Pervasive computing 2007 Cloud computing 2004 Utility computing

ﺭﻭﻧﺪ ﺣﺮکﺖ ﺳیﺴﺘﻢ ﻫﺎی ﻣﺤﺎﺳﺒﺎﺗی Pervasive computing … 2007 Cloud computing 2004 Utility computing Grid Computing Volunteer computing 1997 1996 P 2 P 1990 Cluster … 2

! ﺷﻤﺎیی ﺍﺯ ﺗﻮﺩﻩ ﺍﺑﺮی Salesforce Microsoft Google Yahoo The Cloud Amazon Zoho Rackspace 4

ﻣﺜﺎﻟی ﺍﺯ ﻻیﻪ ﻫﺎی ﻣﺤﺎﺳﺒﺎﺕ ﺍﺑﺮی Cloud Application (Google Docs) Cloud Service (Apps Service) Cloud Client (Firefox) Cloud Platform (Apps Engine) Big. Table Cloud Storage (Infrastructure ) (Big. Table DB) 5

ﻣﻮﺿﻮﻋﺎﺕ ﻣﻄﺮﺡ ﺩﺭ ﻣﺤﺎﺳﺒﺎﺕ ﺍﺑﺮی atio n Ene rgy Billing Reliability VM pe Scalability Pub lic bility pera o Inter cien re Wa Hy rv is or s Clou d Paa. S Storage Effi cy on E Saa. S Iaa. S Res C 2 on S 3 aliz Qo. S Amaz Virt u lity oca L a Dat Am az ng Prici es ic rv our ce M l ve e t e L en c i m rv e Provision Se gre ing A on Deman d eter W ing y urit Sec Web 2. 0 eb Se Migr ation Utility Management Man jraso ft An eka Privacy Private Clo ud Software Eng. Complexity se rpri e t En oud Cl Goo gle om e. c rc Fo les Sa Trust App Eng ine Cloud Computing Mosso ESX 6

Pricing Pay as you Go Model 7

ﺍﺳﺘﻔﺎﺩﻩ ﺑﺮ ﻣﺒﺘﻨی ﻗیﻤﺖگﺬﺍﺭی ﻣﺪﻝ n Traditional Model n n 100 servers * $1, 500 + 3 years * $13, 140 electricity/year + 3 years * 2 staff * $100, 000 salary/year = $789, 420 Pay as you go Model n 100 servers * $0. 40 instance-hour * 3 years * 8, 760 hours/year = $1, 051, 200 * 0. 75% = $788, 400 8

Pay as you Go 9

Application Architecture 11

Database sharding 15

Amazon Cloud Provider 18

Amazon Cloud Services n n n Amazon Elastic Cloud Compute (Amazon EC 2) Amazon Simple Storage Service (Amazon S 3) Amazon Simple Queue Service (Amazon SQS) Amazon Cloud. Front Amazon Simple. DB 19

Amazon Elastic Cloud Compute (EC 2) 20

Amazon Simple Storage Service (S 3) 21

Amazon Simple Queue Service (SQS) 22

Amazon Cloud. Front 23

Amazon Simple. DB 24

Aneka A Platform for Enterprise Grid/Cloud Computing

ﻫﺎ ﻭیژگی n n n Middleware for Enterprise Grids/Clouds Service oriented architecture. NET/Mono based environment n Languages: n n n Platforms: n n 2 7 C#, C++, VB, Delphi, Java/IKVM… … and 20 more languages Windows XP/2000/2003 Linux & Mac OS X

Why Aneka? n n n 2 8 Multiple programming/deployment models Multiple scheduling strategies Multiple authentication models Multiple persistence backends Multiple platform and OSs

Aneka & Clouds Software as a Service Platform as a Service Private Cloud Public Cloud Aneka Infrastructure as a Service Aneka fits into the cloud architecture at the platform layer. This means that it provides a programming based interface for developing distributed application and a virtual execution environment in which the applications developed according to the published APIs can run. 2 9

Current Applications n Scientific n n n Commercial n n n Engineering: Go Front (China): Train models rendering Media and games: platform for on-line gaming Financial: risk analysis Office automation: Excel integration Educational n n n 3 0 Distributed evolutionary computation Proteine structure prediction Image filtering Image rendering Distributed systems teaching

ﺳیﺴﺘﻢ ﻣﻌﻤﺎﺭی Aneka enterprise Cloud Aneka Container work units Executor Manager Executor internet work units Scheduler internet Manager Client Applications 3 1 Manager(s) Workers Executor

ﻧﻮیﺴی ﺑﺮﻧﺎﻣﻪ ﻫﺎی ﻣﺪﻝ n Development n n Aneka is Platform as a Service cloud middleware This means: n n n More precisely… n 3 2 It exposes an API for development It provides access to the cloud at programming level It provides different programming models

Aneka n Scenario Aneka enterprise Cloud public Dumb. Task: ITask { … public void Execute() { for(int i=0; i<n; i++) …… } { … } Dumb. Task task = new Dumb. Task(); app. Submit. Execution(task); } Executor internet work units Scheduler internet Executor Manager Programming / Deployment Model 3 3

ﻧﻮیﺴی ﺑﺮﻧﺎﻣﻪ ﻫﺎی ﻣﺪﻝ n Overview infrastructure Task Model Map. Reduce Model scheduling execution Task. Scheduler Map. Reduce. Scheduler coordination Task. Executor Map. Reduce. Executor client component Task. Manager Map. Reduce. Manager abstractions Task Mapper end users units of execution 3 4 Reducer

ﻧﻮیﺴی ﺑﺮﻧﺎﻣﻪ ﻫﺎی ﻣﺪﻝ n 3 5 Currently supported: n Task Programming Model n Thread Programming Model n Map Reduce Programming Model n Parameter Sweeping Model n. . Implement your own. .

Task Programming Model n Used to model Independent Bag of Tasks (Bo. T) applications n n n 3 6 The application is a collection of execution unit Each execution unit is not related to the others There is no order in the execution of the units

Thread Programming Model n Based on the concept of distributed thread n n Like a local thread but executed remotely Implements a subset of the common operations on thread n n n 3 7 Start Stop State Query Join Provides a quick way for porting on a distributed middleware, multi-threaded applications

Map. Reduce Programming Model n n n 3 8 Based on the Map. Reduce framework from Google Functional-style like primitives: A distributed application becomes a collection of map and reduce operations.

Map. Reduce Programming Model n Developing Map. Reduce based Applications n n n Define map and reduce operations Provide the data Run the Map. Reduce engine Input data Execution: -File staging -Task scheduling -Failed task resubmission -Replication and Fault tolerance -Collection of result Map. Reduce engine 3 9 map & reduce Map & Reduce network

Parameter Sweeping Model n n Based on the Task Programming Model Provides a set of facilities to run applications where n n n . . all the tasks are homogeneous (same task). . the specific instance of tasks is specialized by parameters. . all the possible combination of parameters are explored by generating a task instance for each of the combinations [B, 2, #] 4 0 [A, 1, #] [A, 2, #]

Saa. S Software as a Service 41

Calendar 44

Contact 45

CRM 52

58