Actions Panel

TGI: Virtualization and Softwarization of Cloud and Network Infrastructures

By Prof. Tanja Zseby

When and where

Date and time

Mon, 29 Jun 2015 09:00 - Fri, 3 Jul 2015 12:00 BST


TSSG NetLabs Boardroom, WIT West Campus, Carriganore, Co. Waterford


Recent advances in virtualization and softwarization technologies are revolutionizing the IT and telecommunications landscapes accelerating their convergence. For instance, the last few years have witnessed the rise of cloud computing, a new computing model that aims at providing resources to applications and services like public utilities such as water and electricity. Cloud computing leverages compute, storage and network virtualization for better sharing of data center resources among multiple tenants and applications offering them infrastructure, platform and software as services following a “pay-as-you-go” model. It has already become a multibillion dollar industry, however many key research challenges still exist, calling for further research. Another instance is the emergence of Software-Defined Networking (SDN) as a means to enable network programmability and faciliate network management. SDN separates a network's control logic from the underlying data forwarding, and offers an open interface between networking devices (switches and routers) and the software controlling them. SDN is already deployed in many data center networks and has the potential to change the entire networking industry in the future. A natural extension of the softwarization paradigm is to encompass the integrated management of network, compute and storage resources, referred to as Software-Defined Infrastructure (SDI). SDI enables an organization to deliver its services in the most efficient way by dynamically adapting the compute, storage and networking infrastructure transforming a static IT infrastructure into a workload-aware environment. Along this trend, Network Function Virtualization (NFV) has recently gained significant industry attention to move in-network packet processing from vertically integrated hardware middleboxes to software running on cheap commodity servers. In the same time, significant research progress has been made in the area of Network Virtualization propounded as a diversifying attribute of the future inter-networking paradigm and as a means to fend off current Internet ossification. By introducing a plurality of heterogeneous network architectures cohabiting on a shared physical substrate, Network Virtualization promotes network innovations and diversified applications. This course will cover latest advances and research challenges in Cloud computing, SDN, SDI, NFV and Network Virtualization.Hands on experience on cloud deployment, experimenting with SDN, and container based middlebox chaining will also be covered in this course.

Learning Outcomes

On successful completion of this module, students will learn about:

  1. Cloud computing basic concepts and enabling technologies
  2. Research challenges pertaining to datacenter network design, resource and performance management, energy management, security and pricing.
  3. Cloud workload characterization and scheduling
  4. Hands on experience with OpenStack
  5. Network Softwarization concepts and research chellenges including Control and Data Plane Scalability, Programming Languages, Verification, Consistent Network Update, Debugging, Fault-Tolerance, Security, Monitoring and Traffic Engineering.
  6. Software-Defined Cloud Infrastrcutures and Network Function Virtualization
  7. Experimentation with SDN and NFV technologies
  8. Network Virtualization concepts and algorithms
  9. Data Center Virtualization

Indicative Syllabus

  • Introduction to Cloud computing
  • Cloud computing research challenges
  • Cloud resource management
  • Overview of OpenStack; Deploying and migrating VMs using OpenStack Nova
  • Deployment of application stacks and auto-scaling with OpenStack Heat
  • Introduction to SDN, SDI and NFV
  • Research challenges in SDN, SDI and NFV
  • Distributed OpenFlow Testbed for large scale SDN experiments
  • NFV chaining, SDN controllers and network slicing
  • Introduction to network virtualization
  • Virtual network embedding, survivability, and management
  • Data center virtualization algorithms, reliability and energy efficiency

Assessment Methodology

  • Fullfillment of laboratory exercises on day 3 and day 5 (30%)
  • 3 research paper critiques (30%)
  • Class written test on the last day afternoon on lectures from day 1 – 5 (40%).

Essential and Supplementary Reading/Resources

Supplementary reading lists:

Contact Hours :

  • 25 hours preparatory reading (at least one paper from each category in the provided reading list).
  • 20 hours lectures, 20 hours laboratory sessions, and 10 hours nighttime reading/revision within a single week.
  • 25 hours follow-up reading and revision.

About the organiser

Organised by
Prof. Tanja Zseby

Bio to be included