Cloud Computing Service Models

A cloud service model represents the services and the capabilities that will be offered via the cloud. The fundamental service models based on the NIST definition are Software as a Service

(SaaS), Platform as a Service (PaaS) and Infrastructure as a Service (IaaS) (Mell & Grance, 2011), which are discussed as follows:

Software as a Service (SaaS)

Software as a Service (SaaS) delivers applications running on a cloud infrastructure online and makes them accessible from different types of user devices, usually through the Internet (Mell & Grance, 2011). In this model, there is no need to install and run the application on the user’s computer and all maintenance and update procedures will be carried out by the service provider (Voorsluys et al., 2011). The SaaS model also eliminates upfront cost investment while providing users with business-specific capabilities such as email and document management (Marston et al., 2011). In this model, the users have control over their data only, while the service provider has full control of the underlying cloud infrastructure (Mell & Grance, 2011).

SaaS offers advantages over traditional packaged software as it lowers the cost of implementing the applications since there is no upfront cost required (Voorsluys et al., 2011). Additionally, maintenance and upgrade costs are decreased because these operations are carried out on the provider’s site (Mather et al., 2009). Scalability is another advantage of SaaS because organisations can scale their services based on consumer demand, using simple software configurations in a short time (Marston et al., 2011). This feature is also important for start-up and small- and medium-size organisations with minimal software requirements at the start, and can be scaled up later.

Issues regarding SaaS are mainly related to the security and confidentiality of the data since the organisation does not control the cloud infrastructure (Mather et al., 2009). Other potential issues associated with SaaS include ensuring service availability, since the access to an SaaS application is through the Internet (Mather et al., 2009); integration with other services and applications (Gupta et al., 2013); and vendor lock-in SaaS model, since the data format will be dependent on the service provider format (Phaphoom et al., 2015). When organisations plan to use cloud solutions, they often consider an SaaS model. A study by IDG Enterprise reported that SaaS applications are responsible for 67% of the organisations’ cloud payment (IDG Enterprise, 2016).

Platform as a Service (PaaS)

Platform as a Service (PaaS) offers services for the users to develop and deploy applications in the cloud environment (Mell & Grance, 2011). This model allows the users to maintain the complete application development lifecycle from designing until debugging the application (Hudaib et al., 2014). In this model, the users have control over their data and applications under development while the service provider has full control of the underlying cloud infrastructure (Mell & Grance, 2011). The potential users of PaaS are software developers since this model provides them with programming language execution environments, web servers and databases (Dai et al., 2012). PaaS enables rapid development of software applications at low costs (Mather et al., 2009). It is also useful in developing specific applications that require powerful computing resources such as Big Data analysis (Dai et al., 2012). The PaaS model can provide standardisation for application development since it allows different teams to work on a single platform (Mather et al., 2009). Additionally, some cloud providers offer consultancy in SaaS or PaaS platforms for organisations (Hudaib et al., 2014).

Cloud provider requirements are considered an issue when working with PaaS since, for example, it may be that only specific programming languages can be used (Voorsluys et al., 2011). As with SaaS, lack of portability is also considered an issue with PaaS adoption (Phaphoom et al., 2015). Since the PaaS model is relevant to a specific type of activities (i.e. application development), it is less popular than other types of cloud service models (Chuang et al., 2015). However, this model is becoming more popular in some countries with less developed IT infrastructures, such as in some Asian and Latin American countries. A study by Oxford Economics and SAP (2015) found that 58% of Asia-Pacific organisations and 34% of Latin American organisations offered PaaS platforms to their developers, compared to 23% of North American organisations.

Infrastructure as a Service (IaaS)

Infrastructure as a Service (IaaS) offers IT resources such as operating systems, servers and storage devices as a service through the Internet (Mell & Grance, 2011). In this model, the consumer can control storage, operating systems and applications but does not have any control over the underlying cloud infrastructure (Mell & Grance, 2011). The resources in this model will be utilised as virtualised resources and the consumer can control them using Virtual

Infrastructure Manager (Voorsluys et al., 2011). In IaaS, computing services are offered as utility services and the user pays for the amount of computing resources used as a pay-per-use model where there is no need for upfront IT investment (Mather et al., 2009).

The IaaS model offers advantages over the traditional IT model since it provides scalability features where IT resources can be scaled up based on actual demands more quickly than in traditional IT (Pauly, 2011). IaaS also provides cost savings because it does not require capital expenditure investment and the cost of infrastructure management comes under the cloud provider’s responsibilities (Marston et al., 2011). The IaaS model allows an organisation to reach Best-of-Breed technology and services which are usually not available to organisations, either for financial or technical reasons (Mather et al., 2009).

One concern with using IaaS is the possibility of losing governance of the infrastructure, which may lead to security issues (Lin & Chen, 2012). Although security aspects are considered as a challenge for most cloud service models, security is seen as a more challenging issue in an IaaS environment because the organisations – rather than the service provider – are responsible for all aspects of their application security (Mather et al., 2009). Another issue with the use of IaaS is related to the type of user because an IaaS approach requires users with specific system administration skills (Hudaib et al., 2014). Thus, the IaaS model is more suitable for larger organisations with existing IT capabilities (Hsu et al., 2014). The IaaS model serves more consumer types (i.e. end users, IT professionals with or without specialised system administration skills) than the PaaS model but less than the SaaS model (Chuang et al., 2015), and IaaS has been recognised by many organisations as providing a cloud service model with more promising benefits, although requiring more technical skills. A study by IDG Enterprise (2016) found that the studied organisations allocated 30% of their cloud budget to the IaaS model.

Table 2.1 presents a comparison between the three Cloud Computing Service Models. Cloud Computing support delivers other capabilities as services which are referred to as X Service (XaaS); examples include Data as a Service, Security as a Service, Business Processes as a Service, etc. (Dai et al., 2012; Loebbecke et al., 2011; Armbrust et al., 2009). The three models (SaaS, PaaS, IaaS) are the fundamental service models based on the NIST definition (Mell & Grance, 2011).

Table 2.1 Comparison between Cloud Computing Service Models

Features SaaS PaaS IaaS

Control of resources (for all models, there is no control over the underlying cloud infrastructure)

Control is over the data only

Control is over the data and deployed applications

Control is over operating systems,

storage, deployed applications and data

Consumer type End users without

high-level IT skills Software developers without specialised system administration skills IT professionals with specialised system administration skills Industrial Examples Salesforce, NetSuite, Google Apps, Office 365 Microsoft's Azure Services Platform, Salesforce's Force.com, Google AppEngine Amazon's Simple Storage Service (S3), Microsoft's Azure Services Platform