cloud computing introduction - cinvestavvjsosa/clases/tssd/01... · 2016-07-08 · cloud computing...

07/07/2016

Cinvestav-Tamaulipas 2009-2010 1

Cloud Computing Intro 1

Cloud Computing IntroductionInstructor

Víctor J. [email protected]

Cinvestav

2Selected topics in Distributed Systems

1. Cloud Computing

1.1 Introduction

1.2 Essential characteristics

1.3 Related technologies

1.4 Taxonomy of Cloud Services

1.4.1 Infrastructure as a Service (IaaS)

1.4.2 Platform as a Service (PaaS)

1.4.3 Software as a Service (SaaS)

1.5 Types of deployment:

Private, Public, Community, and Hybrid.

1.6 Technical aspects

1.7 Cloud Computing tools.

1.7.1 IaaS, PaaS and SaaS tools

1.7.2 GAE vs Win Azure

1.7.3 Cloud management platforms

1.8 Standard initiatives

2. Virtualization

2.1 Introduction

2.2 Virtualization scenarios

2.3 Types of virtualization

2.4 Virtualization tools

2.5 Virtual infrastructure management

3. Practical case: Installation and use of a small private

cloud

3.1 Installation of a Virtual Machine Monitor (VMM)

3.2 Instantiation of a Virtual Machine

3.3 Installation of a Cloud Management Platform

INTRODUCTION AND TECHNICAL BACKGROUND

Content

07/07/2016



Bibliography• U.S. National Institute for Standards and Technology (NIST). A Cloud Computing Definition.

Available online at: http://csrc.nist.gov/publications/PubsSPs.html#800-145

• Dan C. Marinescu. Cloud Computing: Theory and Practice. Morgan Kaufmann. ISBN: 124046274.

2013

• Cloud Computing Use Case Discussion Group. Cloud Computing Use Case Whitepaper v4.0.

July/2010. Available online at: http://cloud-computing-use-

cases.googlegroups.com/web/Cloud_Computing_Use_Cases_Whitepaper-4_0.pdf

• The Open Cloud Manifesto Website. http://www.opencloudmanifesto.org/

• Susanta Nanda and Tzi-cker Chiueh. A Survey on Virtualization Technologies. Technical Report (TR-

179). Department of Computer Science, SUNY at Stony Brook. NY 2005. Available online at:

http://www.ecsl.cs.sunysb.edu/tr/TR179.pdf

• Joshua S. White and Adam W. Pilbeam. A Survey of Virtualization Technologies With Performance

Testing. Technical Report, Cornell University Library. Available online at:

http://arxiv.org/pdf/1010.3233v1

• J. C. Duenas, J. L. Ruiz, F. Cuadrado, B. Garcia, H. A. Parada G., "System Virtualization Tools for

Software Development," IEEE Internet Computing, pp. 52-59, September/October, 2009.

• Chaudhary, V., Minsuk Cha, Walters, J.P., Guercio, S.; Gallo, S. "A Comparison of Virtualization

Technologies for HPC,". In Proc. on 22nd Int Conf on Advanced Information Networking and

Applications, 2008 (AINA 2008). pp.861-868, 25-28 March 2008 doi: 10.1109/AINA.2008.45


Bibliography

• R. A. Meyer and L. H. Seawright."A virtual machine time-sharing system". IBM Systems Journal,

Vol.9 Issue 3, IBM Corp. Riverton, NJ, USA. Sept. 1970.

• K. Adams and O. Agesen. ” A comparison of software and hardware techniques for x86

virtualization.” In Proceedings of the 12th international conference on Architectural support for

programming languages and operating systems, October 21-25, 2006, San Jose, California, USA.

• J. Sahoo, S. Mohapatra, R. Lath, "Virtualization: A Survey on Concepts, Taxonomy and Associated

Security Issues." In Proceedings of the Second International Conference on Computer and

Network Technology, pp. 222-226. 2010.

• A. Cleeff, W. Pieters, R.J. Wieringa, "Security Implications of Virtualization: A Literature Study." In

Proceedings of the IEEE Intl. Conf. on Computational Science and Engineering, pp. 353- 358, 2009.

• Sanderson, Dan. Programming Google App Engine. O'Reilly Media, Inc., 2009.

• Roche, Kyle , Valdez, Iein, Douglas, Jeff. Beginning Java Google App Engine. Apress, 2009.

• Moore, Dana, Goodwill, James, Lan, Ikai. Professional Google App Engine Programming with Java.

John Wiley & Sons, 2010.

• Smeets, Bram, Boness, Uri, Bankras, Roald. Beginning Google Web Toolkit: From Novice to

Professional. Apress, 2008.

• Dewsbury, Ryan. Google web toolkit applications. Prentice Hall, 2008.

• Google Code. Google App Engine, Developer's Guide. http://code.google.com/appengine/docs/

07/07/2016



Background: Network-centric computing

� Information processing can be done more efficiently on large farms ofcomputing and storage systems accessible via the Internet.

� Grid computing – initiated by the National Labs in the early 1990s;targeted primarily at scientific computing.

� Utility computing – initiated in 2005-2006 by IT companies and targeted atenterprise computing.

� The focus of utility computing is on the business model for providingcomputing services; it often requires a cloud-like infrastructure.

� Cloud computing is a path to utility computing embraced by major ITcompanies including: Amazon, HP, IBM, Microsoft, Oracle, and others.


Network-centric computing and content

� Data-intensive: large scale simulations in science and engineeringrequire large volumes of data. Multimedia streaming transfers largevolume of data.

� Network-intensive: transferring large volumes of data requireshigh bandwidth networks.

� Low-latency networks for data streaming, parallel computing,computation steering.

� The systems are accessed using thin clients running on systemswith limited resources, e.g., wireless devices such as smart phonesand tablets.

� The infrastructure should support some form of workflow

management.

07/07/2016



Evolution of concepts and technologies

� The concepts and technologies for network-centric computing andcontent evolved along the years.

� The web and the semantic web - expected to support composition ofservices. The web is dominated by unstructured or semi-structureddata, while the semantic web advocates inclusion of sematic content inweb pages.

� The Grid - initiated in the early 1990s by National Laboratories andUniversities; used primarily for applications in the area of science andengineering.

� Peer-to-peer systems.

� Computer clouds.


Cloud Computing (CC)

� Nowadays there is not a unique definition for CloudComputing (CC).

� The following definition of CC is literally taken from onepublished by the U.S. National Institute of Standards andTechnology (NIST)

“”Cloud computing is a model for enabling convenient, on-

demand network access to a shared pool of configurable

computing resources (e.g., networks, servers, storage,

applications, and services) that can be rapidly provisioned and

released with minimal management effort or service provider

interaction. This cloud model promotes availability and is

composed of five essential characteristics, three service

models, and four deployment models.””

07/07/2016



CC in General Terms

� A new class of network based computing that takesplace over the Internet,

� basically a step on from Utility Computing

� a collection/group of integrated and networkedhardware, software and Internet infrastructure (called aplatform).

� Using the Internet for communication and transportprovides hardware, software and networking servicesto clients

� These platforms hide the complexity and details of theunderlying infrastructure from users and applications byproviding very simple graphical interface or API(Applications Programming Interface).


� Provides on demand services, that are alwayson, anywhere, anytime and any place.

� Pay for use and as needed, elastic

� scale up and down in capacity and functionalities

� The hardware and software services are availableto

� general public, enterprises, corporations andbusinesses markets

CC in General Terms

07/07/2016



� Shared pool of configurable computing resources

� On-demand network access

� Provisioned by the Service Provider

CC in General Terms


• Cloud Computing integrates different technologies:

07/07/2016



Cloud Computing CharacteristicsEssential Characteristics:

Resource Pooling

Broad Network Access Rapid Elasticity

Measured Service

On Demand Self-Service

Common Characteristics:

Low Cost Software

Virtualization Service Orientation

Advanced Security

Homogeneity

Massive Scale Resilient Computing

Geographic Distribution


Cloud activities

� Service management and provisioning including:� Virtualization.

� Service provisioning.

� Call center.

� Operations management.

� Systems management.

� QoS management.

� Billing and accounting, asset management.

� SLA management.

� Technical support and backups.

07/07/2016



Cloud activities (cont’d)

� Security management including:� ID and authentication.

� Certification and accreditation.

� Intrusion prevention.

� Intrusion detection.

� Virus protection.

� Cryptography.

� Physical security, incident response.

� Access control, audit and trails, and firewalls.


Cloud activities (cont’d)

� Customer services such as:� Customer assistance and on-line help.

� Subscriptions.

� Business intelligence.

� Reporting.

� Customer preferences.

� Personalization.

� Integration services including:� Data management.

� Development.

07/07/2016


NIST cloud reference modelCarrier

S

e

c

u

r

i

t

y

P

r

i

v

a

c

y

Service

Consumer BrokerService Provider

Auditor

Security

audit

Privacy

impact audit

Performance

audit

Service

Management

Business

support

Provisioning

Portability/

Interoperability

IAASIaaS

SaaS

Service Layer

PaaS

Carrier

Hardware

Facility

Physical resource

layer

Resource

abstraction and

control layer

Intermediation

Aggregation

Arbitrage


The Cloud Computing Use Cases Discussion Group, a group of expert whose aim is to bring together cloud consumers and cloudvendors to define common use cases for CC, has proposed a taxonomy for CC. In this taxonomy, they visualize three main actorsthat can play different roles: Service Consumers that use the services provided through the cloud, Service Providers that managethe cloud infrastructure and Service Developers that create the services themselves. They also emphasize the need for openstandards that can make possible the interaction between these actors

Cloud Taxonomy

07/07/2016



Cloud Service Model


Cloud Service Model: Responsabilities

07/07/2016




Cloud Service Models: ProvidersSoftware as a

Service (SaaS)Platform as a

Service (PaaS)Infrastructure as a

Service (IaaS)

Google App Engine

SalesForce CRM

LotusLive

07/07/2016



Providers

Application Service(SaaS)

Application Platform

Server Platform

Storage Platform Amazon S3, Dell, Apple, ...

3Tera, EC2, SliceHost, GoGrid, RightScale, Linode

Google App Engine, Mosso,Force.com, Engine Yard,Facebook, Heroku, AWS

MS Live/ExchangeLabs, IBM, Google Apps; Salesforce.comQuicken Online, Zoho, Cisco

23


Services

Application

Development

Platform

Storage

Hosting

CC Service Layers Examples

Description

Services – Complete business services such as PayPal, OpenID, OAuth, Google Maps, Alexa

Services

ApplicationFocused

InfrastructureFocused

Application – Cloud based software that eliminates the need for local installation such as Google Apps, Microsoft Online

Storage – Data storage or cloud based NAS such as CTERA, iDisk, CloudNAS

Development – Software development platforms used to build custom cloud based applications (PAAS & SAAS) such as SalesForcePlatform – Cloud based platforms, typically provided using virtualization, such as Amazon ECC, Sun Grid

Hosting – Physical data centers such as those run by IBM, HP, NaviSite, etc.

24

07/07/2016



� All Software as a Service (SaaS) applications share thefollowing characteristics:1. The software is available over the Internet globally through a

browser on demand2. The typical license is subscription-based or usage-based and is

billed on a recurring basis.3. In a small number of cases a flat fee may be changed, often

coupled with a maintenance fee.4. The software and the service are monitored and maintained by

the vendor, regardless of where all the different softwarecomponents are running. Theremay be executable client-side code, but the user isn'tresponsible for maintaining that code or its interaction with theservice.

5. Reduced distribution and maintenance costs and minimal end-user system costs generally make SaaS applications cheaperto use than their shrink-wrapped versions.

SaaS Characteristics


6. SaaS applications feature automated updates and patchmanagement and much faster rollout of changes.

7. SaaS applications often have a much lower barrier toentry than their locally installed competitors, a knownrecurring cost, and they scale on demand (a property ofcloud computing in general).

8. All users have the same version of the software so eachuser's software is compatible with another's.

9. SaaS supports multiple users and provides a shared datamodel through a single-instance, multi-tenancy model.

SaaS Characteristics (Cont..)

07/07/2016



SaaS Maturity Model

Level 2: Configurable percustomer

Level 3: configurable & Multi-Tenant-Efficient

Level 1: Ad-Hoc/Custom –One Instance per customer

Level 4: Scalable, Configurable & Multi-Tenant-Efficient


Licensing: Shrink-Wrapped vs SaaS

07/07/2016



Platform-as-a-Service (PaaS) - AppEngine

Cloud Service Model: PaaS


PaaS tool: Win Azure

07/07/2016



Paas tool: GAE


Infrastructure

Automation

IaaS

Cloud Service Model: IaaS

07/07/2016



Cloud Service Model: IaaS


� Private.

� Exclusive use of a company (could be rented)

� Community

� A set of institutions share its infrastructure

� Public.

� Open to everyone that pays for the service

� Hybrid.

� A combination of public and private

Cloud Computing: Deployment

07/07/2016



Public Cloud


Private Cloud

07/07/2016



Community Cloud


Hybrid Cloud

07/07/2016



Cloud Storage

� Several large Web companies are nowexploiting the fact that they have data storagecapacity that can be hired out to others.

� allows data stored remotely to be temporarily cachedon desktop computers, mobile phones or otherInternet-linked devices.

� Amazon’s Elastic Compute Cloud (EC2) andSimple Storage Solution (S3) are well knownexamples

� Mechanical Turk

39


Amazon Simple Storage Service (S3)

� Unlimited Storage.

� Pay for what you use:

� $0.20 per GByte of data transferred,

� $0.15 per GByte-Month for storage used,

� Second Life Update:

� 1TBytes, 40,000 downloads in 24 hours - $200,

40

07/07/2016



Utility Computing – EC2

� Amazon Elastic Compute Cloud (EC2):

� Elastic, marshal 1 to 100+ PCs via WS,

� Machine Specs…,

� Fairly cheap!

� Powered by Xen – a Virtual Machine:

� Different from Vmware and VPC as uses “para-virtualization”where the guest OS is modified to use special hyper-calls:

� Hardware contributions by Intel (VT-x/Vanderpool) and AMD(AMD-V).

� Supports “Live Migration” of a virtual machine between hosts.

� Linux, Windows, OpenSolaris

� Management Console/AP

41


EC2 – The Basics

� Load your image onto S3 and register it.

� Boot your image from the Web Service.

� Open up required ports for your image.

� Connect to your image through SSH.

� Execute you application…

42

07/07/2016



Why cloud computing could be successful when other paradigms have failed?

� It is in a better position to exploit recent advances in software, networking,storage, and processor technologies promoted by the same companieswho provide cloud services.

� It is focused on enterprise computing; its adoption by industrialorganizations, financial institutions, government, and so on could have ahuge impact on the economy.

� A cloud consists of a homogeneous set of hardware and softwareresources.

� The resources are in a single administrative domain (AD). Security,resource management, fault-tolerance, and quality of service are lesschallenging than in a heterogeneous environment with resources inmultiple ADs.


Challenges for cloud computing

� Availability of service; what happens when the service providercannot deliver?

� Diversity of services, data organization, user interfaces availableat different service providers limit user mobility; once a customer ishooked to one provider it is hard to move to another.Standardization efforts at NIST!

� Data confidentiality and auditability, a serious problem.

� Data transfer bottleneck; many applications are data-intensive.

07/07/2016



More challenges

� Performance unpredictability, one of the consequences of resourcesharing.

� How to use resource virtualization and performance isolation for QoSguarantees?

� How to support elasticity, the ability to scale up and down quickly?

� Resource management; are self-organization and self-managementthe solution?

� Security and confidentiality; major concern.

� Addressing these challenges provides good research

opportunities!!


Ethical issues

� Paradigm shift with implications on computing ethics:

� The control is relinquished to third party services.

� The data is stored on multiple sites administered by severalorganizations.

� Multiple services interoperate across the network.

� Implications

� Unauthorized access.

� Data corruption.

� Infrastructure failure, and service unavailability.

07/07/2016



De-perimeterisation

� Systems can span the boundaries of multiple organizations and crossthe security borders.

� The complex structure of cloud services can make it difficult todetermine who is responsible in case something undesirable happens.

� Identity fraud and theft are made possible by the unauthorized accessto personal data in circulation and by new forms of disseminationthrough social networks and they could also pose a danger to cloudcomputing.


Privacy issues

� Cloud service providers have already collected petabytes ofsensitive personal information stored in data centers around theworld. The acceptance of cloud computing therefore will bedetermined by privacy issues addressed by these companies andthe countries where the data centers are located.

� Privacy is affected by cultural differences; some cultures favorprivacy, others emphasize community. This leads to an ambivalentattitude towards privacy in the Internet which is a global system.

07/07/2016



Cloud vulnerabilities

� Clouds are affected by malicious attacks and failures of theinfrastructure, e.g., power failures.

� Such events can affect the Internet domain name servers andprevent access to a cloud or can directly affect the clouds:

� in 2004 an attack at Akamai caused a domain name outageand a major blackout that affected Google, Yahoo, and othersites.

� in 2009, Google was the target of a denial of service attackwhich took down Google News and Gmail for several days;

� in 2012 lightning caused a prolonged down time at Amazon.


Advantages of Cloud Computing

� Lower computer costs:� You do not need a high-powered and high-priced computer

to run cloud computing's web-based applications.

� Since applications run in the cloud, not on the desktop PC,your desktop PC does not need the processing power orhard disk space demanded by traditional desktop software.

� When you are using web-based applications, your PC can beless expensive, with a smaller hard disk, less memory, moreefficient processor...

� In fact, your PC in this scenario does not even need a CD orDVD drive, as no software programs have to be loaded andno document files need to be saved.

50

07/07/2016




� Improved performance:

� With few large programs hogging your computer's memory, youwill see better performance from your PC.

� Computers in a cloud computing system boot and run fasterbecause they have fewer programs and processes loaded intomemory…

� Reduced software costs:

� Instead of purchasing expensive software applications, you canget most of what you need for free-ish!

� most cloud computing applications today, such as the Google Docs suite.

� better than paying for similar commercial software� which alone may be justification for switching to cloud applications.

51



� Instant software updates:

� Another advantage to cloud computing is that you are no longer facedwith choosing between obsolete software and high upgrade costs.

� When the application is web-based, updates happen automatically� available the next time you log into the cloud.

� When you access a web-based application, you get the latest version� without needing to pay for or download an upgrade.

� Improved document format compatibility.

� You do not have to worry about the documents you create on yourmachine being compatible with other users' applications or OSes

� There are potentially no format incompatibilities when everyone issharing documents and applications in the cloud.

52

07/07/2016




� Unlimited storage capacity:

� Cloud computing offers virtually limitless storage.

� Your computer's current 1 Tbyte hard drive is small compared tothe hundreds of Pbytes available in the cloud.

� Increased data reliability:

� Unlike desktop computing, in which if a hard disk crashes anddestroy all your valuable data, a computer crashing in the cloudshould not affect the storage of your data.

� if your personal computer crashes, all your data is still out there inthe cloud, still accessible

� In a world where few individual desktop PC users back up theirdata on a regular basis, cloud computing is a data-safecomputing platform!

53



� Universal document access:

� That is not a problem with cloud computing, because you do nottake your documents with you.

� Instead, they stay in the cloud, and you can access themwhenever you have a computer and an Internet connection

� Documents are instantly available from wherever you are

� Latest version availability:

� When you edit a document at home, that edited version is whatyou see when you access the document at work.

� The cloud always hosts the latest version of your documents� as long as you are connected, you are not in danger of having an outdated

version

54

07/07/2016




� Easier group collaboration:

� Sharing documents leads directly to better collaboration.

� Many users do this as it is an important advantages of cloudcomputing

� multiple users can collaborate easily on documents and projects

� Device independence.

� You are no longer tethered to a single computer or network.

� Changes to computers, applications and documents follow youthrough the cloud.

� Move to a portable device, and your applications and documentsare still available.

55


Disadvantages of Cloud Computing

� Requires a constant Internet connection:

� Cloud computing is impossible if you cannot connectto the Internet.

� Since you use the Internet to connect to both yourapplications and documents, if you do not have anInternet connection you cannot access anything, evenyour own documents.

� A dead Internet connection means no work and inareas where Internet connections are few orinherently unreliable, this could be a deal-breaker.

56

07/07/2016




� Does not work well with low-speed connections:

� Similarly, a low-speed Internet connection, such as that foundwith dial-up services, makes cloud computing painful at best andoften impossible.

� Web-based applications require a lot of bandwidth to download,as do large documents.

� Features might be limited:

� This situation is bound to change, but today many web-basedapplications simply are not as full-featured as their desktop-based applications.

� For example, you can do a lot more with Microsoft PowerPoint thanwith Google Presentation's web-based offering

57



� Can be slow:

� Even with a fast connection, web-based applications cansometimes be slower than accessing a similar software programon your desktop PC.

� Everything about the program, from the interface to the currentdocument, has to be sent back and forth from your computer tothe computers in the cloud.

� If the cloud servers happen to be backed up at that moment, or ifthe Internet is having a slow day, you would not get theinstantaneous access you might expect from desktopapplications.

58

07/07/2016




� Stored data might not be secure:

� With cloud computing, all your data is stored on the cloud.

� The questions is How secure is the cloud?

� Can unauthorised users gain access to your confidential data?

� Stored data can be lost:

� Theoretically, data stored in the cloud is safe, replicated acrossmultiple machines.

� But on the off chance that your data goes missing, you have nophysical or local backup.

� Put simply, relying on the cloud puts you at risk if the cloud lets youdown.

59



� HPC Systems:

� Not clear that you can run compute-intensive HPC applications thatuse MPI/OpenMP!

� Scheduling is important with this type of application

� as you want all the VM to be co-located to minimize communicationlatency!

� General Concerns:

� Each cloud systems uses different protocols and different APIs

� may not be possible to run applications between cloud based systems

� Amazon has created its own DB system (not SQL 92), and workflowsystem (many popular workflow systems out there)

� so your normal applications will have to be adapted to execute on theseplatforms.

60

07/07/2016



The Future

� Many of the activities loosely grouped together undercloud computing have already been happening andcentralised computing activity is not a new phenomena

� Grid Computing was the last research-led centralisedapproach

� However there are concerns that the mainstreamadoption of cloud computing could cause many problemsfor users

� Many new open source systems appearing that you caninstall and run on your local cluster

� should be able to run a variety of applications on these systems

61


� Messaging and team collaboration applications

� Cross enterprise integration projects

� Infrastructure consolidation, server, and desktopvirtualization efforts

� Web 2.0 and social strategy companies

� Web content delivery services

Adopters of cloud computing

07/07/2016



� Data analytics and computation

� Mobility applications for the enterprise

� CRM applications

� Experimental deployments, test bed labs, anddevelopment efforts

� Backup and archival storage

Adopters of cloud computing


Tools from Public Clouds Providers

07/07/2016



Commercial and Open-Source Software tools for virtual infrastructure management

. . . . . . . .. .


. . . . . . . .. .


07/07/2016





Initiatives for Standards

Interoperability-group initiatives

07/07/2016




Interoperability –vendor initiatives


Initiatives for StandardsSecurity –group initiatives

Security –vendor initiatives

07/07/2016




Portability –group initiatives

Portability –vendor initiatives


Governance, risk and compliance -vendor initiatives


07/07/2016




Other key standards-Group initiatives