🧪 Custom Exporter for Prometheus

A flexible Prometheus exporter with YAML-based configuration for collecting metrics from shell commands, HTTP endpoints, or files.

🔧 Features

• ⚙️ YAML-Driven Configuration - Define exporters easily in a single YAML file.
• 🧾 Multiple Sources - Collect metrics from:
  • Shell commands
  • HTTP APIs
  • Local files
• 🔍 Flexible Parsing - Use regex, json, line, or split to extract metrics.
• 📊 Gauge & Counter metric types.
• 📡 Exposes /metrics, /uptime, and /health endpoints.
• 🖥️ Serves a static HTML page at /.

🚀 Getting Started

1. Clone and Build

git clone https://github.com/kirillsaidov/custom-exporter.git
cd custom-exporter
go build -o custom-exporter cmd/custom-exporter/main.go

2. Run the Exporter

Run manully

./custom-exporter --config export.yaml --port 9100

• Default config file: export.yaml
• Default port: 9100

Setup systemd service

Create custom-exporter.service file:

[Unit]
Description=Custom Exporter for Prometheus
After=network.target
StartLimitBurst=5

[Service]
Type=simple
ExecStart=%h/custom-exporter/custom-exporter --config %h/custom-exporter/export.yaml --port 9100
Restart=on-failure
RestartSec=5

[Install]
WantedBy=default.target

%h - stands for HOME directory.

Run the service:

# 1. create service directory and copy file
mkdir -p ~/.config/systemd/user
cp custom-exporter.service ~/.config/systemd/user/

# 2. enable and start the service
systemctl --user daemon-reload
systemctl --user enable custom-exporter
systemctl --user start custom-exporter

# 3. check status
systemctl --user status custom-exporter

3. Example YAML Configuration

Here’s a minimal example of export.yaml:

exporters:
  - name: "example_command_metric"
    type: "command"
    command: "echo 42"
    interval: 10
    metric_type: "gauge"
    parser:
      type: "regex"
      pattern: "(\\d+)"
    labels:
      source: "shell"
    description: "A simple shell-based metric"

For full configuration examples, see export.yaml.example.

🧠 Parser Types

Type	Description
regex	Extracts value via regex (first capture group).
json	Navigates JSON using dot-path (key.subkey).
line	Gets a specific line by number (line_num).
split	Splits text by delimiter and selects index.

📦 Output Example

# curl http://localhost:9100/metrics

# HELP example_command_metric A simple shell-based metric
# TYPE example_command_metric gauge
example_command_metric{source="shell"} 42

🖥️ HTTP Endpoints

Endpoint	Description
/metrics	Prometheus metrics
/uptime	Uptime in seconds
/health	Simple health check (OK)
/	Serves static HTML (optional)

🛠 Advanced Usage

• Multiple exporters with different intervals and sources
• Dynamic labels
• Built-in metric caching
• Error logging for failed fetches/parsing

LICENSE

MIT.