234
F. Firouzi and B. Farahani
Infrastructure
Hypervisor
Bins/Lib
Guest OS
App 1
Bins/Lib
Guest OS
App 2
Infrastructure
Operating system
Bins/Lib
App 1
Bins/Lib
App 2
Container engine
Virtual machine
Container
Fig. 4.36 The architecture of virtual machine and container
for deployment while sharing the host OS. Deploying IoT in Cloud requires
a large network of servers, protocols, sensors, and applications with many
endpoints including mobile, web, and firmware that need extensive integration
among applications, data, and devices. These need to expand development
timelines and increase the effort needed, even in an agile environment. Both
microservices and containerization support efficient development by dividing
IoT functions into more manageable, independent modules able to work
in isolation without negatively impacting IoT performance. In an advanced
IoT ecosystem, individual microservices are built with diverse programming
languages such as Java, C, C++, Ruby, or Python. Containers increase service
speed and ease of use by providing isolation within a single host and managing
framework or library dependencies. Containers can also be turned on and off
without adversely affecting subsequent containers located on the same host.
– Function-Level Virtualization: The term “serverless” does not mean that
servers are not required, but that developers do not have to spend much
time thinking about them anymore because the developer’s focus can be
shifted from managing servers to code. Serverless is also known as Functions
as a Service (FaaS). As previously discussed, microservices are used to
break monolithic applications down into smaller services that are more easily
developed, managed, and scaled separately. FaaS takes this idea further by
breaking microservices into even smaller pieces/functions (see Fig. 4.37). In
FaaS and serverless computing, we run the code (function) only on-demand
on a per-request basis (i.e., in response to events based on event-driven
programming model). The current leader in this area is Amazon Lambda,
while the products such as Azure Functions and PubNub Blocks are also
F. Firouzi and B. Farahani
Infrastructure
Hypervisor
Bins/Lib
Guest OS
App 1
Bins/Lib
Guest OS
App 2
Infrastructure
Operating system
Bins/Lib
App 1
Bins/Lib
App 2
Container engine
Virtual machine
Container
Fig. 4.36 The architecture of virtual machine and container
for deployment while sharing the host OS. Deploying IoT in Cloud requires
a large network of servers, protocols, sensors, and applications with many
endpoints including mobile, web, and firmware that need extensive integration
among applications, data, and devices. These need to expand development
timelines and increase the effort needed, even in an agile environment. Both
microservices and containerization support efficient development by dividing
IoT functions into more manageable, independent modules able to work
in isolation without negatively impacting IoT performance. In an advanced
IoT ecosystem, individual microservices are built with diverse programming
languages such as Java, C, C++, Ruby, or Python. Containers increase service
speed and ease of use by providing isolation within a single host and managing
framework or library dependencies. Containers can also be turned on and off
without adversely affecting subsequent containers located on the same host.
– Function-Level Virtualization: The term “serverless” does not mean that
servers are not required, but that developers do not have to spend much
time thinking about them anymore because the developer’s focus can be
shifted from managing servers to code. Serverless is also known as Functions
as a Service (FaaS). As previously discussed, microservices are used to
break monolithic applications down into smaller services that are more easily
developed, managed, and scaled separately. FaaS takes this idea further by
breaking microservices into even smaller pieces/functions (see Fig. 4.37). In
FaaS and serverless computing, we run the code (function) only on-demand
on a per-request basis (i.e., in response to events based on event-driven
programming model). The current leader in this area is Amazon Lambda,
while the products such as Azure Functions and PubNub Blocks are also