Back to Projects
Openstack Mongosh Jupyter Elastic Kibana

Implementing Virtualized Microservices and AI Workloads on OpenStack Private Cloud

This project demonstrates the design and deployment of a private cloud infrastructure using OpenStack to host virtualized microservices and high-performance AI workloads, offering a cost-effective open-source alternative to expensive enterprise licenses.

Implementing Virtualized Microservices and AI Workloads on OpenStack Private Cloud

Project Overview

In the modern era of technology, high-performance computing has become a necessity. However, enterprise licensing costs often become a barrier. This project explores OpenStack as a cost-effective solution for heavy workloads such as AI. It proves that building high-performance infrastructure does not have to be expensive

System Architecture

1

Horizon (Dashboard)

The web-based dashboard used to manage and monitor the entire cloud infrastructure.

2

Keystone (Identity)

The identity service that handles authentication and authorization for all users and services.

3

Nova (Compute)

The compute engine responsible for managing and provisioning virtual machines (instances).

4

Neutron (Networking)

The networking component that manages virtual networks, IP addresses, and connectivity.

5

Glance (Image)

The image service used to store and manage the virtual disk images needed to boot instances.

Key Features

Database VM (MongoDB)

Stores and manages large datasets required for AI analysis efficiently.

AI Engine VM

Equipped with Anaconda, Jupyter, and Python for model training and analysis.

Monitoring VM

Dedicated to system health tracking to ensure stable infrastructure.

Full Scalability

Placement service optimizes resource allocation across CPU, RAM, and Disk.

System Flow

1

Data Collection

Honeypot T-Pot captures attack interactions and records logs and payloads in the primary database.

2

Pre-processing

VM 2 extracts files and normalizes log data into a consistent format suitable for analysis.

3

Dual Analysis

The system performs static signature checks and dynamic runtime monitoring to identify malicious behaviors.

4

AI Classification

Processed data is fed into an AI model to generate threat scores and risk categories.

5

Indexing and Visualization

Results are indexed in Elasticsearch and displayed on Kibana for real-time monitoring and research.

Project Outcome

The project successfully established a functional OpenStack private cloud capable of handling diverse workloads. It achieved an end-to-end automated pipeline for malware analysis, from honeypot capture to AI-based classification and visualization. The implementation demonstrates that organizations can maintain full operational control and data sovereignty while running scalable, high-performance security infrastructures.

Screenshots

OpenStack Infrastructure Diagram
OpenStack Dashboard