Metric - Basic Example

Code Source: https://github.com/gogf/examples/tree/main/observability/metric/basic

Description

This example demonstrates the basic usage of various metric types in GoFrame using OpenTelemetry and Prometheus integration. It shows how to:

• Create and use different types of metrics
• Configure metric attributes
• Export metrics in Prometheus format
• Set up a metrics endpoint

Structure

- `go.mod`: The Go module file for dependency management
- `go.sum`: The Go module checksums file
- `main.go`: The main application demonstrating metric usage

Features

The example showcases the following metric types:

1. Counter

   • Cumulative measurements
   • Only increases
   • Used for counting events

2. UpDownCounter

   • Bidirectional counter
   • Can increase and decrease
   • Used for measuring varying quantities

3. Histogram

   • Distribution of measurements
   • Configurable buckets
   • Used for latency/size measurements

4. Observable Metrics

   • Counter
   • UpDownCounter
   • Gauge
   • Updated via callbacks

Requirements

• Go 1.22 or higher
• Git
• GoFrame
• GoFrame OpenTelemetry Metric

Usage

1. Run the example:

   go run main.go

2. Access the metrics:

   # Using curl
   curl http://localhost:8000/metrics
   
   # Or open in browser
   http://localhost:8000/metrics

3. Example metrics output:

   # HELP goframe_metric_demo_counter This is a simple demo for Counter usage
   goframe_metric_demo_counter{const_attr_1="1"} 11
   
   # HELP goframe_metric_demo_histogram This is a simple demo for histogram usage
   goframe_metric_demo_histogram_bucket{const_attr_3="3",le="0"} 0
   goframe_metric_demo_histogram_bucket{const_attr_3="3",le="10"} 1
   ...

Implementation Details

The example demonstrates:

1. Proper metric initialization and configuration
2. Different ways to update metrics
3. Attribute handling
4. Prometheus integration
5. Observable metric callbacks

Notes

• Metrics are exported in Prometheus format
• Default port is 8000
• Metrics endpoint is at /metrics
• All metrics include constant attributes
• Observable metrics are updated via callbacks
• Proper cleanup is handled via defer