What is Gigapipe Mcp

Gigapipe MCP is a Machine Control Protocol server designed to provide a unified interface for querying Prometheus metrics, Loki logs, and Tempo traces.

Use cases

Use cases include monitoring application metrics, analyzing logs for troubleshooting, and tracing requests across microservices.

How to use

To use Gigapipe MCP, clone the repository, build the server using Go, and configure it with the necessary environment variables. You can then utilize Prometheus and Loki tools to query metrics and logs.

Key features

Key features include Prometheus integration for querying metrics with PromQL, Loki integration for querying logs with LogQL, and Tempo integration for querying traces by trace ID.

Where to use

Gigapipe MCP can be used in various fields such as cloud monitoring, application performance management, and log analysis.

Gigapipe MCP Server

A Machine Control Protocol (MCP) server for Gigapipe, providing a unified interface to query Prometheus metrics, Loki logs, and Tempo traces.

Features

Prometheus Integration

• Query metrics using PromQL
• List all available labels
• Get values for specific labels
• Support for time range queries

Loki Integration

• Query logs using LogQL
• List all available labels
• Get values for specific labels
• Support for time range queries

Tempo Integration

• Query traces by trace ID
• List all available trace tags
• Get values for a specific trace tag
• Support for JSON trace format

Installation

# Clone the repository
git clone https://github.com/lmangani/gigapipe-mcp.git
cd gigapipe-mcp

# Build the server
go build -o gigapipe-mcp

Configuration

The server can be configured using environment variables:

# Required: Gigapipe server address (default: http://localhost:3100)
# Must include protocol (http:// or https://)
export GIGAPIPE_HOST="https://your-host:3100"

# Optional: HTTP Basic Authentication
export GIGAPIPE_USERNAME="your-username"
export GIGAPIPE_PASSWORD="your-password"

Note: If no protocol is specified in GIGAPIPE_HOST, http:// will be used as the default.

Usage

Prometheus Tools

1. Query Metrics

# Query metrics with PromQL
prometheus_query --query="rate(http_requests_total[5m])" --start="2024-01-01T00:00:00Z" --end="2024-01-01T01:00:00Z" --step="1m"

2. List Labels

# List all available labels
prometheus_labels --start="2024-01-01T00:00:00Z" --end="2024-01-01T01:00:00Z"

3. Get Label Values

# Get values for a specific label
prometheus_label_values --label="instance" --start="2024-01-01T00:00:00Z" --end="2024-01-01T01:00:00Z"

Loki Tools

1. Query Logs

# Query logs with LogQL
loki_query --query='{job="varlogs"}' --start="2024-01-01T00:00:00Z" --end="2024-01-01T01:00:00Z" --limit="100"

2. List Labels

# List all available labels
loki_labels --start="2024-01-01T00:00:00Z" --end="2024-01-01T01:00:00Z"

3. Get Label Values

# Get values for a specific label
loki_label_values --label="job" --start="2024-01-01T00:00:00Z" --end="2024-01-01T01:00:00Z"

Tempo Tools

1. Query Traces

# Query a trace by ID
tempo_query --trace_id="1234567890abcdef"

2. List Tags

# List all available trace tags
tempo_tags

3. Get Tag Values

# Get values for a specific trace tag
tempo_tag_values --tag="service.name"

API Endpoints

The server communicates with Gigapipe using the following endpoints:

Prometheus

• Query Range: /api/v1/query_range
• Query: /api/v1/query
• Labels: /api/v1/labels
• Label Values: /api/v1/label/:name/values

Loki

• Query Range: /loki/api/v1/query_range
• Query: /loki/api/v1/query
• Labels: /loki/api/v1/label
• Label Values: /loki/api/v1/label/:name/values

Tempo

• Query Traces: /api/traces/:traceId
• Query Traces (JSON): /api/traces/:traceId/json
• Tags: /api/search/tags
• Tag Values: /api/search/tag/{name}/values

Development

Prerequisites

• Go 1.16 or later
• Git

Building from Source

# Clone the repository
git clone https://github.com/lmangani/gigapipe-mcp.git
cd gigapipe-mcp

# Install dependencies
go mod download

# Build the server
go build -o gigapipe-mcp

# Run tests
go test ./...

License

MIT License