Grafana dashboard for Lotus
Lotus supports exporting a wide range of metrics that can be visualized with Grafana. This tutorial guides you through setting up Prometheus and Grafana to monitor your Lotus node.
Overview
Lotus nodes export metrics that can be collected and visualized to gain insights into node performance and behavior. This tutorial will show you how to:
- Set up Prometheus to collect metrics from your Lotus node
- Set up Grafana to visualize these metrics
- Import the pre-configured dashboards from the Lotus repository
Prerequisites
- A Linux or macOS system with a running Lotus node
- Administrative/root access to install software
- The Lotus repository cloned on your system (for config files and dashboards)
Setting up monitoring
The Lotus repository includes detailed instructions for setting up Prometheus and Grafana, along with pre-configured dashboards in its metrics documentation.
Follow these steps:
Clone the Lotus repository if you haven’t already:
git clone https://github.com/filecoin-project/lotus.git cd lotusFollow the instructions in the Lotus metrics README to:
- Install and configure Prometheus
- Install and configure Grafana
- Import the pre-configured dashboards
Available Dashboards
The Lotus repository includes several pre-configured dashboards:
- F3Dashboard.json: Monitors F3 metrics like GPBFT instance completion rates, error counts, phase transitions, and encoding performance.
- MessageExecution.json: Provides insights into message execution metrics, focusing on ApplyBlocks timing data.
- Node system health: When using node_exporter, you can monitor system-level metrics like CPU, memory, disk, and network usage.
Troubleshooting
If you encounter issues:
- Make sure your Lotus node is running and exporting metrics
- Verify Prometheus is configured to scrape your Lotus node endpoint
- Check that Grafana can connect to your Prometheus instance
- Ensure you’re importing the dashboard JSON files correctly
For more detailed troubleshooting, refer to the Lotus metrics documentation.