Metrics are the foundation of understanding how Metaform performs in real-world workloads. They provide visibility into system health, query execution, and resource utilization, allowing you to diagnose issues and optimize performance.

Each metric is reported on a per-Node basis, reflecting the behavior of an individual Metaform node rather than the cluster as a whole. By collecting and analyzing these metrics, you can answer questions such as:

• Is the Node handling queries efficiently?
• Where are the bottlenecks in data processing?
• How are resources like CPU and memory being consumed?

This section explains how to access Metaform’s metrics and how to interpret them so you can integrate these insights into your monitoring and analytics workflows.

Accessing Metrics

Once Metaform is running, you can inspect its internal metrics through a simple HTTP endpoint. These metrics expose real-time information about the health and performance of the system, making them essential for monitoring and troubleshooting.

To view the metrics:

1. Open a web browser and navigate to the following URL:

http://<ip>:8080/metrics

Replace <ip> with the IP address of the Metaform node you want to monitor.

2. For example:

• If you are running Metaform locally, enter:

http://localhost:8047/metrics
```

• If your deployment is hosted at 192.168.1.2, enter:

http://192.168.1.2:8047/metrics
```

3. The endpoint returns a text-based report with key metrics, such as query execution times, memory consumption, and active sessions. The output will look similar to the following:

Node-reported load (local view; may differ from OS aggregate).

node.load.avg

Process uptime (seconds). Useful for normalizing rates: per-second = count / uptime.

node.uptime

These metrics can be consumed directly by your monitoring tools or inspected manually to better understand how Metaform is performing.