Mochi 🍡

The sweetest way to optimize your Kubernetes resources! Mochi delivers intelligent resource optimization with advanced analysis, confidence-scored recommendations, safety-first application, and complete change tracking. It features a beautiful terminal interface (TUI) and powerful backend services that make resource optimization as smooth as mochi itself! ✨

The go-to tool for Kubernetes resource optimization - because kubectl top is so 2024

🚀 Features

• Beautiful Terminal Interface: Interactive TUI that makes resource optimization a joy to use
• Smart Resource Analysis: Deep dive into CPU, memory, and container-level metrics with historical insights
• Intelligent Recommendations: Three optimization strategies (Conservative, Balanced, Aggressive) with confidence scoring
• One-Click Application: Seamlessly apply recommendations to your Kubernetes workloads with built-in safety
• Recommendation History: Track and manage all your applied recommendations with easy rollback capabilities
• Cost Analysis: See exactly how much you're spending and how much you could save
• Export Magic: Export your analysis and recommendations results in JSON, Markdown, or HTML formats
• REST API: Full programmatic access with health monitoring and workload discovery
• TimescaleDB Storage: Time-series database that keeps your historical data safe and sound
• Kubernetes Integration: Works with any cluster using metrics-server - no special setup needed!
• Container Analysis: Get granular insights into individual container resource usage
• Safety First: Configurable safety margins and validation rules to keep your workloads happy

🚀 Quick Start

CLI Installation

Get the mochi CLI up and running in seconds:

curl -sSL https://raw.githubusercontent.com/thread-koder/mochi/main/install.sh | bash

Server Installation

Deploy the mochi server using Helm:

helm install mochi ./charts/mochi \
  --set timescaledb.enabled=true

📖 Detailed Installation Guides:

• CLI Installation Guide - Get the mochi CLI running
• Server Installation Guide - Deploy with Helm

📋 Prerequisites

• kubernetes cluster (1.31+) with metrics-server installed
• kubectl configured and connected to your cluster
• TimescaleDB (for storage)
• Go 1.25+ (for development and building from source)

🏗️ Architecture

Mochi consists of these components:

• TUI Client: Beautiful terminal interface for analysis, recommendations, monitoring, and recommendation history
• API Server: REST API server with health monitoring and workload discovery - the heart of mochi!
• Metrics Collector: Gathers resource usage data from Kubernetes metrics-server like a pro
• Analysis Engine: Analyzes CPU, memory, performance, and cost metrics with intelligence
• Recommendation Engine: Generates optimization recommendations with confidence scoring
• Storage Backend: TimescaleDB for time-series data storage that keeps everything organized
• Export System: JSON, Markdown, and HTML export formats for sharing your insights

See Architecture Documentation for the full mochi recipe! 🍡

🤝 Contributing

We'd love to have you join our mochi-making team! 🍡 Please see our Contributing Guide for all the details.

📄 License

This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details.

📞 Support

Need help with your mochi? We've got you covered! 🍡

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Documentation: docs/

---

Made with ❤️ for the Kubernetes community