hephaestus

README

Hephaestus

Hephaestus is a distributed system for automated log analysis and solution generation. It uses machine learning models to analyze system logs and propose solutions for identified issues.

Features

• Node-based distributed architecture
• Real-time log processing and analysis
• Integration with remote repositories
• Automated solution generation using AI models
• Prometheus metrics collection
• gRPC-based API interface

Requirements

• Go 1.21 or later
• Prometheus (for metrics collection)
• Access to a remote repository service
• Access to an AI model service provider (OpenAI or Anthropic)

Installation

go get github.com/HoyeonS/hephaestus

Configuration

The system requires configuration for:

• Remote repository settings
• Model service settings
• Logging configuration
• Repository settings

Example configuration:

remoteSettings:
  authToken: "your-auth-token"
  repositoryOwner: "owner"
  repositoryName: "repo"
  targetBranch: "main"

modelSettings:
  serviceProvider: "openai"
  serviceApiKey: "your-api-key"
  modelVersion: "gpt-4"

loggingSettings:
  logLevel: "info"
  outputFormat: "json"

repositorySettings:
  repositoryPath: "/path/to/repo"
  fileLimit: 1000
  fileSizeLimit: 10485760 # 10MB

Usage

1. Initialize a new node:

node, err := manager.CreateSystemNode(ctx, config)

1. Process logs:

err := processor.ProcessLogs(ctx, nodeID, logs)

1. Generate solutions:

solution, err := service.GenerateSolution(ctx, nodeID, logEntry)

Metrics

The system collects various metrics using Prometheus:

• Node operations
• Node status
• Log processing duration
• Model latency
• Repository errors

Development

To run tests:

go test ./...

To build:

go build ./...

License

This project is licensed under the MIT License - see the LICENSE file for details.