'
     metrics_path: /metrics    
     static_configs:
       - targets: ['localhost:8009']

5. Setup prometheus for your execution client

Append the applicable job snippet for your execution client to the end of prometheus.yml.

Spacing matters. Ensure all job_name snippets are in alignment.

   - job_name: 'geth'
     scrape_interval: 15s
     scrape_timeout: 10s
     metrics_path: /debug/metrics/prometheus
     scheme: http
     static_configs:
       - targets: ['localhost:6060']

   - job_name: 'besu'
     scrape_interval: 15s
     scrape_timeout: 10s
     metrics_path: /metrics
     scheme: http
     static_configs:
     - targets:
       - localhost:6060

   - job_name: 'nethermind'
     static_configs:
       - targets: ['localhost:6060']

   - job_name: 'erigon'
     scrape_interval: 10s
     scrape_timeout: 3s
     metrics_path: /debug/metrics/prometheus
     scheme: http
     static_configs:
       - targets: ['localhost:6060']

   - job_name: 'reth'
     metrics_path: "/"
     scrape_interval: 10s
     static_configs:
       - targets: ['localhost:6060']

Here's an example of a Lighthouse-Nethermind config:

global:
  scrape_interval:     15s # By default, scrape targets every 15 seconds.

  # Attach these labels to any time series or alerts when communicating with
  # external systems (federation, remote storage, Alertmanager).
  external_labels:
    monitor: 'codelab-monitor'

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
   - job_name: 'node_exporter'
     static_configs:
       - targets: ['localhost:9100']
   - job_name: 'Lighthouse'
     metrics_path: /metrics
     static_configs:
       - targets: ['localhost:5054']
   - job_name: 'validators'
     metrics_path: /metrics
     static_configs:
       - targets: ['localhost:5064']
   - job_name: 'nethermind'
     static_configs:
       - targets: ['localhost:6060']

To exit and save, press Ctrl + X, then Y, then Enter.

Update file permissions.

sudo chmod 644 /etc/prometheus/prometheus.yml

Restart the services.

sudo systemctl restart grafana-server prometheus prometheus-node-exporter

Verify that the services are running.

sudo systemctl status grafana-server prometheus prometheus-node-exporter

6. Create a SSH Tunnel to Grafana

Each time you want to access Grafana, create a SSH tunnel with port 3000 forwarded.

Example of how to create a SSH tunnel in Linux or MacOS:

ssh -N -v <user>@<staking.node.ip.address> -L 3000:localhost:3000

#Full Example
ssh -N -v ethereum@192.168.1.69 -L 3000:localhost:3000

Navigate to Connection > SSH > Tunnels > Enter Source Port 3000 > Enter Destination localhost:3000 > Click Add

Now save your configuration. Navigate to Session > Enter a session name > Save

Click Open to open a connection

7. Setup Grafana Dashboards

Login with admin / admin
Change password
Click the configuration gear icon, then Add data Source
Select Prometheus
Set Name to "Prometheus"
Click Save & Test
Click Create + icon > Import
Add the consensus client dashboard via Upload JSON file
If needed, select Prometheus as Data Source.
Click the Import button.
Repeat steps 12-15 for the execution client dashboard.
Repeat steps 12-15 for the node-exporter dashboard.

Symptom 1: Your dashboard is missing some data_._

Solution 1: Ensure that the execution or consensus client has enabled the appropriate metrics flag.

cat /etc/systemd/system/execution.service
cat /etc/systemd/system/consensus.service

Geth: --http --metrics --pprof
Besu: --metrics-enabled=true
Nethermind: --Metrics.Enabled true
Erigon: --metrics
Reth: --metrics 127.0.0.1:9001
Lighthouse: --validator-monitor-auto
Nimbus: --metrics --metrics-port=8008
Teku: --metrics-enabled=true --metrics-port=8008
Lodestar: --metrics true

Symptom 2: Don't want to use SSH tunnels and you want to expose port 3000 to access Grafana, but understand the security concerns.

Solution 2: Open port 3000 in your ufw firewall. Access grafana at http://<node ipaddress>:3000

sudo ufw allow 3000

Example of Grafana Dashboards for each consensus client.

Example of Grafana Dashboards for each execution client.

Example of Node-Exporter Dashboard

General system monitoring

Includes: CPU, memory, disk IO, network, temperature and other monitoring metrics。

8. Setup Alert Notifications

Setup alerts to get notified if your validators go offline.

Get notified of problems with your validators. Choose between email, telegram, discord or slack.

Option 1: Email Notifications

Sign up for an account
Verify your email
Search for your validator's public address
Add validators to your watchlist by clicking the bookmark symbol.

Option 2: Telegram Notifications

On the menu of Grafana, select Alerting.
Click on Contact points menu, then +Create contact point button.
Give the contact point a name.
Select Telegram from the Integration list.
Create a new telegram group.
Invite the bot to your new group.
Type at least 1 message into the group to initialize it.
In the JSON response, find and copy the Chat ID. Find it between **chat **and title. Example of Chat ID: -1123123123
```
"chat":{"id":-123123123,"title":
```
Paste the Chat ID into the corresponding field in Grafana.
Save and test the notification channel for your alerts.

Option 3: Discord Notifications

On the menu of Grafana, select Alerting.
Click on Contact points menu, then +Create contact point button.
Give the contact point a name.
Select Discord from the Integration list.
Enter the Webhook URL in the Discord notification settings panel.
Click Send Test, which will push a confirmation message to the Discord channel.

Option 4: Slack Notifications

On the menu of Grafana, select Alerting.
Click on Contact points menu, then +Create contact point button.
Give the contact point a name.
Select Slack from the Integration list.
Enter the Slack Incoming Webhook URL in the URL field.
Click Send Test, which will push a confirmation message to the Slack channel.