IoT is the next big paradigm shift in computing. The move to super-dense sensor networks creates a completely new set of opportunities and challenges for developers, designers and end-users. The databases we designed for the computing environments of the early 90s can no longer support modern, mobile super-scale web applications. In this talk, Joe discussed some of these changes and how they impact the requirements for a modern database.
Internet of Things Cologne 2015: Why Your Dad’s Database won’t Work for IoT and Industrie 4.0
1. MongoDB
IOT City Tour – Cologne
10th September 2015
MongoDB Speaker:
Joe Drumgoole
Director of Solutions Architecture, EMEA
2. Why Your Dad’s database is not
suitable for the Internet of Things
and Industrie 4.0
Joe Drumgoole
Director of Solutions Architecture (EMEA)
@jdrumgoole
15. 15
Closing the Manufacturing Loop
Integrated Product/Service Design,
Product Manufacturing, Service Development
Distribution &
Activation
Marketing & Sales
Service Operations
Remote Condition Monitoring, Remote
Maintenance, Predictive Maintenance
Solution Support & Product
Maintenance
DigitalServices
Resale & Retirement
FieldData/ProductPerformance,
ProductUsageData
Source:
www.enterprise-iot.org
16. 16
Service Led Products
Traditional Manufacturer Servitization
Products Solutions
CAPEX OPEX
Transaction Relationship
Connected Asset
Lifecycle Management
Digital Services
New,
enabled
by IoT
21. 21
Sensor- Density
Gartner – 4.9 billion devices
ABI Research – 16 billion devices
Datamation– 12.9 billion devices
Five Billion Sensors Deployed – Bosch SI
22. 22
• Ubiquitous, cheap sensors and controllers
• Ubiquitous cheap bandwidth
• HTTP/TCP/IP as a universal protocol
• On demand storage at cents per GB
What enables the IoT?
23. 23
What is the next big thing?
The thing we have been doing
badly for the last ten years
Why Now?
25. 25
• Expensive Storage
• Cheap Programmers
• Tables of strings, ints, floats, dates
• One big machine
• A small number of connected users
• A well defined unchanging set of requirements
These are IoT Anti-Patterns
Relational Database Assumptions
26. 26
MongoDB
• Dynamic Schemas
• Automatic Scaling
• Text Search
• Aggregation Framework
and MapReduce
• Hadoop Integration
• GEO Search
• Full, Flexible Index Support
and Rich Queries
• Built-In Replication for High
Availability
• Advanced Security
• Large Media Storage with
GridFS
• Pluggable Storage Engine
27. For startups that want to be enterprises
and enterprises that want to be startups.
Editor's Notes
The ARPANET begins the year with 14 nodes in operation. BBN modifies and streamlines the IMP design so it can be moved to a less cumbersome platform than the DDP-516. BBN also develops a new platform, called a Terminal Interface Processor (TIP) which is capable of supporting input from multiple hosts or terminals.
The Network Working Group completes the Telnet protocol and makes progress on the file transfer protocol (FTP) standard. At the end of the year, the ARPANET contains 19 nodes as planned.
Database lauched in 1970. This internet didn’t even have TCP/IP. Unix hadn’t been invented. IBM resisted Ted Codds innovation allowing Oracle to launch the worlds first SQL database.
This is the internet today. Millons of nodes. All big server farms, giant distributors of data. This is the Internet of people. Nearly all of the data on this internet was created by humans for consumption by humans,
Here is google. Its pretty big.
All the nodes that google indexes are read-only. The are the leaves of an enormous broadcast network. This is the internet as we know it today. Acres of servers doling out web pages to google search requests.
Its still asynchronous. We read more than we write. Ajax and HTML4 have changed the way we interact with. Reflects the way the web was designed. Predominantly broadcast based. More readers than writers. More downloads than uploads. More clients than servers. There has been once big sea change since 1998. Mobile.
Input via, mouse, keybooard, voice, camera. Dumb consumption on both sides. It’s a desktop computer.
9 hidden sensors + camera 10 + voice 11. Intrinsic connection to the Internet. Constantly capturing context.
Real time camera streams, hear sensor, quantified self with personal fitness monitor, RFID tag, Smart watches, smart phones, smart homes.
Mobile networks, speed cameras, TV monitor, iPad , car navigation, credit cards and door keys.
Super dense network. System Steps. High resolution video. What I watch. Where I go what I buy, when I arrive and leave work. Where I drive. What I eat. What I play. What I browse. Inbound data. Feedback loops.
Sensor machines redefining our world. Tesla continues to improve mileage on their cars by aggregating car data. Streetview cars have changed our view of the world. Drones are giving us safe, fast access.
Assess damage in Nepal.
Protocols were analog or proprietary. No feedback loop. Data was often compressed. Usually discarded after lying idle and unused.
We ain’t in Kansas anymore and that database is more often than not a NoSQL database. Why ? Because the stream of data we capture from sensors is unstructured /semi structured and located in time and space.
Just under 1.7bn smart phone users.
953m websites. So you can expect about 1.7 users to login into to your website on average.
Just under 1.7bn smart phone users.
953m websites. So you can expect about 1.7 users to login into to your website on average.
Mobile phones, smart phones, iPhone, Android. Analog cameras, digital cameras. Wais, Gopher archie to WWW. Dial uo to broadband. We reduce friction and people do something more.
Contextual, temporal, all the sensors on that phone related data to time and place. We store everything for ever. The data