CS 854: Cloud Computing and Software-Defined Networking

Information

Instructor: Bernard Wong
Email: bernard@uwaterloo.ca
Time: Every Monday and Wednesday at 1:30-2:50 Winter 2015
Location: DC 3313
URL : http://www.cs.uwaterloo.ca/~bernard/courses/cs854-W15

Description

CS854 is a graduate course on current research topics in cloud computing and software-defined networking (SDN). It is offered as a "paper-chase" course: we read and discuss important papers on different subtopics each week. Students must also individually complete a final project that explores a new idea in one of the sub-topics covered in the course. The goal is that by the end of the course, the final project can serve as a starting point for a workshop paper submission.

Grading

The following is the tentative course grades breakdown.

Paper Reviews

Before each class, each student must submit a review for one of the papers to be discussed that day. A review should include a paragraph summarizing the core contributions of the paper, and a second paragraph that identifies weaknesses in the paper and areas in which the paper can be expanded upon.

Please inform me via email the circumstances for each late or missed submission.

Paper Presentations

Each student is expected to present the papers to be discussed in class for at least one class. The paper presentations should follow the same format as a conference talk. The presenter should be prepared with sufficient background knowledge of the related works in the area to answer broad questions and lead the class discussion.

Presenters: Please send me a copy of your slides at least two days before your presentation. I will provide feedback on the slides the next day. Do not just re-use slides provided by the paper authors. You may borrow, with attribution, figures and animations, but your slides should be created independently.

Reading

Date Presenter Topic
1/5 Bernard Introduction: General course description. Cloud computing and SDN overview.
1/7 Bernard Background (1/3)

Above the Clouds: A Berkeley View of Cloud Computing

1/12 Bernard Background (2/3)

Ethane: Taking Control of the Enterprise

1/14 Bernard Background (3/3)

OpenFlow: Enabling Innovation in Campus Networks

NOX: Towards an Operating System for Networks

1/19 Pei Network Architectures (1)

A Scalable, Commodity Data Center Network Architecture

BCube: A High Performance, Server-centric Network Architecture for Modular Data Centers

1/21 Haibo Network Architectures (2)

High Throughput Data Center Topology Design

FireFly: A Reconfigurable Wireless Datacenter Fabric using Free-Space Optics

1/26 Nouach Network Virtualization (1)

Can the Production Network Be the Testbed?

FairCloud: Sharing the Network in Cloud Computing

1/28 Guoyao Network Virtualization (2)

Chatty Tenants and the Cloud Network Sharing Problem

Network Virtualization in Multi-tenant Datacenters

2/2 Tony Flow Scheduling and Datacenter Transport Protocols (1)

Hedera: Dynamic Flow Scheduling for Data Center Networks

Data Center TCP (DCTCP)

2/4 Boris? Flow Scheduling and Datacenter Transport Protocols (2)

TCP ex Machina: Computer-Generated Congestion Control

pFabric: Minimal Near-Optimal Datacenter Transport

2/9 Lalit Simulation/Emulation/Testing

A Network in a Laptop: Rapid Prototyping for Software-Defined Networks

Reproducible Network Experiments Using Container-Based Emulation

2/11 Prateek Software-Defined WANs

B4: Experience with a Globally-Deployed Software Defined WAN

Achieving High Utilization with Software-Driven WAN

2/23 Akshat Cloud Storage (1)

Dynamo: Amazon's Highly Available Key-value Store

Bigtable: A Distributed Storage System for Structured Data

2/25 Srihari Cloud Storage (2)

A Self-Configurable Geo-Replicated Cloud Storage System>

Spanner: Google’s Globally-Distributed Database

3/2 Manish Performance Isolation

End-to-end Performance Isolation Through Virtual Datacenters

Performance Isolation and Fairness for Multi-Tenant Cloud Storage

3/4 Boris High Availability and Reliability

Characterizing Cloud Computing Hardware Reliability

A Self-Organized, Fault-Tolerant and Scalable Replication Scheme for Cloud Storage

3/9 Guoyao? Middleboxes

FlowTags: Enforcing Network-Wide Policies in the Presence of Dynamic Middlebox Actions

SIMPLE-fying Middlebox Policy Enforcement Using SDN

3/11 Bernard Scheduling

Decentralized Task-aware Scheduling for Data Center Networks

Efficient Coflow Scheduling with Varys

3/16 Tony? Distributed Computation

MapReduce: Simplified Data Processing on Large Clusters

Resilient Distributed Datasets: A Fault-Tolerant Abstraction for In-Memory Cluster Computing

3/18 Nouach? Software Support

The Chubby lock service for loosely-coupled distributed systems

ZooKeeper: Wait-free coordination for Internet-scale systems

3/23 Michael Green Computing

Don't Lose Sleep Over Availability: The GreenUp Decentralized Wakeup Service

It's Not Easy Being Green

3/25 Srihari? Network Architectures (3)

Jellyfish: Networking Data Centers Randomly

REWIRE: An optimization-based framework for unstructured data center network design

3/30 Everyone Final Project Presentations (1)
4/1 Everyone Final Project Presentations (2)

Last modified: Jan 5, 2015