25
Click to edit Master subtitle style 25/08/09 Cloud Computing Course Administration Vasudeva Varma

Academic Perseptive [ppt]

Embed Size (px)

Citation preview

Page 1: Academic Perseptive [ppt]

Click to edit Master subtitle style

25/08/09

Cloud Computing Course Administration

Vasudeva Varma

Page 2: Academic Perseptive [ppt]

25/08/09

Background

Cloud Computing research in IIIT Hyderabad

Hadoop in SIEL/Other places in IIIT Hyderabad

Previous edition of cloud computing course

Collaboration with Yahoo!

Collaboration with Amazon

Collaboration with others

Google, Intel, Microsoft, ITsAP…

Page 3: Academic Perseptive [ppt]

25/08/09

Course Administration

TAs:

GRJ Srinivas

Padmini Priyadarshini  

Reference Material

No Text book

Relevant material (or pointers to them) will be provided on the course web site

Page 4: Academic Perseptive [ppt]

25/08/09

Course Administration

Course website: http://search.iiit.ac.in/cloud

Course mailing list: [email protected]

Two Mid Semester Exams (10%) each

End Semester Exam (20%)

Assignments (30%)

Project 40%

Rules

Strict Plagiarism Policy– Punishment: F Grade

Be on-time for the class

Page 5: Academic Perseptive [ppt]

Click to edit the outline text format

Second Outline Level

Third Outline Level Fourth Outline Level Fifth Outline Level Sixth Outline Level Seventh Outline Level Eighth Outline Level

Ninth Outline LevelClick to edit Master text styles

Second level

Third level

Fourth level

Fifth level

25/08/09

Course Outline

Introduction to Cloud Computing

Cloud introduction and overviewDifferent clouds Risks RequirementsIntroduction Cloud computing architectureOn Demand ComputingVirtualization at the infrastructure levelSecurity in Cloud computing environmentsCPU Virtualization - A discussion on HypervisorsStorage Virtualization - A discussion on SAN, ISCSINetwork Virtualization - A discussion on VLAN

Cloud Computing Architecture - Amazon Case study Introduction to Amazon Cloud Computing services, Amazon EC2, Amazon S3 , Amazon DB, Queues and Cloud FrontPractical Amazon

Page 6: Academic Perseptive [ppt]

25/08/09

Course outline…

Introduction to Distributed Computing

Fallacies of Distributed ComputingSources of failure in distributed computingLarge Data sets and the handling issuesMoore Law & relationship to storage, CPU speed, etc

Introduction Map Reduce

Discussion of Google PaperDiscussion of BigTables, GFS, HDFSMapReduce - Hadoop FrameworkMap Reduce – Hadoop Framework examplesHadoop and related Projects, HBase, Pig, Mahut etc

Current topics of CC @ IIIT-H

Scheduling

Test bed creation

Hardware aware middleware

Page 7: Academic Perseptive [ppt]

25/08/09

Planned Guest Lectures

Grid and Cloud Computing Group - Yahoo

Amazon Web Services – Amazon

Google App engine – Google

Cloud in the enterprise by Ramesh Logananthan, Progress software

Some others…?

Page 8: Academic Perseptive [ppt]

Click to edit Master subtitle style

25/08/09

Migration to Cloud Computing: An Academic and Research Perspective

Page 9: Academic Perseptive [ppt]

25/08/09

What we do?

Crawling, Parsing and Indexing the web - vertically and horizontally

Cross Lingual Information Retrieval

Digital Libraries - Book Search

Enterprise search

Information extraction from Semi-structured and un-structured data sources

Multi-document Summarization

Machine Translation

Speech Processing

Page 10: Academic Perseptive [ppt]

25/08/09

Academic Research is getting more complexNature of the tasks

Even lab prototypes are at the web scale

Need to deal with large data

More experiments to validate research claims/assumptions

Increase in overall Computing infrastructure costs

Hardware

Bandwidth

Maintenance

Common resources

Local

Central

Global

Page 11: Academic Perseptive [ppt]

25/08/09

Motivation to Migrate to Cloud ComputingSeveral projects need to process huge amount

of data

Computational needs are far beyond capacity of a single machine

On single nodes programs can run for days (On cluster its a matter of hours)

Less time for experiments => More experiments!

We needed a parallel programming framework that is

Easy to program, Easy to learn, Easy to deploy,

Does most of the things 'automagically’

Page 12: Academic Perseptive [ppt]

25/08/09

Cloud Computing in the NEWS

IBM, Google, UW partnership

Yahoo!, HP, Intel + 6 research organizations

Amazon

Page 13: Academic Perseptive [ppt]

25/08/09

Market Oriented Cloud Architecture

ApplicationP2P (Skype), Web Apps (Facebook), SaaS (Google Aps, SaleForce), Software+Services (MS Online Services)

Hardware – Android

Infrastructure EC2, Sun Grid

Platform Google App Engine, Mosso

Services: AWS, Mechanical Turk

Storage: S3, Live Mesh, Mobile Me

Page 14: Academic Perseptive [ppt]

Click to edit the outline text format

Second Outline Level

Third Outline Level Fourth Outline Level Fifth Outline Level Sixth Outline Level Seventh Outline Level Eighth Outline Level

Ninth Outline LevelClick to edit Master text styles

Second level

Third level

Fourth level

Fifth level

25/08/09

Cloud Accelerates Innovation

Expand the Innovation playground

Reduce the Time to market/conference

When you work on a cloud, entry barriers look lower and smaller

Page 15: Academic Perseptive [ppt]

25/08/09

Comparison of Cloud Computing Offerings

Page 16: Academic Perseptive [ppt]

25/08/09

Cloud Computing market Place

Page 17: Academic Perseptive [ppt]

25/08/09

Different Clouds

Virtual Private Cloud (VPC)

An IT solution platform that integrates local and commercial cloud infrastructure under a single management

Regional Clouds

Geographical Cloud Computing - Response based scaling on a regional basis

Regional Scaling and Monitoring

Offline Clouds

What happens when your lose connectivity?

What happens when your infrastructure goes down?

Page 18: Academic Perseptive [ppt]

Click to edit the outline text format

Second Outline Level

Third Outline Level Fourth Outline Level Fifth Outline Level Sixth Outline Level Seventh Outline Level Eighth Outline Level

Ninth Outline LevelClick to edit Master text styles

Second level

Third level

Fourth level

Fifth level

25/08/09

The next generation Cloud has many challenges to tackle

Security

Scalability

Availability

Reliability

Cost Efficient

Data Intensive

Personalization

Mobility

Latency

Manageability

1818

Page 19: Academic Perseptive [ppt]

25/08/09

Challenges – At the Service LevelHow to translate business concept

requirements to infrastructure requirements?

Define a Service Definition Language?

Must be able to handle rollback on deployment failures

How to determine the mapping of high level service requirements and metrics (e.g., response time) to infrastructure level requirements and metrics (e.g., CPU utilization)

Support multiple levels of QoS1919

Page 20: Academic Perseptive [ppt]

25/08/09

Management Challenges

Support policy based management across Clouds

Dynamically automatically hire additional 'power‘ from a another cloud

Enable intra-site and inter-site workload optimization

Protect Service Level Agreements

Detect violations (SLA monitoring)

Provide for dynamic relocation of resources

Provide accountability

Bill for services used, even across Clouds

Different billing and accounting systems may be used.

2020

Page 21: Academic Perseptive [ppt]

25/08/09

Challenges at the Infrastructure level

Provide for relocation of resources without boundaries

Live migration across subnet boundaries

Migration to a different physical host without shared storage

Speculative Execution – Heterogeneous infrastructure

Provide standardized interfaces for lifecycle management to Virtualized Execution Environment

Analyze end-to-end performance in a virtualized environment to understand bottlenecks

Be able to handle surges in 3-5 orders of magnitude in service requests

RESERVOIR @ NOTUR 20082121

Page 22: Academic Perseptive [ppt]

25/08/09

Some more questions and challengesThe Long tail of Web services

Accelerating academic skills and research (CC Research Beds)

Cloud OS – EyeOS, AIR (Adobe), Open-Laszlo

UI Challenges – as rich experience as workstations?

Multi-platform, multi-language and multi-platform

Need for new programming paradigms?

Page 23: Academic Perseptive [ppt]

25/08/09

We know about …Outsourci

ngAre we now getting into …

Right-sourcing

Cloud

Sourcing?If so, who will be

emerging as superpower?

Page 24: Academic Perseptive [ppt]

25/08/09

Key Message: One cloud

"I think there is a world market for about five computers" — Remark attributed to Thomas J. Watson (Chairman of the Board of International Business Machines) – 1943

“… In a sense, says Yahoo Research Chief PrabhakarRaghavan, there are only five computers on earth. He lists Google, Yahoo, Microsoft, IBM, and Amazon. Few others, he says, can turn electricity into computing power with comparable efficiency …”

From , by Steven Baker - BusinessWeek.com

“… The World Wide Web is becoming one vast, programmable machine. As NYU's Clay Shirky likes to say, Watson was off by four …” – Nicholas Carr

From

Page 25: Academic Perseptive [ppt]

Click to edit Master subtitle style

25/08/09

Thank You

<[email protected]>