Data Center Topology, Protocols, Security, Resilience, and AI Usage

This repository contains my CSEN 329 - Network Technology (Fall 2025) research project at Santa Clara University.

The project provides a structured overview of modern data center networking, focusing on architecture choices, core protocols, security mechanisms, resiliency techniques, and the role of AI in data center operations.

Author

Wineel Wilson Dasari
MS Computer Science, Santa Clara University

Project Overview

Modern cloud services depend on data centers that must deliver low latency, high bandwidth, strong security, and high availability.

This project serves as a conceptual survey and learning document, bringing together key data center networking concepts that are often spread across textbooks, research papers, and vendor documentation.

The focus is on networking aspects of data centers, explained in a clear and structured way while reflecting real-world design practices.

Topics Covered

1. Data Center Basics and Topologies

• Role of networking in data centers
• Traditional three-tier architecture
• Spine-leaf architecture
• Clos and fat-tree topologies
• Comparison of scalability, latency, bandwidth, fault tolerance, and cost
• Example data center design and migration scenario

2. Data Center Protocols and Technologies

• Ethernet and VLAN-based segmentation
• Layer 2 vs Layer 3 design approaches
• ECMP routing using OSPF and BGP
• Overlay networking with VXLAN
• Software Defined Networking (SDN)
• Storage and transport technologies:
  • SAN, Fibre Channel, iSCSI
  • RDMA and RoCE
• Network management and automation concepts

3. Security in Data Centers

• Threat landscape and attack surfaces
• Network segmentation and micro-segmentation
• Firewalls, IDS, IPS, and zero-trust concepts
• Access control, authentication, RBAC, and MFA
• Logging, monitoring, and SIEM integration
• Defense-in-depth and operational best practices

4. Resiliency and High Availability

• Common data center failure types
• Link-level and device-level redundancy
• Load balancing at Layer 4, Layer 7, and DNS level
• Fast failover mechanisms
• Disaster recovery concepts (RPO and RTO)
• Example failure scenarios and response analysis

5. AI Usage in Data Centers

• Motivation for AI in large-scale operations
• AI-based monitoring and anomaly detection
• Traffic engineering and optimization using ML
• AI-driven security analytics
• Challenges and limitations of AI systems

Repository Structure

• README.md – Project overview and documentation
• Research Project.pdf – Final submitted report
• Research Project.docx – Editable source version
• Images/ – Figures and diagrams used in the report

Intended Audience

• Computer Science and Computer Engineering students
• Networking and IT beginners
• Readers seeking a conceptual yet practical introduction to modern data center networking

Prerequisites:
Basic understanding of OSI/TCP-IP models, IP addressing, switching, routing, and basic security concepts.

Scope and Limitations

• Focuses primarily on networking aspects of data centers
• Does not deeply cover:
  • Electrical and power design
  • Cooling systems
  • Physical building architecture
  • Detailed server hardware internals
• Intended as a learning and reference document, not a production deployment guide

References

The project is based on standard textbooks, RFCs, and industry whitepapers, including:

• Kurose and Ross - Computer Networking: A Top-Down Approach
• VXLAN RFC 7348
• Cisco ACI Design Guide
• VMware NSX Design Guide
• NREL report on AIOps for data center operations

See the full reference list in the PDF.

License

This project is shared for educational and learning purposes.
You may reference or cite it with proper attribution.