OpenStack Approach
OpenStack can be used as the focal point for infrastructure cloud solutions in hybrid, in-house,
commercial, and hybrid deployments along with a complete OpenSourceeco system that is
filled with frameworks and tools. Several EU projects like the CloudWave are using this ecosys-
tem to design their own cloud infrastructure.
In addition, with diverse partnership due to an independent consortium that includes
small and large industrial companies, OpenStack has the influence to handle a wide variety
of approaches and paradigms. Therefore, it promises to remain relevant amidst the highly
crowded IaaS (infrastructure as a service) space.
While working with the implementation’s detail in the IoT environments, the resource-
level restrictions and standard embedded systems may need ad-hoc fixes.
Going by the above, one of the best choices for communication standard can be CoAP or
constrained application protocol. The bus layer is the straddling of centralized subsystems and
distributed things. In addition to CoAP, there is also scope for other RESTful end-to-end protocols.
A bus helps to provide not only a forwarding model but also a queuing system that does
not generally exist in common end-to-end communication. For forwarding model, flexibility
is attained whereas scalability is improved for the queuing system. Furthermore, transparent
delivery is also a byproduct through the use of publish-subscribe techniques. This view also
factors in that the OpenStack architecture and its “bus” contribute to the solution.
Therefore, to update analysis for the cloud subsystem, you can use OpenStack’s Ceilometer
for billing and metering. For the dashboard, there is the Horizon. These frameworks can help
to solve issues pertaining to data visualization and data acquisition where the “things” provide
the samples.
Ceilometer’s metering is a necessary functionality in billing applications, a common part
of cloud solutions, though here it is reversed because measurements are of prime importance
that can help with several important urban parameters.
Is Open Stack Cloud architecture better than any other IoT architecture? Does it come
out cheaper for the organizations to implement this over the others listed here?
Flash Question
HADOOP
Hadoop is also used with IoT. This section discusses the use of StreamSets Data Collector for the
migration of IoT sensor data to the Hadoop. To do this, there are two patterns.
Pattern 1 (Real-Time IoT Events Streaming)
The Internet of Things produces voluminous datasets of event data. During the ingestion of such
data, there are several challenges that have to be factored in.
• You have to manage dierent firmware versions of devices in the midst of the rolling
upgrades. All the firmware versions can have their own slightly distinct schemas and
data payloads.
Chapter 4 IoT Architecture Choices 89
Internet_of_Things_CH04_pp081-104.indd 89 9/3/2019 10:13:31 AM
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.