Presented by, Dr Christian Geuer-Pollmann, Senior Technology Evangelist at Microsoft.
The presentation gives a solid overview to the Microsoft Azure platform, with a special emphasis on scenarios for IoT workloads. First, Christian provides an introduction to Microsoft Azure’s IaaS compute and networking infrastructure (i.e. virtual machines, virtual networks, load balancers and HA concepts). The second part of the presentation focuses on higher-order services in Azure, such as relational data bases, machine learning, search, and NoSQL offerings. Last, Christian explains how the Azure Service Bus and the Intelligent Systems Services fit into the overall IoT landscape.
16. some selected platform services …
SQL
Azure
RDBMS as a Service
HDInsight Hadoop Cluster as a Service
API
Management
API Proxy for Security, etc.
Azure
Cache
Redis Cache as a Service
Machine
Learning
Machine Learning as a Svc
Traffic
Manager
DNS Loadbalancer
Document
DB
{} Managed NoSQL Doc DB
Azure
Search
Managed Search Service
27. ML API API service service and the Developer
Developer
• Tested models available as an url that can be called from any end point
Portal
Azure Ops Team
Studio
Data Scientist
HDInsight
Azure Storage
Desktop Data
&
ML API service
ML Studio
• Access and prepare data
• Create, test and train models
• Collaborate
• One click to stage for
production via the API service
Azure Portal & ML API service
• Create ML Studio workspace
• Assign storage account(s)
• Monitor ML consumption
• See alerts when model is ready
• Deploy models to web service
29. The IoT ecosystem has been fragmented
Secure?
Connect Configure Harness Administer Extend
Limited flexibility
Inaccessible data
Slow implementation Unsecure data and assets
Incompatible
with infrastructure Unreliable service
30. Connect Configure Harness Administer Extend
Flexible and extensible solution
Accessible data
Intelligent Systems Service
An integrated solution
Finished service provided by
Partner built
Protect
Faster implementation Protect data and assets
Compatible with existing infrastructure
Reliable service
31. Accelerate implementation and time-to-value
Connect across a range of endpoints
OSS Agent (C-library) for arbitrary
systems
Enable broad connectivity options
Connect directly to ISS
or though a local gateway (for
constrained devices)
Connect quickly
Intelligent Systems Service SDK
for Windows
Integrate existing devices and infrastructure
Existing apps through OData
Connect
Agent
Agent
Agent
Agent
Agent
Gateway
32. Optimize performance and reliability
Configure
Deploy out-of-the-box cloud services Automate alarms and response options
preferences
adding or decommissioning devices
changing alarm actions and severity levels
adding new rules
changing connectivity and storage options
Reduce costs with a finished SaaS solution
that does not require development time
and building infrastructure
Use built-in metadata that works with
multiple data schema to drive
intelligent actions and insights such as
command and control
Configure alarms and response options; ISS
provides a number of alarms that can be
configured and customized to support a
number of response options
Drive intelligent actions Adjust as needs change
33. Produce data-driven business insights
Harness
Capture a variety of data Use familiar analytics tools
Apply configurable and customizable
business rules to enable alarming and
eventing based on ingressed data and
through a complex event processing
engine
Capture machine-generated, user-generated and
transaction data using a variety of protocols:
MQTT, AMQP, HTTPS and plug-in protocols
Enable data ingress to Azure Tables, SQL
Azure and on-premises SQL, and access
Intelligent Systems Service BLOB data
seamlessly with your HDInsight account
Use common tools such as Excel and
HDInsight for deep analytics, and enable
data egress through OData interfaces to
other analytics tools both cloud-based and
on-premises
10101110101010101010
10010101010101010110
10101010101110101001
01001101010101010101
10101010101001010101
10101010101001101010
1110101010101
Simplify analysis with rationalized data Apply business rules
34. Achieve new levels of control
Operate from a central dashboard Manage remotely
Support configuration of connected
devices so specific actions are taken on
device groups rather than on an
individual basis, reducing manual
intervention
Use the Intelligent Systems Service Operator
Portal to remotely manage devices, including
monitoring, maintenance, data transfer and
deployment of software
Securely log in from remote devices
and products to retrieve data, control
devices, and diagnose and resolve
issues
Group endpoints for simplified management
Administer
Distribute packages and commands
Leverage the command and control dashboard to
support multiple management activities such as
distributing packages, sending commands and
setting timed transmissions
35. Innovate and grow on a flexible platform
Extend
Integrate existing systems Scale as needed
Work with familiar SI, ISV and OEM partners
that have deep industry expertise to create
rich, customized experiences and vertical
solutions
Connect your on-premises
environment with solution services
running in the Azure public cloud
Incorporate new devices, apps, data and
infrastructure into your existing setup with
the Intelligent Systems Service SDK or the
public SDK
Address variable demands with scalable and
efficient data collection and storage in the Azure
cloud through support of Azure Tables and
Azure BLOB
+
Capitalize on cloud capabilities Innovate with third-party solutions
36. Feel confident your data is protected
Protect
Unify protection system-wide Configure granular permissions
Secure your data with automatic geo-replication
of data across datacenters that
are geographically separate
Simplify security relationships by using secure
protocols like HTTPS and AMQP
Enable data ingress and egress to and
from the cloud via secure protocols with
the Azure Service Bus
Federate granular permissions to ensure
the right people get the right access, and
manage permissions with a consistent
approach across datacenters and the cloud
Transport data through secure channels Access full data recovery features
0101
1100
37. Thank you / chgeuer@microsoft.com
Build with PaaS Use as SaaS
The presentation gives a solid overview to the Microsoft Azure platform, with a special emphasis on scenarios for IoT workloads. First, Christian provides an introduction to Microsoft Azure’s IaaS compute and networking infrastructure (i.e. virtual machines, virtual networks, load balancers and HA concepts). The second part of the presentation focuses on higher-order services in Azure, such as relational data bases, machine learning, search, and NoSQL offerings. Last, Christian explains how the Azure Service Bus and the Intelligent Systems Services fit into the overall IoT landscape.
Slide Objectives:
Explain differences between Push and Pull
Transition:
This is a continuation of Relay vs. Broker discussion
Slide Objectives:
This slide and the next slide list some of integration patterns enabled by queues – load leveling, offline/batch, load balancing (competing consumers)
Slide Objectives:
This slide introduces some integration patterns enabled by topics and subscriptions
<Alternate slide with no animation>
Here’s a simplified snapshot of the whole solution, from storing and managing data, to business users accessing results and making decisions. If you already have a Microsoft Azure subscription or data in the cloud – especially in HDInsight – you are more than halfway there to realizing the benefit of this solution.
Let’s start in the bottom left with the Azure Portal.
The Azure ops team, maybe already accustomed to managing storage accounts or provisioning Azure virtual machines, can get a machine learning environment set up right from the Azure Portal. They can:
Create an ML Studio workspace and dedicated storage account to get their data scientists up and running
Monitor ML consumption to keep track of expenses
See alerts when a model is ready to be published
And deploy models as web services with the ML API Service
Now, moving right, to the ML Studio experience. This where the data scientist will spend her time:
She can execute every step in the data science workflow in one place – ML Studio
She can access and prepare data
Create, test and train models, as well as import her company’s proprietary models securely into her private workspace
Work with R and over 300 of the most popular R packages along with Microsoft’s business class algorithms
Collaborate with colleagues within the office or across the globe as easy as clicking “share my workspace”
Deploy models within minutes rather than weeks or months
And the data scientist has her choice of what data she wants to pull into her models. She can access data already in Azure, query across Big Data in HDInsight, or pull datasets in right from her desktop.
Once the data scientist is ready to publish, that’s when tested models become available to developers via the API service. The business users can access results, from anywhere, on any device. And any model updates simply refresh the model in production with no new development work needed.
Companies may be creating solutions today, but not in a standardized way. The result has been:
Slow implementation. It might take 18th months to get a solution up and running and even more time to see return on your investment.
Incompatibility with infrastructure. It’s hard to deal with the complexity of multiple protocols, form factors and connectivity methods. It’s hard to leverage your existing technology investments for the long term.
Unsecure data and assets. With complex security models, it’s hard to secure your endpoints or your data.
Unreliable service. It’s hard to guarantee the reliability of your business-critical devices. You don’t want to run the risk of deploying immature technology.
Limited flexibility. Solutions cannot offer scalable data storage at a sustainable price or they are built in an entirely custom, unrepeatable manner.
Inaccessible data. And in the end, you are often left unable to use the data you’re generating for analysis and insight.
Or you’ve looked into IoT before, but it’s been beyond your scope, budget or interest. It’s just been too hard.
T: Microsoft Azure Intelligent Systems Service makes it easier to securely connect, manage, and capture and transform data from industry endpoints.
Silos of data storage, formats, authentication, access and experiences
Centralized command a control is difficult (think app for everything)
Devices are heterogeneous in every way (apis, apps, control…)
The ROI dream of what is possible is why we want to manage devices…
<jonathan>
<jonathan>
Microsoft can help create an intelligent system simply by building on existing investments and providing a foundation for you to achieve your IoT vision.
The solution was built with three guiding principles:
Accelerate Time-to-value. Deploy an out-of-the-box solution that is easy to extend and positioned to scale to quickly realize ROI.
Build on a Trusted Platform. Benefit from the credibility, functionality and innovation of Microsoft Assets and future investments.
Increase Flexibility. Gain greater business insights and control with a single solution for heterogeneous environments.
And it enables the core capabilities required by any customer in any industry:
Connect. Connect endpoints regardless of form-factor, operating system or intelligence to other devices, cloud-based services and infrastructure.
Configure. Apply configurable and customizable business rules that define actions on devices to automate and improve business processes.
Harness. Efficiently capture, store, join, visualize, analyze and share data to drive meaningful business insights.
Administer. Remotely manage data transfer, maintenance, configuration, and software deployment on convenient asset dashboards.
Extend. Address variable demands with scalable and efficient data collection and storage in the cloud. Innovate on top of the solution to create rich, customized experiences.
&
Protect. Underlying all of these capabilities is a unified, enterprise-grade approach to security developed and supported by Microsoft.
T: First, let’s take a look at how Microsoft Azure Intelligent Systems Service easily connects all the endpoints in your environment.
Connect quickly. Connect with Intelligent Systems Service agents for Windows devices delivered through the Intelligent Systems Service SDK.
Connect across a range of endpoints. Easily develop open-source agents supporting other operating systems.
Integrate existing devices and infrastructure. Connect existing devices, data and infrastructure such as LoB applications, Active Directory and others through OData interfaces.
Enable broad connectivity options. Provide connections for unintelligent sensors and actuators through a secure gateway with an Intelligent Systems Service agent.
T: Once your devices are connected and integrated, configure your setup to optimize performance and predict needs.
Deploy out-of-the-box cloud services. Reduce IT burden and costs with a finished SaaS solution that does not require development time and resources to figure out the foundational infrastructure for an IoT solution.
Automate alarms and response options. Configure alarms and response options; Intelligent Systems Service provides a number of alarms that can be configured and customized to support a number of response options.
Drive intelligent actions. Use built-in meta data that works with multiple data schema to drive intelligent actions and insights such as command and control.
Adjust as needs change. Easily manipulate and add preferences as needs change, such as adding or decommissioning devices, changing alarm actions and severity levels, adding new rules and changing connectivity and storage options.
T: The next step is to capture meaningful data.
Capture a variety of data. Capture machine-generated, user-generated and transaction data using a variety of protocols: MQTT, AMQP, HTTPS and plug-in protocols.
Use familiar analytics tools. Use common tools such as Excel and HDInsight for deep analytics, and enable data egress through OData interfaces to other analytics tools both cloud-based and on-premises.
Simplify analysis with rationalized data. Enable data ingress to Azure Tables, SQL Azure and on-premises SQL, and access Intelligent Systems Service BLOB data seamlessly with your HDInsight account.
Apply business rules. Apply configurable and customizable business rules to enable alarming and eventing based on ingressed data and through a complex event processing engine.
T: Intelligent Systems Service then offers simplified management to make the most of endpoint connections and data capturing.
Operate from a central dashboard. Use the Intelligent Systems Service Operator Portal to remotely manage devices, including monitoring, maintenance, data transfer and deployment of software.
Manage remotely. Securely log in from remote devices and products to retrieve data, control devices, and diagnose and resolve issues.
Distribute packages and commands. Leverage the command and control dashboard to support multiple management activities such as distributing packages, sending commands and setting timed transmissions.
Group endpoints for simplified management. Support configuration of connected devices so specific action are taken on device groups rather than on an individual basis, reducing manual intervention.
T: As your needs change, you can easily scale your solution. And to meet additional needs in your industry, work with a trusted partner.
Integrate existing systems. Connect your on-premises environment with solution services running in the Azure public cloud.
Scale as needed. Address variable demands with scalable and efficient data collection and storage in the Azure cloud through support of Azure Tables and Azure BLOB.
Capitalize on cloud capabilities. Incorporate new devices, apps, data and infrastructure into your existing setup with the Intelligent Systems Service SDK or the public SDK.
Innovate with third-party solutions. Work with familiar SI, ISV and OEM partners that have deep industry expertise to create rich, customized experiences and vertical solutions. Microsoft and its partners are industry leaders in innovation.
T: And, lastly, you won’t need to worry about security issues.
Unify protection system-wide. Simplify security relationships by using secure protocols like HTTPS and AMQP.
Configure granular permissions. Federate granular permissions to ensure the right people get the right access, and manage permissions with a consistent approach across datacenters and the cloud.
Transport data through secure channels. Enable data ingress and egress to and from the cloud via secure protocols with the Azure Service Bus.
Access full data recovery features. Secure your data with automatic georeplication of data across datacenters that are geographically separate.
Datenschutz in Azure und die Europäische Article 29 Working Party: http://blogs.technet.com/b/microsoft_blog/archive/2014/04/10/privacy-authorities-across-europe-approve-microsoft-s-cloud-commitments.aspx