Unit 5

IoT Physical Servers, Cloud

Offerings & IoT Case
Studies
 Conten
t• Introduction to Cloud Storage Models
• Communication API
• WAMP: AutoBahn for IoT
• Xively Cloud for IoT
• Python Web Application Framework: Djanjo
• Amazon Web Services for IoT
• SkyNet IoT Messaging Platform.
• Case Studies:
• Home Intrusion Detection,
• Weather Monitoring System,
• Air Pollution Monitoring,
• Smart Irrigation.
Introduction to Cloud Storage
Models
• Popular Models are
• Amazon Web Service (AWS)
• Xively Cloud (PAAS)
Communication
API
• Cloud Models are relied on Communication API
• Communication API facilitate data transfer, control information
transfer from application to cloud, one service to another
• It also exist in the form of Communication Protocols
• It supports RPC, PUBSUB and WAMP
• Eg. Popular API is RESTful API (communication in cloud model)
• Django web framework is used to implement Communication
API
WAMP: AutoBahn for
IoT
• WAMP : Web Application Messaging Protocol
• Mainly used in cloud storage model for IoT & other messaging services
• WAMP is a routed protocol, with all components connecting to a WAMP Router,
where the WAMP Router performs message routing between the component
• It is protocol for Web Socket (PUBSUB based protocol) : uses RPC Messaging Pattern
• Some Important Key Terminologies
• Transport
• Session
• Clients (Publisher & Subscriber)
• Router
• Broker
• Dealer
• Application Code
 WAMP for IoT

• Web Application Messaging Protocol (WAMP) is a sub-protocol of WebSocket

which provides publish–subscribe and remote procedure call (RPC) messaging
patterns.

Book website: http://www.internet-of-things-book.com Bahga & Madisetti, © 2015

 WAMP – Concepts
• Transport: Transport is a channel that connects two peers.
• Session: Session is a conversation between two peers that runs over a transport.
• Client: Clients are peers that can have one or more roles.
• In the publish–subscribe model, the Client can have the following roles:
– Publisher: Publisher publishes events (including payload) to the topic maintained by the Broker.
– Subscriber: Subscriber subscribes to the topics and receives the events including the payload.
• In the RPC model, the Client can have the following roles:
– Caller: Caller issues calls to the remote procedures along with call arguments.
– Callee: Callee executes the procedures to which the calls are issued by the Caller and returns the results to the Caller.
• Router: Routers are peers that perform generic call and event routing.
• In the publish–subscribe model, the Router has the role of a Broker.
– Broker: Broker acts as a Router and routes messages published to a topic to all the subscribers subscribed to the topic.
• In the RPC model, the Router has the role of a Dealer.
– Dealer: Dealer acts a router and routes RPC calls from the Caller to the Callee and routes results from the Callee to the
Caller.
• Application code: Application code runs on the Clients (Publisher, Subscriber, Callee or Caller).

Book website: http://www.internet-of-things-book.com Bahga & Madisetti, © 2015

 Xively Cloud Services™
Commercial Platform as a Service for the Internet of Things

• Supports hundreds of platforms, millions of

gateways and billions of smart devices
• Comprehensive and secure infrastructure
services
• Online development tools and dev center
• Best of breed approach

Built on LogMeIn’s Gravity platform connecting

255+ m devices for 55+ m users
Xively …
Contd…
• Xively is an IoT Cloud Platform
• It is an enterprise platform for building, managing, and deriving
business value from connected products.
• It also provides the cloud base API with an SDK
• It supports platforms and technologies like Android, Arduino, C
etc.
• Xively is a PaaS (Platform as a Service) which exposes its service via
RESTful API
• It supports messaging service based on MQTT
Case Study: “Smart Plant System“
[3]
Detail
s
• This IoT project uses Arduino Uno and a set of sensors:
• DHT11: Temperature and humidity sensor
• TEMT6000: Light intensity sensor
• YL-38 + YL-69: Soil moisture sensor
• We want to send all values read from these sensors to Xively and
create a dashboard so we can monitor these parameters.
• Code : Download
Results by
Xively
 Amazon EC2 (Elastic Compute Cloud)
Python Example
• Boto is a Python package that provides interfaces to Amazon Web Services (AWS)
#Python program for launching an EC2 instance
• In this example, a connection to EC2 service is import boto.ec2
first established by calling the from time import sleep
boto.ec2.connect_to_region function. ACCESS_KEY="<enter access key>"
SECRET_KEY="<enter secret key>"

• The EC2 region, AWS access key and AWS secret REGION="us-east-1"
key are passed to this function. After connecting AMI_ID = "ami-d0f89fb9"
EC2_KEY_HANDLE = "<enter key handle>"
to EC2 , a new instance is launched using the INSTANCE_TYPE="t1.micro"
conn.run_instances function. SECGROUP_HANDLE="default"

conn =
• The AMI-ID, instance type, EC2 key handle and boto.ec2.connect_to_region(REGION,
security group are passed to this function. aws_access_key_id=ACCESS_KEY,
aws_
secre
t_ac
cess
_key
=SEC
RET_
Book website: http://www.internet-of-things-book.com KEY) Bahga & Madisetti, © 2015
 Amazon AutoScaling – Python Example
#Python program for creating an AutoScaling group (code excerpt)
• AutoScaling Service import boto.ec2.autoscale
:
• A connection to the AutoScaling service is first established print "Connecting to Autoscaling Service"
by calling the boto.ec2.autoscale.connect_to_region conn = boto.ec2.autoscale.connect_to_region(REGION,
function. aws_access_key_id=ACCESS_KEY,
aws_secret_access_key=SECRET_KEY)

• Launch Configuration print "Creating launch configuration"

• After connecting to the AutoScaling service, a new launch
configuration is created by calling lc = LaunchConfiguration(name='My-Launch-Config-2',
conn.create_launch_con f iguration. Launch configuration image_id=A
contains instructions on how to launch new instances MI_ID,
including the AMI-ID, instance type, security groups, etc.
key_name=EC2_KEY_HANDLE,

• AutoScaling Group instance_type=INSTANCE_TYPE,

sec
• After creating a launch configuration, it is then associated ur
with a new AutoScaling group. The AutoScaling group is ity
created by calling conn.create_auto_scaling_group. There _g
are settings for the AutoScaling group such as the ro
maximum and minimum number of instances in the u
group, the launch configuration, availability zones, ps
optional load balancer to use with the group, etc. =
launch_config=lc, min_size=1, max_size=2,
connection=conn) [
SECGROUP_HANDLE, ])
conn.create_auto_scaling_group(ag
Book website: http://www.internet-of-things-book.com conn.create_launch_configuration(lc)
)
 Amazon AutoScaling – Python Example

• AutoScaling Policies #Creating AutoScaling policies

• After creating an AutoScaling group, the policies for scale_up_policy = ScalingPolicy(name='scale_up',

scaling up and scaling down are defined. adjustment_type='ChangeInCapacity',
• In this example, a scale up policy with adjustment type
as_name='My-Group',
ChangeInCapacity and scaling_ad justment = 1 is
defined. scaling_adjustment=1,
• Similarly, a scale down policy with adjustment type cooldown=180)
ChangeInCapacity and scaling_ad justment = −1 is
defined. scale_down_policy = ScalingPolicy(name='scale_down',

adjustment_type='ChangeInCapacity',
as_name='My-Group',

scaling_adjustment=-1,

cooldown=180)

conn.create_scaling_policy(scale_up_policy)
conn.create_scaling_policy(scale_down_policy)

Book website: http://www.internet-of-things-book.com Bahga & Madisetti, © 2015

 Amazon AutoScaling – Python Example
#Connecting to CloudWatch
cloudwatch = boto.ec2.cloudwatch.connect_to_region(REGION,
• CloudWatch Alarms aws_access_key_id=ACCESS_KEY,
• With the scaling policies defined, the next step is to aws_secret_access_key=SECRET_KEY
)
create Amazon CloudWatch alarms that trigger these alarm_dimensions = {"AutoScalingGroupName": 'My-Group'}
policies.
#Creating scale-up alarm
• The scale up alarm is defined using the CPUUtilization scale_up_alarm =
metric with the Average statistic and threshold greater MetricAlarm( name='scale_up_on_cpu',
70% for a period of 60 sec. The scale up policy created namespace='AWS/EC2', metric='CPUUtilization',
statistic='Average', comparison='>',
previously is associated with this alarm. This alarm is threshold='70',
triggered when the average CPU utilization of the period='60', evaluation_periods=2,
alarm_actions=[scale_up_policy.policy_arn],
instances in the group becomes greater than 70% for dimensions=alarm_dimensions)
more than 60 seconds. cloudwatch.create_alarm(scale_up_alarm)
• The scale down alarm is defined in a similar manner #Creating scale-down alarm
with a threshold less than 50%. scale_down_alarm =
MetricAlarm( name='scale_down_on_cpu',
namespace='AWS/EC2', metric='CPUUtilization',
statistic='Average', comparison='<', threshold='40',
period='60', evaluation_periods=2,
alarm_actions=[scale_down_policy.policy_arn],
dimensions=alarm_dimensions)
cloudwatch.create_alarm(scale_down_alarm)
Book website: http://www.internet-of-things-book.com Bahga & Madisetti, © 2015
 Amazon S3 (Simple Storage Service)
Python Example
• In this example, a connection to the S3 service is first established by calling the boto.connect_s3
function.
• The upload_to_s3_bucket_path function uploads the file to the S3 bucket specified at the specified
path.
# Python program for uploading a file to an S3 bucket
import boto.s3

conn = boto.connect_s3(aws_access_key_id='<enter>',
aws_secret_access_key='<enter>')

def percent_cb(complete, total):

print ('.')

def upload_to_s3_bucket_path(bucketname, path, filename):

mybucket = conn.get_bucket(bucketname)
fullkeyname=os.path.join(path,filename)
key = mybucket.new_key(fullkeyname)
key.set_contents_from_filename(filename, cb=percent_cb,
num_cb=10)

Book website: http://www.internet-of-things-book.com Bahga & Madisetti, © 2015

 Amazon RDS (Relational Database Service)
Python Example
#Python program for launching an RDS instance (excerpt)
import boto.rds
• In this example, a connection to the RDS service is
first established by calling the ACCESS_KEY="<enter>"
SECRET_KEY="<enter>"
boto.rds.connect_to_region function. REGION="us-east-1"
INSTANCE_TYPE="db.t1.micro"
• The RDS region, AWS access key and AWS secret key ID = "MySQL-db-instance-3"
are passed to this function. USERNAME = 'root'
PASSWORD = 'password'
• After connecting to the RDS service, the DB_PORT = 3306
DB_SIZE = 5
conn.create_dbinstance function is called to launch a DB_ENGINE =
new RDS instance. 'MySQL5.1'
DB_NAME = 'mytestdb'
SECGROUP_HANDLE="default
• The input parameters to this function include the "
instance ID, database size, instance type, database
#Connecting to RDS
username, database password, database port, conn = boto.rds.connect_to_region(REGION,
database engine (e.g., MySQL5.1), database name, aws_access_key_id=ACCESS_KEY,
aws_secret_access_key=SECRET_KEY)
security groups, etc.
#Creating an RDS instance
db = conn.create_dbinstance(ID, DB_SIZE, INSTANCE_TYPE,
USERNAME, PASSWORD, port=DB_PORT, engine=DB_ENGINE,
Book website: http://www.internet-of-things-book.com db_name=DB_NAME, security_groups = Bahga
[ SECGROUP_HANDLE,
& Madisetti, © ]2015
 Amazon DynamoDB –Non Relational Databases
Python Example
# Python program for creating a DynamoDB table (excerpt)
import boto.dynamodb
• In this example, a connection to the DynamoDB
service is first established by calling ACCESS_KEY="<enter>"
boto.dynamodb.connect_to_region. SECRET_KEY="<enter>"
REGION="us-east-1"
• After connecting to the DynamoDB service, a schema
for the new table is created by calling #Connecting to
DynamoDB
conn.create_schema. conn = boto.dynamodb.connect_to_region(REGION,
aws_access_key_id=ACCESS_KEY,
• The schema includes the hash key and range key aws_secret_access_key=SECRET_KEY)
names and types. table_schema = conn.create_schema(
hash_key_name='msgid',
• A DynamoDB table is then created by calling the hash_key_proto_value=str,
conn.create_table function with the table schema, range_key_name='date',
range_key_proto_value=str
read units and write units as input parameters. )

#Creating table with schema

table =
conn.create_table( na
me='my-test-table',
schema=table_schema,
read_units=1,
Book website: http://www.internet-of-things-book.com write_units=1 Bahga & Madisetti, © 2015
 Python Packages of Interest

• JSON
• JavaScript Object Notation (JSON) is an easy to read and write data-interchange format. JSON is used as an alternative to XML
and is easy for machines to parse and generate. JSON is built on two structures: a collection of name–value pairs (e.g., a Python
dictionary) and ordered lists of values (e.g., a Python list).
• XML
• XML (Extensible Markup Language) is a data format for structured document interchange. The Python minidom library provides
a minimal implementation of the Document Object Model interface and has an API similar to that in other languages.
• HTTPLib & URLLib
• HTTPLib2 and URLLib2 are Python libraries used in network/internet programming.
• SMTPLib
• Simple Mail Transfer Protocol (SMTP) is a protocol which handles sending email and routing email between mail servers. The
Python SMTPLib module provides an SMTP client session object that can be used to send email.
• NumPy
• NumPy is a package for scientific computing in Python. NumPy provides support for large multi-dimensional arrays and
matrices.
• Scikit-learn
• Scikit-learn is an open source machine learning library for Python that provides implementations of various machine learning
algorithms for classification, clustering, regression and dimension reduction problems.

Book website: http://www.internet-of-things-book.com Bahga & Madisetti, © 2015

 Python Web Application Framework – Django

• Django is an open source web application framework for developing web applications in Python.
• A web application framework in general is a collection of solutions, packages and best practices
that allows development of web applications and dynamic websites.
• Django is based on the Model–Template–View architecture and provides separation of the data
model from the business rules and the user interface.
• Django provides a unified API to a database backend.
• Thus, web applications built with Django can work with different databases without requiring
any
code changes.
• With this flexibility in web application design combined with the powerful capabilities of the Python
language and the Python ecosystem, Django is best suited for cloud applications.
• Django consists of an object-relational mapper, a web templating system and a regular-expression-
based URL dispatcher.

Book website: http://www.internet-of-things-book.com Bahga & Madisetti, © 2015

 Django Architecture

• Django uses a Model–Template–View (MTV) framework.

• Model
• The model acts as a definition of some stored data and handles the interactions with the database. In a web application,
the data can be stored in a relational database, non-relational database, an XML file, etc. A Django model is a Python
class that outlines the variables and methods for a particular type of data.
• Template
• In a typical Django web application, the template is simply an HTML page with a few extra placeholders. Django’s
template language can be used to create various forms of text files (XML, email, CSS, Javascript, CSV, etc.).
• View
• The view ties the model to the template. The view is where you write the code that actually generates the web pages.
View determines what data is to be displayed, retrieves the data from the database and passes the data to the
template.

Book website: http://www.internet-of-things-book.com Bahga & Madisetti, © 2015

Django Project
Layout
django-admin.py startproject
<PROJECT_ROOT>
manage.py
<PROJECT_DIR>

init

.py
settings.py
settings.p
y
• Defines settings used by a Django application
• Referenced by wsgi.py to bootstrap the project loading
• Techniques for managing dev vs prod settings:
• Create settings-dev.py and settings-prod.py and use symlink to link settings.py
to the correct settings
• Factor out common settings into base-settings.py and import. Use
conditionals to load correct settings based on DEBUG or other setting
 Sample
Settings…
DEBUG = True
TEMPLATE_DEBUG = True
ALLOWED_HOSTS = []
# Application definition
INSTALLED_APPS = (
'django.contrib.admi
n',
'django.contrib.auth',
'django.contrib.contenttypes',
'django.contrib.sessions',
'django.contrib.messages',
'django.contrib.staticfiles',
)
Django
Apps
• Reusable modules
• django-admin.py startapp <app_name>
• Creates stub layout:
<APP_ROOT>
admin.py
models.py
templates
(directory
)
tests.py
views.py
urls.py
Django
Models
• Defined in models.py
•Typically inherit from django.db.models.Model
Example Model:
from django.db import models
class TestModel(models.Model):
name = models.CharField(max_length = 20)
age = models.IntegerField()
Models
(cont’d)
• Default is to set NOT NULL on all fields. Override by adding null
= True to field definition:
name = models.CharField(max_length=20, null = True)
• Relationships defined through special field types:
models.OneToOneField(model)
models.ForeignKey(model)
models.ManyToManyField(model)
Models
(cont’)
• Need Nulls in a Boolean Field? Use models.NullBooleanField()
• Set Default value with “default”:
count = models.IntegerField(default = 0)
• Use a inner Meta class to define additional options, especially useful
for abstract classes:
class TestModel(models.Model):
class Meta:
abstract = True
Model
Methods
• model.save(self, *args, **kwargs)
• model.delete(self, *args, **kwargs)
• model.get_absolute_url(self)
• model. str (self) [Python 3] model.
unicode (self) [Python 2]
• Override with super(MODEL, self).save(*args,
**kwargs)
Activating a
Model
• Add the app to INSTALLED_APPS in settings.py
• Run manage.py validate
• Run manage.py syncdb
• Migrations
• Write custom script or manually handle migrations
• Use South
Selecting
Objects
• Models include a default manager called objects
• Manager methods allow selecting all or some instances
Question.objects.all()
Question.objects.get(pk = 1)
Use try block, throws DoesNotExist exception if no match
Question.objects.filter(created_date lt = ‘2014-01-01’)
• Returns QuerySet
Introspecting Legacy
Models
• manage.py inspectdb
• Cut and paste generated code into models.py – Easy!!
 Full
Sample
from django.db import models from datetime import
datetime
class TimestampedModel(models.Model): created_datetime =
models.DateTimeField() updated_datetime = models.DateTimeField()
def save(self, *args, **kwargs): if self.id is None:
self.created_datetime = datetime.now() updated_datetime = datetime.now()
super(TimestampedModel,self).save(*args, **kwargs)
class Meta: abstract = True
Full Sample
(cont’d)
class Question(TimestampedModel):
question_text = models.CharField(max_length = 200)
def str (self):
return self.question_text
Function vs. Class
Views
• Django allows two styles of views – functions or class based views
• Functions – take a request object as the first parameter and must
return a response object
• Class based views – allow CRUD operations with minimal code.
Can inherit from multiple generic view classes (i.e. Mixins)
 Sample – Viewing a List of
Questions
• Function based:
from .models import Question
from django.shortcuts import render_to_response

def question_list(request): questions =Question.objects.all()

return render_to_response(‘question_list.html’,
{‘questions’:questions})
Quick CRUD Operations with Generic
Views
• ListView
• UpdateView
• CreateView
• If Model is specified, automagically creates a matching ModelForm
• Form will save the Model if data passes validation
• Override form_valid() method to provide custom logic (i.e sending
email or setting additional fields)
Sample – As Class Based
View
from .models import Question
from django.views.generic import ListView

class QuestionList(ListView):
model = Question
context_object_name =
‘questions’
Django
Templates
• Very simple syntax:
variables = {{variable_name}}
template tags = {%tag%}
• Flexible – can be used to render
html, text, csv, email, you name
it!
• Dot notation – template engine attempts to resolve by looking for
matching attributes, hashes and methods
Question List
Template
<!doctype html>
<html lang=en>
<head>
<meta charset=utf-8>
<title>List of Questions</title>
</head>
<body>
{%if questions%}
<ul>
{%for q in questions%}
<li>{{q.question_text}}</li>
{%endfor%}
</ul>
{%else%}
<p>No questions have been defined</p>
{%endif%}
</body>
</html>
urls.p
y
• Defines routes to send urls to various views
• Can use regular expressions
• Extract parameters from a url and pass to the view as a named
parameter:
r(‘^question/(?P<question_id>\d+)/$’,’views.question_detail’)
• Extensible – urls.py can include additional url files from apps:
r(‘^question/’,include(question.urls))
 Hooking up the Question
List
from django.conf.urls import patterns, url, include urlpatterns = patterns(‘’,
(r’^questions/$’,’views.QuestionList’))
OR:
from django.conf.urls import patterns from views import
QuestionListView
urlpatterns = patterns(‘’,(r’^questions/$’,’views.QuestionList.as_view()))
Forms in
Django
• django.forms provides a class to build HTML forms and validation.
• Example:
from django import forms
class EditQuestionForm(forms.Form):
question_text = forms.CharField(max_length = 200)

• Often redundant when creating forms that work on a single

model
 ModelForm
•sAutomatically generate a form from a model.
• Handles saving a bound model
• Can specify fields to be included or excluded in the form
• Sample:
from django.forms import ModelForm
from .models import Question
class QuestionForm(ModelForm):
class Meta:
model = Question
fields = [‘question_text’]
Using a
ModelForm
• Create an instance of an empty form:
form = QuestionForm()
• Create a form to update an existing instance of a model:
question = Question.objects.get(pk = 1)
form = QuestionForm(instance = question)
• Pass the form into the template and use the form methods to render
the form:
form.as_p
form.as_ul
form.<field_name>
form.<field_name>.e
rrors
 Request &
Response
• Request object encapsulate the request and provide access to a number of
attributes and methods for accessing cookies, sessions, the logged in user object,
meta data (i.e environment variables),
• Response objects are returned to the browser. Can set content type,
content length, response does not have to return HTML or a rendered
template
• Special response types allow for common functionality:
HttpResponeRedirect
Http404
HttpStreamingResponse
Django
Extras
• CRSF Middleware – enabled by default. Include template tag in
all forms:
{%csrf_token%}
• Authentication
• Caching
• Sessions
• Messages
• Email
• Logging
Authenticatio
n
• Django’s out of the box Auth system uses database authentication.
• Changed extensively in Django 1.6 to allow custom User objects.
• AUTHENTICATION_BACKENDS setting in settings.py allows overriding
how User objects are authenticated
• If using the Authentication middleware and context_processors the
current user is available to code as request.user and {{user}} is
defined in all templates
 Auth
•Decorators
Live in django.contrib.auth.decorators
• login_required
@login_required
def function_view(request):
….
• user_passes_test (can be used with lambda functions for real power)
–
@user_passes_test(lambda u: u.is_staff)
def function_view(request):
…
• has_perms – test for user permissions
Decorating
CBVs
• Decorator is applied to the dispatch method
• Must be converted to a method_decorator – use
django.utils.decorators.method_decorator function:
class MyView(ListView):

@method_decorator(login_required)
def dispatch(self, *args, **kwargs):
super(MyView,self).dispatch(*args,
**kwargs)
Custom Auth Backend for the
Bubble
Sending
Email
• django.core.mail includes functions and classes for handling email
• Set EMAIL_HOST in settings.py to outgoing mailserver
• Import send_mail for simple mail:
send_mail(subject, message, from, to_emails)
• Use django.template.render_to_string to format a message using a
template
• Use EmailMultiAlternatives to create a text message and attach a
html version as well.
Resources
• Python – http://www.python.org
• Django – http://www.djangoproject.com
• Python Packages – https://pypi.python.org
• Django Packages – https://www.djangopackages.com