Comprehensive Interaction Model for Cloud Management

Cloud computing is readily being adopted by enterprises due to its following benefits: ability to provide better service to customers, improved flexibility, lower barrier to entry for an enterprise, lower maintenance cost on IT service, availability etc. However, the interaction between cloud service provider and customer is not well-defined yet. Understanding of the service offered while approaching cloud computing paradigm and also understanding of the required actions during the period of receiving a cloud service e.g. provision of new resources, scaling up/down, billing, etc. remains a concern for the enterprises. This paper proposes a segregated interaction model to manage the receiving of a cloud service in a hierarchical way. Keywords—Cloud computing; cloud management; cloud customer


I. INTRODUCTION
This Nowadays the concept of a cloud computing is one of the emerging issues in the domain of Information Technology (IT) service industry [1], [2]. Cloud computing [3], [4] is the realization of a long-held desire to offer computing (e.g. networks, servers, storage, and application services) as a service like other utilities e.g. electricity, gas, water etc. It offers a an array of solutions and advantages to business with increased flexibility, scalability and agility with reduced procedural, administrative, hardware and software costs and with higher efficiency [4], [5]. Due to these benefits, the growth in adoption of cloud has been phenomenal [6]. Gartner has the forecast that the cloud computing industry is poised for strong growth through 2014, when worldwide cloud services revenue is projected to reach $148.8 billion [7].
Cloud computing enables us to enroll into subscriptionoriented services in a "pay-as-you-go" manner by leasing IT capabilities through the Internet. As a result, new entrepreneurs no longer require investing capital for IT Infrastructure. Instead, they can take the opportunities of provisioning of resources based on usage requirement and pay based on consumption. Thus, entrepreneurs can transfer risks and maintenance aggravation at the cost of converting capital expenses to operational expenses. Cloud computing also gives an illusion of ample resources available on demand for enormous scalability. Cloud computing has been defined by many thinkers and experts in many different ways. Among them, we will use the definition given by National Institute of Standards and Technology (NIST). According to NIST: "Cloud Computing is a model for enabling ubiquitous, convenient, ondemand 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" [8].
A cloud can be public or private. A private cloud is behind the company firewall and is not avaiable to external entity whereas in a public cloud system (Fig. 1), general public or external enterprises can get computing services as utility. This public cloud now triggers the need for cloud service management on the receiving side, which is addressed by this paper. On the other hand, cloud service providers consider service delivary model e.g. Insfrastructure as a Service (IaaS), Platform as a Service (PaaS), Software as a Service (SaaS) and manage cloud by employing differernt management models e.g. consumer utilization model to maximise their utilization and hence profit [9], [10].
Many people are aware of the concept of cloud but do not have a clear understanding of the management of receiving a cloud service and thus reluctant to adopt the cloud. On the other hand, many people already have shifted to cloud but unable to gain a real benefit of being in the cloud due to absense of a proper approach. This paper offers a comprehensive interaction model for cloud customers covering the enitre interaction scenarios by using a set of use cases, so that users can approach cloud with a systematic way and can realize maximum benefits from cloud.
The remainder of the paper is structured as follows: Section 2 provides an anatomy of services provided by a cloud. A breakdown of a cloud system with a mapping to use cases is provided in Section 3. In Section 4, an interaction model for cloud management is proposed. Finally, Section 5 provides some useful concluding remarks.

II. CLOUD SERVICE
To manage a cloud, it is very essential to obtain an understanding of a typical cloud service. A cloud service is similar to a recurring business activity. A cloud user can get one or more services for a period of time, from one or multiple cloud providers. Cloud providers may charge cloud customer for their contract/consumption of resources by hour, month or year. Cloud service is defined by the Open Group as: "A service is a logical representation of a repeatable business activity that has a specified outcome, is self-contained, may be composed of other services" [11].
A cloud service is initiated, when a cloud provider makes an offering of a cloud service and makes it available for potential customers. A service is ended, when the particular service is no longer available for any customer. A typical cloud service has five states in its lifecycle (see Fig. 2): • Available: When a cloud provider makes an offering of a service along with associated cost, policies, etc. and makes it available for potential customers.
• Contract: When cloud customer(s) make contract to get a service.
• Runtime Maintenance: Cloud provider deploys resources for customer(s) as per contract i.e. makes the resources available for the customer(s) and manages the resources.
• End of Contract: When cloud customer(s) terminate the contract for the service.
• Unavailable: When the cloud provider(s) or other external event(s) make the service unavailable. Now, consider a scenario of a business organization, where IT infrastructure is required. If the organization invests on IT infrastructure more than the actual demand, then a part of the capital is wasted, whereas having less of IT capability will hamper business growth. These businesses can take the benefit of cloud by leasing IT capabilities in a systematic approach. Moreover, demand of IT infrastructure varies with time; some business face periodic demand, some face very high demand for a short period of time. Some of the common features of IT demand are shown in Fig. 3. Instead of leasing a constant IT capability, an organization can realize a tailored service(s) by following a usage pattern from old consumption data in a cloud. To realize the full benefit of being in cloud, users should be aware of all the functionalities within the system boundary and this can be achieved by following a model like we propose in this paper.

III. USE CASE
Use cases [11] are very useful to capture and communicate the functional characteristics of a system in a clear and easy way without internal details. It is a list of steps defining who does what and thus it describes the interactions between stakeholders and helps to achieve the system goal(s). A complete set of use cases can describe all the different ways to use the system and bounds the system scope. In a use case, a stakeholder is represented as an actor. 1) Actors: In a system, an entity that is able to make decisions is represented as an actor which may not necessarily be a human. In cloud system, there are two high level actors: Cloud User and Cloud Provider.
2) Cloud user: Cloud user is the representation of a person or company, who consumes cloud-based service(s) and pays in return and triggers the need for effective cloud usage management.
3) Cloud provider: Cloud provider is a representation of a company, who provides cloud-based services (infrastructure, platform, application software, etc.).
In the proposed interaction model, we have two low level actors in each high-level group: Business Manager (for business decision making) and Technical Manager (for technical decision making) as shown in Fig. 4. In the proposed model, we use the term "customer" to represent an entity that pays for receiving cloud services while user is the superset of customer. 345 | P a g e www.ijacsa.thesai.org

IV. INTERACTION MODEL
We have already defined the actors in the previous section. Now, we will define the roles of the actors to describe all the functions in a cloud system. The system functions can be categorized into three groups: business contract level, operational level and feedback level functions. In business contract level, we will describe the system functions that are attached with legal bindings. The system functions that are actually used to receive a cloud service are grouped into operational level, whereas feedback level functions feed into business contract level so that business contracts can be done or changed in an optimal way. A complete and comprehensive set of use cases are shown in Fig. 5.

A. Business Contract Level Functions
Establish Relationship: It is essential for a cloud provider to be certain of the identity of the cloud customer and provide it with security credentials (Table I).
Establish Contract: Cloud customer and provider should enter into legal bindings before providing or receiving any service. Cloud Service level Agreements (SLA) acts as a key liaison between consumers and providers on renting Anything as a Service (AaaS) [12]. In this stage, an SLA is agreed upon by both parties (Table II). Due to the dynamic nature in cloud, continuous monitoring on Quality of Service (QoS) characteristics is essential to administer SLAs [13]. While preparing SLA, the following points should be considered from the cloud customer point of view: a) Everything should be written in SLA. b) There must be a certain guarantee of performance such as service availability, response time, etc. and there must be a financial penalty in presence of failure to meet such performance level.
c) The licensing of required software(s) must be explicitly defined.
d) It is better to get service from multiple vendors to avoid vendor lock-in.
e) The time required to provision resources must be included in SLA.
f) Data ownership, confidentiality, auditability must be defined and the vendor must return data in case of change in vendor.
g) There must be provisions for change (Table III) and terminate (Table IV)

B. Operational Level Functions
Provision of Resource: By using this function, a cloud customer can reserve resources for its use within the limit mentioned in the SLA and these resources are kept aside for the particular customer until the resources are released (Table V).
Get report: Cloud customer gets a report on the resource consumption to use it for internal purpose and also these reports are used to find a pattern in its usage (Table VI).
Billing: Cloud customer gets an invoice for billing and makes the payment (Table VII).

C. Feedback Level Functions
Generate Usage Patterns: In this function, a cloud customer tries to find a pattern of its use against time. With the understanding of its internal demand, a cloud customer can make changes in SLA or get multiple different type of contracts to maximizing its goal(s) ( Table VIII). In this paper, we have considered the problem of receiving a service in a cloud computing system. We propose a segregated interaction model to manage cloud service in a hierarchical way. The developed interaction model covers all the major functions in the defined system boundary. The presented model identifies the major decision makers for cloud management and their relationship and provides a segregation of system functions to approach the considered problem easily and effectively. At the end, the model analyses the system functions with the help of use cases. In future, data mining services can be incorporated in order to analyze usage patterns more effectively [14], [15], [16], [17].