Implementation of User-End Broker Policy to Improve the Reliability of Cloud Services

Implementation of User-End Broker Policy to Improve the Reliability of Cloud Services

Jitendra Singh (University of Delhi, New Delhi, India) and Vikas Kumar (Society for Education & Research Development, Yamunanagar, Haryana, India)
Copyright: © 2013 |Pages: 15
DOI: 10.4018/ijcac.2013100102
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Outage in cloud computing services is a critical issue and is primarily attributed to the single data center connectivity. To address the cloud outage, this work proposes a model for the subscription and selection of more than one data center. Selection of data center can be determined by the usage of broker at the user ends itself. Provision of broker at user's end reduces the overhead at provider's end; as a result performance of cloud data center improves. For the selection of appropriate data center, broker takes the feedback from the available data centers, and select one of them. During the selection of cloud, their status (up/down) at that particular time is also considered. In case of outage at one data center, other can be selected from the available list. Broker also facilitates the homogeneous use of cloud by allotting the load to less congested data centers. Experimental results revealed that multiple data center approach is not only helpful in countering the outage (as other data center can be selected from the broker) but also the usage cost.
Article Preview

1. Introduction

Cloud computing is a new style of computing, where flexible and scalable information technology (IT) related capabilities are provided as a services (Lin, Fu, Zhu, & Dasmalchi, 2009). Cloud computing deals with how IT is provisioned instead of only improvement of data centers (Creeger, 2009). It is a shift from currently used desktop to laptops and smart devices etc. which have become indispensable (Chien, Calder, Elhert & Bhatia, 2003). The economic appeal of cloud computing is often mentioned as converting capital expense (Capex) to operating expenses (Opex) (Armbrust et al., 2009). Cloud services can offer the traditional IT services with significant cost advantage (Lin et al. 2009; Pallis et al. 2010). Users can access the IT resources from anywhere and anytime using multiple devices such as computers, smart phones, tablets and laptops etc. (Park & Ryoo, 2012).

Even though cloud computing has many benefits, still it has not matured enough to cater the needs of various category of users. Therefore, enterprises are considering the benefits and other impact before adoption (Khajeh, Greenwood, & Sommerville, 2010). Experts believe that to get the full adoptions of cloud paradigm approximately 10 to 15 years are needed (Sullman, 2009). Even though cloud computing has been adopted by a number of enterprises, yet cloud paradigm has not yet reached to threshold of its maturity (CSA & ISACA, 2012). Same is evident from the recent outages, which has been witnessed in many of the prominent cloud providers. Some of the prominent cloud outages and their cloud providers have been depicted in Table 1(source: http:iwgcr.org/).

Table 1.
Cloud outages in major cloud providers
DateServiceDurationCritical Data Lost
27-Oct-12Windows Azure6 hoursNo
26-Oct-12Google App11 hoursNo
26-Oct-12Tumblr2 hoursNo
26-Oct-12Salesforce1 hourNo
22-Oct-12Salesforce20 minutesNo
19-Oct-12Windows Azure3 hoursNo
23-Oct-12Amazon12 hoursNo
28-Sep-12Windows Azure4.10 hoursNo
25-Sep-12Google Apps – Gmail2 hoursNo
10-Sep-12Windows Azure1.5 hoursNo
10-Sep-12Windows Azure1.5 hoursNo
10-Sep-12iCloud mail30 hoursNo
3-Aug-12Windows Azure0.5 hoursNo
26-Jul-12Twitter3 hoursNo
10-Jul-12SaleForce13 hoursNo
2-Jan-12MailChimp14 hoursYes
15-Feb-12Amazon<1 hourNo
29-Feb-12Windows Azure8 hoursNo
27-Mar-12OVH2 hoursNo
17-Apr-12Google Gmail1 hourNo
17-Apr-12Google Gmail1 hourNo
27-Mar-12OVH2 hoursNo
1-Mar-12Microsoft Azure8 hoursNo
9-Aug-07Cisco3 hoursNo
13-Oct-11RIM72 hoursNo

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 7: 4 Issues (2017)
Volume 6: 4 Issues (2016)
Volume 5: 4 Issues (2015)
Volume 4: 4 Issues (2014)
Volume 3: 4 Issues (2013)
Volume 2: 4 Issues (2012)
Volume 1: 4 Issues (2011)
View Complete Journal Contents Listing