4 Architecting IoT Cloud
177
4.2.1 Cloud Computing Key Characteristics
According to NIST, each Cloud should have the following basic characteristics:
• On-Demand Self-Service – Users are able to independently utilize computing
capability including network storage and server time as needed without human
intervention from individual Cloud service providers.
• Broad Network Access – Computing is accessible via a network and typical
mechanisms that are useable by heterogeneous thick or thin client platforms such
as laptops, mobile phones, workstations, or tablets.
• Resource Pooling – Computing resources such as processing, storage, network bandwidth, and memory are combined to serve many consumers via a
multi-tenant model. Various virtual and physical components are allocated and
reassigned based on client demand. Customers can sometimes at a higher level
of abstraction specify the location (i.e., state, country, or datacenter), but they
do not control or have exact knowledge of where provided resources are coming
from.
• Rapid Elasticity – It is possible to provide and release (sometimes automatically)
capabilities and quickly scale out on-demand. It may appear to clients that
provided capabilities are limitless and always obtainable in any quantity.
• Measured Service – Cloud computing systems control and maximize the use of
resources by utilizing metering abilities pertinent to a specific service type (i.e.,
active user accounts, processing, bandwidth, or storage). The consumption of
resources can be tracked, controlled, and documented for reporting purposes,
creating transparency for consumers and service providers.
4.2.2 Service Models
Providers of Cloud computing services utilize different models to offer services.
According to the NIST, the three typical models include Platform as a Service
(PaaS), Software as a Service (SaaS), and Infrastructure as a Service (IaaS).
Figure 4.2 illustrates the deployment options for IoT applications as well as the
three service models of Cloud. Note that IoT users can deploy their applications in
a private system (called on-premises) or on the Cloud as shown in Fig. 4.2.
• Infrastructure as a Service (IaaS) – This model gives organizations and companies computing resources such as networking, servers, data center space, and
storage, using a pay-per-use structure.
• Platform as a Service (PaaS) – This model creates an environment that includes
all things needed to facilitate complete life cycles for constructing and distributing Cloud applications without the complications or cost purchasing and
managing all fundamental software, hardware, or hosting.
177
4.2.1 Cloud Computing Key Characteristics
According to NIST, each Cloud should have the following basic characteristics:
• On-Demand Self-Service – Users are able to independently utilize computing
capability including network storage and server time as needed without human
intervention from individual Cloud service providers.
• Broad Network Access – Computing is accessible via a network and typical
mechanisms that are useable by heterogeneous thick or thin client platforms such
as laptops, mobile phones, workstations, or tablets.
• Resource Pooling – Computing resources such as processing, storage, network bandwidth, and memory are combined to serve many consumers via a
multi-tenant model. Various virtual and physical components are allocated and
reassigned based on client demand. Customers can sometimes at a higher level
of abstraction specify the location (i.e., state, country, or datacenter), but they
do not control or have exact knowledge of where provided resources are coming
from.
• Rapid Elasticity – It is possible to provide and release (sometimes automatically)
capabilities and quickly scale out on-demand. It may appear to clients that
provided capabilities are limitless and always obtainable in any quantity.
• Measured Service – Cloud computing systems control and maximize the use of
resources by utilizing metering abilities pertinent to a specific service type (i.e.,
active user accounts, processing, bandwidth, or storage). The consumption of
resources can be tracked, controlled, and documented for reporting purposes,
creating transparency for consumers and service providers.
4.2.2 Service Models
Providers of Cloud computing services utilize different models to offer services.
According to the NIST, the three typical models include Platform as a Service
(PaaS), Software as a Service (SaaS), and Infrastructure as a Service (IaaS).
Figure 4.2 illustrates the deployment options for IoT applications as well as the
three service models of Cloud. Note that IoT users can deploy their applications in
a private system (called on-premises) or on the Cloud as shown in Fig. 4.2.
• Infrastructure as a Service (IaaS) – This model gives organizations and companies computing resources such as networking, servers, data center space, and
storage, using a pay-per-use structure.
• Platform as a Service (PaaS) – This model creates an environment that includes
all things needed to facilitate complete life cycles for constructing and distributing Cloud applications without the complications or cost purchasing and
managing all fundamental software, hardware, or hosting.