A Social Compute Cloud Allocating and Sharing

7/23/2019 A Social Compute Cloud Allocating and Sharing

http://slidepdf.com/reader/full/a-social-compute-cloud-allocating-and-sharing 1/10

A Social Compute Cloud: Allocating and Sharing

Infrastructure Resources via Social Networks

Abstract:

Social network platforms have rapidly changed the way that

people communicate and interact. They have enabled the

establishment of, and participation in, digital communities as well as

the representation, documentation and exploration of social

relationships. We believe that as ‘apps’ become more sophisticated, it

will become easier for users to share their own services, resources

and data via social networks. To substantiate this, we present a Social

ompute loud where the provisioning of loud infrastructure

occurs through !friend" relationships. #n a Social ompute loud,

resource owners offer virtuali$ed containers on their personal

computer%s& or smart device%s& to their social network. 'owever, as

users may have complex preference structures concerning with whomthey do or do not wish to share their resources, we investigate, via

simulation, how resources can be effectively allocated within a social

community offering resources on a best effort basis. #n the

assessment of social resource allocation, we consider welfare,

allocation fairness, and algorithmic runtime. The key findings of this

work illustrate how social networks can be leveraged in the

construction of cloud computing infrastructures and how resources

can be allocated in the presence of user sharing preferences.



ALGORITHM

(atching )lgorithms*

ompare The +ser profile Separate ame -mail #d

reference matching )lgorithm:

/ating 0riend 1ist hecking

(atching in2 is a subset of edges(⊆- such that at most one

edge is incident to each vertexin 3

(atching )lgorithm*

) larger set of identification links across the networks.

0or all the pairs %u, v& with u ∈ 24 and v ∈ 2 and such that d24

)ssign to %u, v& a score e5ual to the numberof similarity

witnesses between u and v

#f %u, v& is the pair with highest score in whicheither u or v

appear and the score is above Tadd %u, v& to 1.



S6ST-( ))16S#S

EXISTING SYSTEM

The greedy strategy seems to providbetter welfare than the random

strategy and at the same time is computationally as efficient.While the runtime for

2)T) per allocation is around 47seconds, both random and greedy run almost

instantlyWe get similar results for the number ofunstable pairs, which are most

often lower for the greedystrategy than the random strategy here not sharing the

user friend user their our wish of not count for the 1ist of friends the system

enables resourcesharing using social networks without the exchange ofmoney and

relying on a notion of trust to avoid freeriding. 1ike our approach, they use a

virtual containerto provide virtuali$ation within the existing virtualmachine

instance, however our approach using Seattle’sprogramming level virtuali$ation

provides a muchmore lightweight model at the expense of flexibility.

PROPOSED SYSTEM

We propose using a social adapter, rather than implementing the

platform as a social net8work application as we have observed that users often

misunderstand the sep8aration between social networks and their applications

propose extensions to several well known scheduling mecha8nisms for

task assignments. Their approach considers resource endowment, and physical

network structure as core factors in the allocation problem, which are different

considerations for resource allocation. They analyse the potential of a Social loud

via simulation, using several co8authorship and friendship networks as input. They

observe how a Social loud performs based upon vari8ations in load, participation

and graph structure. 1ike 0riend 1ist ount #ncrease their 1ist of view



Advantage

Every User Can be Seeing i!e "riend ist Un#i!e $riend

ist can be increase and Decrease

%&' Many Pe&(#e i!e A Average O$ Rating S)&' T)e*

Every "riend O+r ist &$ Rating ,n&' As-





User

Social Networks

Cloud Computing

Social Cloud Computing

!reference"#ased Resource Allocation$USER

Sign Up:

In this module new user regiter the information in

order to use the Social Network$

Sign In:

In this module user can login #% using his&her userId

and password$

S&cia# Net'&r!s

"riends:

In this module can #e 'ispla%ed (ur )riend *ist

Information for Using

"riend Re.+est:

In +his ,odules New *ist (f )riends Serach +o Seeing

+hen -hen %ou *ike +o Send +he )riend Re.uest !urpose )or

Using *et *ike Us Con/rm (ur )riend *ist )or Using$

My Ga#ary:



In this module used on 'ispla%ed !hotots ,essage ,%

!ro/le Information for using In +his ,odule

Add P)&t&:

In this module each user Upload 0our !hoto And 0our

,essage Can

Share 0our )riends )or Using$

Edit Pr&$i#e:

In this module user can 1e 2dit 0our !ro/le Information

Can 0our -ish to 2dit to Update for Using$

Cloud Computing

A model for SocialNetwork information technolog%

services in which resources are retrieved from the internet

through we#"#ased tools and applications rather than a direct

connection to a server$ 'ata and software packages are stored in

servers$ 3owever cloud computing structure allows access to

information as long as an electronic device has access to the we#

Social Cloud Computing

) Social ompute loud is designed to enable ac8cess to elasticcompute capabilities provided through a cloud fabric constructed over resources

contributed by socially connected peers. ) Social loud is a form of

ommunity loud



)s the resources are owned, provided and consumed by members of a

social com8munity. Through this cloud infrastructure consumers are able to

execute programs on virtuali$ed resources that expose %secure& access to

contributed resources, and disk9storage. #, providers host sandboxed lightweight

virtual machines on which consumers can execute applications, potentially in

parallel, on their computing resources. While the concept of a Social ompute

loud can be applied to any type of virtuali$ation environment in this paper we

focus on lightweight programming %application level& virtuali$ation as this

considerably reduces overhead and the burden on providers: , however the time

to create and contextuali$e 3(s was shown to be considerable.

!reference"#ased Resource Allocation$

To support user preferences, we implement several

algorithms for bidirectional preference8based resource allocation. We compare the

runtime of these algorithms finding that for large numbers of participants and

fre5uent allocations it may be impractical to compute allocations in real8time. We

also study the effects of stochastic user participation %i.e., changing supply and

demand& when instant reallo8cation may be impossible due to constraints on

migration. We therefore introduce heuristics and compare their economic

performance against the algorithms based on metrics such as social welfare and

allocation fairness. -very +ser like friend a increase their ount 1ist #n cause

unlike their ount reduced

:

.



SYSTEM SPECI"ICATION

%ard'are Re.+ire*ents:

Syste* : Penti+* I/ 0-1 G%2-

%ard Dis! : 13 G4-

"#&((y Drive : 5-11 Mb-

M&nit&r : 516 C&#&+r M&nit&r-

M&+se : O(tica# M&+se-

Ra* : 750 Mb-

S&$t'are Re.+ire*ents:

O(erating syste* : 8ind&'s 9 U#ti*ate-

C&ding ang+age : ASP-Net 'it) C

"r&nt;End : /is+a# St+di& 0353 Pr&$essi&na#-

Data 4ase : S< Server 033=-



Documents

A Social Compute Cloud Allocating and Sharing