A Cost Efficient Multi-Cloud Data Hosting Scheme Using Charm

Internatonal Journal of Information Technology and ComputerEngineering
Special Issue: NCRASET-16 e-ISSN: 2455-529
A COST EFFICIENT MULTI-CLOUD DATA HOSTING SCHEME

USING CHARM
Priyam Kumar1, Nitish Das 2, Saransh Srivastava 3, Dr. Savita Choudhary 4
1
UG Student, Department of CS&E, Sir MVIT, Bangalore, India
2
UG Student, Department of CS&E, Sir MVIT, Bangalore, India
3
UG Student, Department of CS&E, Sir MVIT Bangalore, India
4
Assistant Professor, Department of CS&E, Sir MVIT Bangalore, India, priyam08@gmail.com
ABSTRACT
In the ever increasing global modern trends, the global deployment of cloud datacenters is enabling
improved performance and delivery of fast responses. More and more enterprises and organizations are
hosting their data into the cloud, to reduce the IT maintenance cost and enhance the reliability of data. This
unprecedented geographical distribution of the computation is doubled by an increase in the scale of the
data handled by such applications, bringing new challenges related to the efficient data management and
storage schemes across sites. High throughput, low latencies or cost-related trade-offs are just a few
concerns for the cloud services users when it comes to handling data across datacenters. Existing solutions
are limited to cloud-provided storage, which offers low performance based on rigid cost schemes.
In this paper, we expand the scope for CHARM, providing a Cost-efficient, Multi-cloud Data Hosting
Scheme that ensures High Availability, also integrating the two key functions that we have a need for. The
first key function is about selecting several suitable clouds and a suitable redundancy strategy to store data
with minimized cost and guaranteed availability, while the second major technique is the triggering of the
transition process to re- distribute data based on the variations of data access pattern and pricing of clouds.
On an evaluation of the performance of CHARM, the results show that our system is able to model
accurately the cloud performance and to leverage this for efficient data dissemination, being able to reduce
the monetary costs and transfer time.
Keywords—Cloud computing; Multi-Cloud; Data Hosting; CHARM; File Uploading with Hashing; Storage
mode Switching; Cost Process; File Download
---------------------------------------------------------------------------------------------------------------------------
1. INTRODUCTION
Cloud computing, is a kind of Internet-based computing that provides shared processing resources and data
to the networking services users and their associates on demand. It is a model which works for the enabling of
all-pervasive, that is, global or worldwide on-demand access to a shared pool of configurable computing
resources (e.g., networks, servers, storage, applications and services), which can be rapidly stipulated and
released with minimal management effort. Cloud computing allows consumers and businesses to use
applications without installation and access their personal files at any computer with internet access. This
technology allows for much more efficient computing by centralizing data storage, processing and bandwidth at
third party data centers.
The recent years have experienced a severe surge in the need and the usage of the cloud services. The cloud
computing environment with their globally distributed datacenters, provide cloud infrastructures that enable
the rapid development of large scale applications. Examples of such applications running as cloud services
between sites ranging from office collaborative tools (Google Drive, Microsoft Office 365), to search engines
(Bing, Google, Ask.com), and global stock market analysis tools to entertainment services (e.g., sporting and
live events broadcasting, parallel games, news and data mining) and scientific workflows. Most of these
applications are deployed on massively parallel sites to leverage proximity to users through content delivery
networks.
98
IJITC| www.ijitc.org|editor@ijitc.org
Over the years, as the enterprise awareness of the cloud has increased, more and more enterprises are
moving towards the outsourcing of their workload deployments to the cloud, in most cases to a single cloud
provider or vendor. However, with the market changing rapidly, greater options are becoming available from a
variety of cloud service providers, with public offerings such as Amazon, HP, IBM, RackSpace as well as
private offerings such as Openstack and VMware. The new deployment options make it possible for the clients
to mix and match their suiting platforms and cloud providers, as well as to set up hybrid clouds which ensures
the resources in an on-premises datacenter or private cloud while migrating parts of the workload to one or
more public clouds.
Heterogeneity in services - Current cloud computing services infrastructure typically surmises a

homogeneous collection of commodity hardware, wherein the data and computational centers are often
limited by power and compute density and efficiency. These exhibit great heterogeneities while considering
both the working performances and the pricing policies. The various cloud vendors build and design their
own respective infrastructures, and keep upgrading with various techniques to make their services
competitive. This diversity is responsible for the variety in the observed performance and differential pricing
for the cloud services providers. The pricing schemes are distinct in both pricing
figure 1. basic principle of multi-cloud data hosting
levels and charging items depending on the web operations, bandwidth consumption and storage spaces [1].
Multi-Cloud strategy is the associated use of two or more cloud services to distribute data across
multiple clouds to gain enhanced redundancy and to minimize the risk of widespread data loss or downtime due
to a localized component failure in a cloud computing environment. Such a failure can occur in hardware,
software, or infrastructure. Such a strategy also improves the overall enterprise performance by avoiding
potential risks of prohibition issues such as "vendor lock-in" and using different infrastructures to meet the
needs of diverse partners and customers [2].
Vendor lock-in risk. The widespread presence of a numerous cloud vendors as well as their heterogeneous
performances and policies schemes, customers may be disconcerted with the right set of clouds that may be
suitable for storing their data and choosing which hosting strategy is cheaper. The general status quo being that
customers generally put all of their data into a single cloud service and then simply depend on luck. This is
subject to the so-called “vendor lock-in risk”, wherein customers may be baffled if they want to switch to other
cloud venders.
Vendor lock-in can be of three types with respect to cloud computing,

-in where the cloud services incline to be assembled on one of some of the
feasible virtualization platform such as VMWare or Xen and the migrating of a user from a cloud provider using
one platform to a cloud provider using a different platform could be very complicated.
-in, where since the cloud‟s standards of ownership, are not yet developed, make
it complicated if cloud computing user decides to move data off a cloud vendor‟s platform.
-in where if the tools built to manage a cloud computing environment are not
compatible with different versions of both virtual and physical infrastructure, these implements tend to solely be
capable to run information or to manage data or apps that live in the vendor's particular cloud environment [4][5]
99
2. PROBLEM DESCRIPTION
A. Existing System
Z. Li, C. Jin, T. Xu, C. Wilson, Y. Liu, L. Cheng, Y. Liu, Y. Dai,and Z.-L. Zhang [3], in this paper the
author introduced the various advantages of Cloud storage services like Google drive, Drop box and many other.
In the existing industrial and enterprise based data hosting systems, data availability are usually guaranteed by
replication or erasure coding. For replication, replicas are put into several clouds, and a read access is only
served by the cheapest cloud, such that the cloud that charges a minimal amount for out-going bandwidth and
GET operation [6].
Data replication is generally adopted for systems with Distributed Applications that are hosted in a
Hybrid Cloud and often require access to the same data from various parallel application components. If the
application components accessing the data are globally distributed, then the data access performance for the
parallel systems may be reduced drastically, comparing if data is only stored in one geographic location. Some
of these locations may only handle a subset of the available data or the data has to be obfuscated, based on the
set of governing and operating laws and corporate regulations
.
The other method is the erasure coding. An erasure code is a forward error correction (FEC) code for
the binary erasure channel, in which a message of k symbols is transformed into a longer message (code word)
with n symbols, such that the original message can be recovered only from a subset of the n symbols. Here, the
fraction r = k/n is called the code rate, the fraction k‟/k, is called the reception efficiency, where k‟ denotes the
number of symbols required for recovery. In cloud computing, data is encoded into „n‟ blocks including „m‟
data blocks and „n [] m‟ coding blocks, and these blocks are put into „n‟ different clouds. In this case, the data
availability can be guaranteed within a lower storage space, but a read access has to be served by multiple
clouds that store the corresponding data blocks [7].
In the multi-cloud scenario also, the replication techniques and the erasure coding mechanisms are used
to meet different availability requirements, but the implementation of these are very different
.
B. Disadvantages of the Existing System
security and availability of the system,
user data vulnerable to potential attacks, A quarantined and secure hybrid architecture ensuring a dependable
and reliable service is missing.
3. PROPOSED MODEL
A. The proposed CHARM scheme

In this paper, we propose using a novel, efficient, and heuristic-based data hosting scheme for
heterogeneous multi-cloud environments with high availability, named “CHARM”. It intelligently
accommodates data into multiple clouds with minimized monetary cost and guaranteed availability, considering
the differential pricing strategies, availability requirements, and data access patterns. It selects suitable clouds
and an appropriate redundancy strategy to store data, while monitoring the variations of pricing policies and data
access patterns, and adaptively triggers the transition process between different data storage modes.
In CHARM, the two widely used redundancy mechanisms, i.e., replication and erasure coding, are combined
into an uniform model to meet the required availability in the presence of different data access patterns.
B. System Modules
1) MULTI-CLOUD
To provide for the cloud computing services, several regions are divided around the globe, and every region is
usually provided with several data centers that belong to the same or different cloud providers. Technically, a
user of the system may access any of the data centers in a certain region, but the user would experience an
altogether different performance.
100
Figure 2 CHARM SYSTEM

ARCHITECTURE
Figure 3.THE ARCHITECTURE OF CHARM. "R" REPRESENTS REPLICATION AND

"E" REPRESENTS ERASURE CODING
The latency of some data centers may be very low while that of others may be intolerably high. CHARM
chooses a set of clouds for storing data from all the available clouds which meet the performance requirement,
that is, they can offer acceptable throughput and latency when they are not in outage.
1) DATA HOSTING
The data hosting model is located in the proxy. There are four main components in CHARM: Data Hosting,
Storage Mode Switching (SMS), Workload Statistic, and Predictor. Workload Statistic keeps collecting and
tackling access logs to guide the placement of data. It sends the statistic information to Predictor which guides
the action of Storage Mode Switching. Data Hosting stores data using, the two widely used redundancy
mechanisms, replication and erasure coding, depending upon the size and access frequency of the data. SMS
decides whether the storage mode of the data should be changed from replication to erasure coding or in
reverse, according to the output of Predictor. The implementation of changing storage mode runs in the
background, without disturbing the Web services. Predictor is used to predict the future access frequency of
files.
2) FILE UPLOADING WITH HASHING
In the File Uploading module, the user selects the file to be uploaded to the cloud by selecting the number of
copies of replication required to be stored .While uploading, the charm server reads the file size, while,
selecting the best cloud storage server, based on the storage availability, pricing cost, predictor and size. For
the Integrity Verification process, it generates the Hash Key (HK1) using MD5 algorithm and keeps it in the
101
user
Figure 4. FILE UPLOADING
Database. Finally based on the Replication details the File will be stored on the Cloud Storage Server.
3) CLOUD STORAGE SECURITY
CHARM uses the availabilities declared in the SLAs of cloud services. Most of the proposed schemes assume
cloud storage service providers or the third parties handling key management, are trusted and cannot be hacked.
Some entities may however, intercept communications between the users and cloud storage providers and then
convince the storage providers to release user secrets by using illegal practices. In this case, encrypted data are
assumed to be known and storage providers are requested to release user secrets. To avoid such situations,
special kind of encryption scheme called deniable encryption scheme is employed that aims at building an
encryption scheme that could help cloud storage providers avoid this predicament. In this approach, the storage
providers create fake user account, from which the outside coercers can only obtain forged data from a user‟s
stored ciphertext.
4) FILE DOWNLOAD
When the user wants to download the file from the Multi Cloud Storage Server, user has to select the file from
the Charm Application, where the Charm Server will select the corresponding cloud sever details from the
associated database and also, will check for the cloud availability for downloading. If the cloud is not available,
then it will download from another cloud server. While downloading it, the server will generate the HASH Key
(HK2), which will check the HK1 &HK2 for the Integrity Check. Finally the file would be downloaded to the
user local System.
5) COST PROCESS
We generate the storage mode table based on the data available about the various clouds guaranteeing higher
availability. We use data for different file sizes varying from 1KB to 1GB and varying read counts to calculate
their corresponding storage modes. This storage mode table depends on prices of the available clouds and
required availability. If the prices of the clouds change, the table will change accordingly. The total costof
CHARM includes storage/bandwidth/operation costs and transition cost. When the price adjustment occurs,
CHARM re- calculates the storage mode table, and uses it to store data and implement transition. This way it
demonstrates great adaptability, when compared with the various other schemes that cannot change the clouds
as dynamically depending upon the price fluctuation.
B. Advantages
 A multi-cloud approach is where an enterprise uses two or more cloud services, hence minimizing
the risk of widespread data loss or outage due to a component failure, as in a single cloud computing
environment.
 Data Hosting with low cost cloud storage Server will make effective use for client benefit.
4. CONCLUSION
This paper expands CHARM, a Cost-efficient, Multi-cloud Data Hosting Scheme that ensures High
Availability, a novel storage scheme which guides customers to distribute data among clouds efficiently in large,
102
geographically distributed and highly dynamic environments. As the demand for decentralisation and
delocalisation of data is growing, Cloud services are experiencing rapid development and the services based on
multi-cloud also become prevailing. A major challenge to moving services to the clouds is the capital
involvement. CHARM provides a superior solution to the complexities involving as to which storage mode to
use and which clouds to place data in. It appeals to the particularities of multi-cloud environment, and hosts data
into multiple clouds cost-effectively, while guaranteeing flexible availability and avoiding vender lock-in.
REFERENCES
[1] Steve Crago, Kyle Dunn, Patrick Eads, Lorin Hochstein, Dong-In Kang, Mikyung Kang, et al --
“Heterogeneous cloud computing” --- University of Southern California / Information Sciences Institute
3811 N. Fairfax, Suite 200. Arlington, VA, 22203 {crago∗, lorin† }@isi. edu,
[2] Elpula Vijaya Kumar, V. Rama Krishna, P. Raja Sekhar Reddy -- “A Novel Approach for Multi Cloud
Storage” --- CSE Dept., AGI, India,
[3] Z. Li, C. Jin, T. Xu, C.Wilson, Y. Liu, L. Cheng, Y. Liu,
Y. Dai, and Z.-L. Zhang, -- “Towards network-level efficiency for cloud storage services,” -- in Proceedings
of the 2014 Conference on Internet Measurement Conference, pp. 115– 128, ACM, 2014
[4] Cloud Computing Issues -- https://en.wikipedia.org/wiki/Cloud_computing_issues [5] Vendor Lock-
In Issues –
http://emereo.net/success/cloud-infrastructure-vendor-lock-in/ [6] Dejene Boru, Dzmitry Kliazovich, Fabrizio
Granelli, Pascal Bouvry, Albert Y. Zomaya – “Energy-efficient data replication in cloud computing
datacenters”
[7] Jun Li and Baochun Li – “Erausre Coding For Cloud Storage Systems”
.
103

A Cost Efficient Multi-Cloud Data Hosting Scheme Using Charm

Uploaded by

A Cost Efficient Multi-Cloud Data Hosting Scheme Using Charm

Uploaded by

Internatonal Journal of Information Technology and ComputerEngineering

Special Issue: NCRASET-16 e-ISSN: 2455-529

A COST EFFICIENT MULTI-CLOUD DATA HOSTING SCHEME

Priyam Kumar1, Nitish Das 2, Saransh Srivastava 3, Dr. Savita Choudhary 4

Heterogeneity in services - Current cloud computing services infrastructure typically surmises a

figure 1. basic principle of multi-cloud data hosting

Vendor lock-in can be of three types with respect to cloud computing,

security and availability of the system,

A. The proposed CHARM scheme

Figure 2 CHARM SYSTEM

Figure 3.THE ARCHITECTURE OF CHARM. "R" REPRESENTS REPLICATION AND

Figure 4. FILE UPLOADING

You might also like