Overcoming Today's Data Challenges with MongoDB

www.mongodb.com
MongoDB Amsterdam
October 3rd 2017
Overcoming Today's Data
Challenges with MongoDB
With the participation of:

• How the World has Changed Since
Relational Databases were Invented
• How to radically transform your IT
environments with MongoDB
• MongoDB, the Database of choice for
multiple Use Cases
• Customer Story: IHS Markit
• Q&A and Conclusion
Agenda

Your Speakers today:
Matthijs Van Vliet
Regional Director Benelux and
Nordics, MongoDB
Matthijs.vanvliet@mongodb.co
m
Roman Gruhn
Director of Information Strategy
(EMEA), MongoDB
Roman.gruhn@mongodb.com
Sander Van Loo
Executive Director, Indices
and Data Delivery, IHS Markit
Eugene Bogaart
Solution Architect, MongoDB
Eurgene.bogaart@mongodb.co
m

Matthijs van Vliet
Regional Director MongoDB, Benelux & Nordics
matthijs.vanvliet@mongodb.com
How the World has Changed
Since Relational Databases were
Invented

Digital Platforms Have Changed
The platforms your end users and customers use to engage with your applications and services have
fundamentally changed at an unprecedented speed over the past 5 years.
UPFRONT SUBSCRIB
E
Busines
s
YEARS / MONTHS WEEKS / DAYS
Applications
P
C
MOBILE / BYOD
Customers
ADS SOCIAL
Engagement
SERVER
S
CLOUD
Infrastructure

TRADITIONAL MODERNIZED
APPS On-Premise, Monoliths SaaS, Microservices
DATABASE Relational (Oracle) Non-Relational (MongoDB)
EDW Teradata, Oracle, etc. Hadoop
COMPUTE Scale-Up Server Containers / Commodity Server / Cloud
STORAGE SAN Local Storage & Data Lakes
NETWORK Routers and Switches Software-Defined Networks
The New Enterprise Stack

800+
employees
About
MongoDB, Inc.
4,000+
customers
29 offices
worldwide
$311M in
funding

Customers MongoDB
Office
Support MongoDB User
Groups
25+ Million Downloads

MongoDB Use Cases
Single View Internet of
Things
Mobile Real-Time
Analytics
Catalog Personalization Content Management

Let our team help you on your journey to efficiently leverage the capabilities of MongoDB, the database that
allows innovators to unleash the power of software and data for giant ideas.
Being successful with MongoDB
We have worked with over 50% of the Fortune 500 companies. While the definition of success metrics
look different for each one of them, 2 key factors are consistent across all of our engagements:
5xProductivity
We help our customers to increase
overall output, e.g. in terms of
development or ops productivity.
80%Cost reduction
We help our customers to dramatically lower
their total cost of ownership for data storage
and analytics by up to 80%.

How to radically transform your IT
environments with MongoDB
Roman Gruhn
Director, Information Strategy, MongoDB
roman.gruhn@mongodb.com

Agenda
• Something has changed…
• Challenges & Opportunities
• The New Operating Models in IT
• Customer Success Stories

The Dominance of Data
“Software is
eating the world”
“Software is king,
but data is queen”
Our Mission:
Be the data platform for innovators everywhere

The World Has Changed
Leverage Data &
Technology to Maximise
Competitive Advantage
Accelerate
Time to Value
Dramatically
Lower TCO
Reduce Risk for
Mission-Critical
Deployments
Data Applications Commercials Risk
Our Value Drivers:
Volume
Velocity
Variety
Time to value
Architectures
Operating Models
Scalability
Opex vs Capex
TCO
24/7 availability
Global impact
Business criticality

Software is disrupting every industry
Source: US Bureau of Economic Analysis
Manufacturing Retail Transportation Publishing,
Broadcast
Education,
Healthcare,
Social
Assistance
Finance,
Insurance,
Real Estate
Arts,
Entertainment,
Food
$1.6T
$1.1T
$1.5T
$6.2T
$5.3T
$2.4T
$1.2T

Key Topics
Regulatory &
Compliance
Customer
Expectations
Startups
(Social, Media, FinTech /
etc.)
Cloud
TCO Talent

Most banking
executives believe
challenges are
enabling growth
Source: http://www.visualistan.com/2016/02/top-tech-trends-for-banks-in-2016.html
Regulatory Challenges pose growth
opportunities

Key decision criteria
What to think about when choosing a cloud data platform
Deployment Flexibility
On-premise, Private, Public, or
Hybrid without vendor lock-in
Reducing Complexity
Broad use case applicability to
avoid additional complexity
Agility
Accelerate time to market and
speed of change for the business
Resiliency
Engineered for high availability
across distributed architectures
Scalability
Elastically grow with demand
Cost
Aligned to actual demand and
value but with predictability
Security
Leverage best in class and
appropriate security controls

Legacy
Legacy systems are falling short
RDBMS systems were not created for today’s requirements and consequently try to bolt-on features to compensate for
the lack of capabilities. But this strategy can’t compete with data systems purpose-built to solve today’s problems.
Rigid Schemas
Resistant to
change
Throughput &
Cost make Scale-
Up Impractical
Relational Model Scale-up
Data changes constantly, which
fits poorly with a relational model
Scale-Up clusters were never meant to handle
today’s volumes

MongoDB combines the best of Relational &
NoSQL
Scalability
& Performance
Always On,
Global Deployments
Flexibility
Strong Consistency
Enterprise Management
& Integrations
NoSQL
Expressive Query Language
& Secondary Indexes
Relational

MongoDB – Multi-purpose operational data
platformMongoDB is the most powerful, holistic data management platform in the market today, helping you to reduce system
complexity, drastically lower TCO, increase productivity and minimise risk for critical operations.
Multi-Model database – rich use cases require
“more”
than just relational queries (document, graph, search,
etc)
Multi-Workload support – combine operational and
analytical workloads in a single, powerful data platform
Multi-structured, polymorphic data – real-life
data
doesn’t fit into rows/columns and changes over
time
Maximum deployment optionality – from on-premise
and
VMs to hybrid/public cloud and Database-as-a-Service
K-V SQLDOC
Cloud / DBaaSOn-premise / self-managed
{
first_name: ‘Paul’,
surname: ‘Miller’,
city: ‘London’,
location:
[45.123,47.232],
cars: [
{ model: ‘Bentley’,
year: 1973,
value: 100000, … },
{ model: ‘Rolls Royce’,
year: 1965,
value: 330000, … }
]
}
+

Strategic
SaaS, Mobile, Social
Microservices /
API Access / JSON
Polymorph Data (structured,
semi-structured, unstructured)
Hadoop, Spark
Commodity HW / Cloud
Local Storage / Cloud
Software-Defined Networks
MongoDB and Enterprise IT Strategy
Our technology can help you transform your IT organization and modernize the entire IT stack by enabling you leverage
strategic solutions on every level to drive business transformation.
Legacy
Apps On-Premise
Data Access
Object-Relational Mapping /
ODBC Access / SOAP
Database Oracle / Microsoft
Data Schemas Relational Data / Structured
Offline Data Teradata
Compute Scale-Up Server
Storage SAN
Network Routers and Switches
MongoDB sits right at the centre of
strategic IT and business / digital
transformation, enabling full stack
modernization.
By removing layers we can:
• Reduce complexity
• Reduce cost
• Increase business agility
• Improve data & service quality
• Facilitate innovation

Technology Landscape Transitions
Platform 1: Mainframes Platform 2: Client/Server Platform 3: Cloud/Mobile
1960s-1980s 1990s-2000s 2010-Beyond

Architecture is shifting
On premises / self-hosted
Monolithic
Proprietary
Fat Client / Web v1
Cloud
Microservices
Open source
Mobile

...as is the Org Structure
Centralized IT
Hierarchical
Specialized
Process heavy
DevOps
Small, autonomous teams
Cross functional
Agile
(a la Amazon, Google, Netflix)

Multi Cloud
On Premises
Desktop
Cloud
Self-
Managed
Fully Managed
MongoDB - Develop and run anywhere

Automated Available On-Demand
Secure Highly Available Automated Backups
Elastically Scalable
MongoDB Atlas - Database as a Service

Problem Why MongoDB ResultsProblem Solution Results
Massive variability in data structured
ingested from customer systems:
highly concurrent batch loads and
continuous queries
Relational databases didn’t provide
schema flexibility or scalability
Hadoop was too complex
MongoDB Enterprise Advanced running on
Azure
Complex queries and aggregations to support
ad-hoc, exploratory queries
MongoDB Connector for BI to provide rich
visualizations in Tableau
MongoDB Cloud Manager for operational
automation and disaster recovery
First to market with unique
management accounting services
50% faster development time than
any other database
5x scale on same infrastructure
footprint
Cloud-Based Data Lake
Industry-first “benchmarking” service for 70,000 French
businesses, built on MongoDB & Azure
France

Problem Why MongoDB Results
Problem Solution Results
• With the advent of mobile banking,
Barclays has experienced a significant
growth of traffic originating from mobile
devices to Mainframe platforms that
supports banking applications. Growth
of traffic, which is expected to continue,
has led to an increased cost of
operations and decreased performance
• Ability to provide high resiliency during
mainframe outages
• Existing ETL processes that load
transaction data into Teradata on a daily
basis are updated to additionally feed
data to MongoDB paving the way for
decommissioning of Teradata. In
subsequent phases of the project,
MongoDB will be updated in near real-
time via a live transactions feed
• De-normalized real time data store using
MongoDB with the benefits to reduce
growth
• Stand-in capability to support Resiliency
during planned and unplanned outages
across mainframe system and other
source systems
• Reduced cost of operations
• Reduced number of read only
transactions to Mainframes , there by
freeing up mainframe resources for
additional growth
Operational Data Source
Data lake to store data from multiple sources for operations on the data.
ODS is built to store and process read only customer transactions for
business operations, analysis and reporting.

Bala Chandrasekaran, Director Data Optimisation & Simplification
“This is because MongoDB architecture is scalable and the mainframe isn’t under as much
pressure. The operational database now has over 13 billions transactions held in 114 million
documents”.
Customer Testimonial: Barclays

100+ Apps and Growing
Faster time-to-market and lower operational overhead
makes MongoDB the new default at Expedia
Developers impeded by rigid relational
model leading to inability to effectively
keep up with business
Significant effort achieving performance
targets and maintaining optimal user
experience
Significant operational overhead
involved in maintaining status quo, and
time-to-deploy new systems
Flexible data model makes it easy to
adapt to unforeseen and frequent
changes, allowing for radical data
model changes with no downtime
Inherent high-performance removed
need for significant ongoing
performance tuning
Native HA and multi-DC support
streamlined production deployments
100+ new apps launched over a 2
years
Enabled Ops to provision new
Production systems in under an hour: >
72X productivity improvement over SQL
Server
1:100+ Ops Engineer to Production
Server ratio compared to 1:28 with SQL
Server

Overcoming Today's Data Challenges with MongoDB

MongoDB, the DB of choice for multiple Use
Cases
Eugene Bogaart
Senior Solution Architect, MongoDB
Eugene@mongodb.com

Cases for Change
MongoDB is a modern, operational database that supports a polyglot data strategy – on-premise and in the cloud. This
allows us to drive several business critical topics with our customers.
Cloud Data Strategy
Leveraging the right data platforms as part of your
overall cloud strategy helps to avoid vendor lock-in.
Legacy Modernisation
Current legacy technology stacks can’t cope with the
range of new business requirements – we can help you
modernise in a highly efficient and effective way.
Mainframe Offloading
Reduce cost and MIPS on legacy mainframes and
enable data to be leveraged for new use cases.
Operational Intelligence
Solving the problem of deriving value from existing
EDW or Hadoop-based data lake solutions in real-time.
Single View
Provide a holistic view of data entities (e.g. customer) across
multiple underlying, disconnected source systems.
Compliance & Regulation (e.g. PSD2)
MongoDB is enabling scalable, highly available data
platforms to banks who are forced to provide data in a more
agile way to comply with the PSD2 regulations.
Internet of Things (IoT)
MongoDB can help you overcome Scalability & Performance
issues that are not being met by many current IoT solutions

Single View Defined
• What
– Single, real-time representation of a business entity
or domain
– Customer, product, supply chain, financial asset
class & more
• Why
– Improves business visibility
– Serve operational applications
– Foundation for analytics
• How
– Gathers and organizes data from multiple,
disconnected sources
– Aggregates information into a standardized format
and joint information model

Single View Use Cases
• Comparative view of
contracts or products
• Agency-wide view of
asset exposure
• Aggregated
transactions for fraud,
waste or abuse
models
• Access logistics domain data
• Agency-wide view of the
processes, resources, and
efforts of providers
• Enables strategic decision
support, deep-dive analytics,
and long-term trend analysis
• Management of patient
medical records for
treatment plans
• Macro-analysis view for
public health
• Medical history to
identify insurance risk
Finance Logistics Healthcare

Why Single View?
• Efficiently retrieve status of any
business entity in real time,
• Foundation for analytics, i.e:
• cross-sell,
• upsell,
• churn risk

Single View with a Relational Database

Solution: Aggregate with a Dynamic Schema
…Mobile
App
Web
Call
Centre CRM Social
Feed
COMMON FIELDS
CustomerID | Activity ID | Type…
DYNAMIC FIELDS
Can vary from record to record
Single View

ETLorMessageQueue
Web
Mobile
CRM
Mainframe
Single View
Call Center
Analytics
Technical
Support
Billing
Source Systems Consuming Systems
Load Reads
High Level Architecture

Architecture for Writes to the Single
View
ETLorMessageQueue
Web
Mobile
CRM
Mainfra
me
Single View
Call
Center
Analytics
Technica
l Support
Billing
Update
Queue
Reads
Writes
Source Systems Consuming Systems
Load

Why MongoDB for Single View
• Data model flexibility with a dynamic
schema
• Real-time analytics
• Rich query, aggregation, search & reporting
• Performance, scale & always-on
• Enterprise deployment model

Single View of the Customer
360° view of the customer increases customer satisfaction,
cross-sell & up-sell with MongoDB, Spark, & Hadoop
Customer data spread across 100+
systems, making it difficult for Air France
to personalize the customer experience
Commercial service agents not able to
retrieve all data about a customer from a
single system
New single view CRM planned to provide
a better experience for agents but legacy
relational systems made it different to
create a common data model
Single View application was built on MongoDB to
take advantage of the database’s flexible data
model, expressive query language, secondary
indexes, & horizontal scalability
Data from old relational systems fed into Spark
for analysis and then stored in MongoDB to
support real-time CRM
The data stored in MongoDB feeds nightly batch
jobs in Hadoop, the results of which go back into
MongoDB to better inform personalized
recommendations
Air France expects increased revenues
from more personalized offerings,
which will drive cross-sell and upsell
Reduced competitive pressures from
addressing a key gap in product
offerings
Current plan is to store 100 TB of
customer data in MongoDB

Single View of Customer
Insurance leader generates coveted single view of
customers in 90 days – “The Wall”
No single view of customer, leading to
poor customer experience and churn
145 years of policy data, 70+ systems,
24 different 1-800 numbers, 15+ front-
end apps that are not integrated
Spent 2 years, $25M trying build single
view with DB2 – failed
Built “The Wall,” pulling in disparate
data and serving single view to
customer service reps in real time
Flexible data model to aggregate
disparate data into single data store
Expressive query language and
secondary indexes to serve any field in
real time
Prototyped in 2 weeks
Deployed to production in 90 days
Decreased churn and improved ability
to upsell/cross-sell

Where to Go from Here?
• Single view projects are challenging
– Partner with a vendor offering proven methodology,
tools & technologies
• Learn More
– Download the whitepaper
– 10-Step Methodology to Building a Single View
• Engage
– MongoDB Global Consulting Services can help you
scope the project and get started
– Book a workshop

Operational Data Layer defined
• What
– Real-time delivery of value from data in existing
EDWs or Hadoop-based data lake solutions
• Why
– Reduce dependency of legacy systems
(Mainframe, RDBMS)
– Minimize High profile outages & downtime
– Speed up time to market
– Reduce Complexity & risk (of change)
– Regulatory Requirement to open backend systems
to public APIs
• How
– Gathers and organizes data and interoperate with
various application types
– Capture multi-structured data and grow into the
petabyte scale

Problem / Solution Overview
RDBMS Files
Mainframe
Application
Microservices / API Layer
ReadsWrites
Key/Value
Store
Files
Mainframe
Application
Typical Architecture
Complex & Fragile
Operational Data Layer (ODL)
Simplified & Resilient
Application Application Application
In-Memory
Cache
RDBMS
Wide-Column
Store
Application Application
Non-standard data access Standardised Data Access
Near Real-
Time CDC
Message
Streaming/Pr
ocessing
Graph Store

Characteristics: Operational Data Layer (ODL)
• Supports Structured, Semi-
Structured and Un-Structured
data with the same level of
functionality
• Native drivers connect
applications to data without need
for conversion (JSON)
• Multi-tenancy through use of a
• Native support for All deployment
types
• On-premise/Bare Metal, Private, Public,
Hybrid and Cross Clouds
• Scale-out architecture supports all
deployment types in mixed mode
• Information Lifecycle Management
easily managed by workload and
geography
Data Agnostic Deployment
Agnostic
&

Why MongoDB for Operational Data Layer
Data:
• Dynamic Data model flexibility with a dynamic
schema
• Workload isolation
• Expressive Queries & Secondary Indexes
Deployment:
• Real-time analytics
• Performance, scale & always-on
• Enterprise deployment model

Real-Time Analytics
Travelers instantly browse billions of recommendation with
MongoDB powered new Amadeus flight search portfolio
Amadeus serves 124 airlines across 190 countries and
needs to process over 1.6 billion data requests each day
25% of travelers have not decided on a destination and
almost ½ don’t know the date they want to travel
Need to provide a more personalized travel experience
that is instant
Expectations for online and mobile services are
incredibly high; must develop an Instant Search
application to browse billions of travel options across
multiple criteria in real time
Unable to provide instant results to multi-dimensional
queries and perform at scale
MongoDB’s flexible data model to accelerate
time to value and handle key data structures at
immense scale, for the industry's most
demanding travel companies
Multi-region distribution for scalability and high
availability
MongoDB Enterprise security features allowed
the ability to easily authenticate and authorize
users
WiredTiger storage engine for compression
and efficiency
MongoDB able to deliver complex searches across
multiple dimensions, returned in seconds. Both the
internal NoSQL DB, and relational DBs couldn't
handle this complexity at scale
MongoDB clusters powering additional apps that
are handling 10s of TBs of data, and 10s of TBs of
throughput
MongoDB WiredTiger storage to compress storage
by 80% significant cost reductions and
performance improvements

API Access Layer
Operational Data
Customers
Products
Accounts
ML Models
Shared Physical Infrastructure
App1 App2 App3
1. Development agility
– UI for self-service provisioning & scaling
2. Data Re-use
– Each service’s data is physically isolated into
its own database instance
– Federated across services with appropriate
permissioning
3. Corporate Governance
– Logically managed as one service
Cloud Data Strategy: Standardized, On-Demand
DB Service
Cloud Agnostic
Any Cloud, Any Where

Eliminating Lock-In
Freedom of choice
Traditional
Data Centres
Cloud IaaS
Cloud PaaS
Ops Mgr
Cloud Mgr Cloud Mgr
Atlas
Ops Mgr
Pure
On-Prem
Pure
IaaS
Hybrid On-
Prem / DBMaaS
Hybrid IaaS
/ DBMaaS
Pure
DBaaS

Why MongoDB for Cloud Data Strategy?
• Freedom of choice
• On premise and/or as Managed Service
• Same code base everywhere

Why MongoDB Atlas?
• Ready for Developers and DevOps
• Scalable back-end for your application on-demand
• Secure by Default
• High Available, even while scaling
• Path maintenance performed for you
• Your own MongoDB cluster in the cloud
(multitenant)

IoT App Running on MongoDB Atlas
Biotechnology giant uses MongoDB Atlas to allow their customers
to track experiments from any mobile device
Thermo Fisher is developing Thermo Fisher
Cloud, one of the largest cloud platforms for the
scientific community on AWS
For scientific IoT applications, internal
developers need a database that could easily
handle a wide variety of fast-changing data
Each experiment produces millions of “rows” of
data, which led to suboptimal performance with
incumbent database
Thermo Fisher customers need to be able to
slice and dice their data in many different ways
MS instrument Connect allows Thermo
Fisher customers to see live experiment
results from any mobile device or browser
MongoDB’s expressive query language
and rich secondary indexes provide
flexibility to support both ad-hoc and
predefined queries to support customers’
scientific experiments
Deployed MongoDB using MongoDB
Atlas, a hosted DB service running on
Amazon EC2
Thermo Fisher customers now can obtain
real-time insights from mass spectrometry
experiments from any mobile device or
browser; not possible before
Improved developer productivity with 40x
less code in testing with MongoDB when
compared to incumbent databases
Improved performance by 6x
Easy migration process & zero downtime.
Testing to production in under 2 months

PSD2 Banking
Global consulting company turns to MongoDB Atlas to
launch a critical application to meet the EU Payment
Services Directive
Needed a solution to help the customer
comply with the EU Payment Services
Directive
Lack of knowledge of security and
operational best practices while
adhering to tough deadlines
A tight budget required the project team
to be nimble while architecting a
solution that meets any future scaling
and performance needs
Built application on top of MongoDB Atlas
with out-of-the-box security controls, end-
to-end encryption, high availability and
continuous backups
Using MongoDB’s flexible data model to
ingest variety of unstructured banking
data and iterate on app quickly
Provided self-service capability for
developers to rapidly develop and deploy
application
Successfully achieved all client project
milestones and met all technical
requirements of the EU Payment Services
Directive
Eliminated the need to rely on operational
and security experts while considerately
reducing TCO for the client
Delivered a distributed future-proof cloud
native application that can scale to
accommodate any data and performance
needs without any downtime

© 2017 IHS Markit. All Rights Reserved.© 2017 IHS Markit. All Rights Reserved.
Financial Markets Data Delivery
Enhancing Data Agility
October, 2017

© 2017 IHS Markit. All Rights Reserved.
Mission
• Distribute data produced by Financial Markets products to customers
> Provide consistent service across multiple asset classes, products and delivery channels
> Minimize delivery latency for products
> Support low cost addition of new data sets and fields
> Provide access control and usage data
> Provide transparency on delivery and performance
78

The inception of a new delivery platform
• Redesign of the platform started at the end of 2013, go-live in 2015
• Design goals
> Standardization and simplification
> Robustness and scalability
> Increase transparency
> Data agility
79

Architecture
80
Web Delivery (through MCS)
File Delivery
API Delivery
Data Acquisition
Customer
Data
Producer
Data Store
Data AccessMongoDB
Messaging System
Kafka
SFTP Server
Price Viewer
Application
Data Dictionary
Data DictionaryData
Dictionary
File
Generator
Feed
conforms
conforms
Screener
Price Viewer
Services
Run-time data onboarding Product Specific Interactions

Data Onboarding Process
81
Sample File Analyze
Create
Dictionary
Configure
Dictionary
Publish Data
Configure API
Configure
File Delivery
Determine:
- Entities
- Fields
- Data types
- Identifiers
Describe using
template:
- Entities
- Fields
- Data types
- Semantics
- Constraints
- Cardinality
Convert template to
DDML
Upload dictionary
Configure meta data
(e.g. batches)
Configure entitlements
Check for new data
Map to dictionary
Publish to feed
Whitelist namespace
Whitelist namespace
Run API
Client
Get sample file
Use API portal for
onboarding
Run File
Delivery
Configure file

Data Agility
Use data dictionaries to support run-time onboarding
82

Data dictionaries
• A data dictionary is required to integrate with Data Delivery
• A dictionary describes the product data using a structure that the platform can understand:
> Entities
> Fields
> Types
> Constraints
> Semantics
• Once a dictionary is created and configured all platform components learn about the new data
from the dictionary definition
83

Example data dictionary
84

Dictionary derived API documentation
85

Files UI dictionary derived field selection
86

Data Storage
Migration from a relational database to MongoDB to support
high performance time series requests
87

Data storage in a relational database
> Schema Generator
– Database schema needs to be able to change at run-time
– Schema generator converts DDML to DDL statements
– Coordination is required to prevent locking and make schema changes appear atomic
– Data migration rules need to be coded into the schema generator
> Performance
– Messages are deconstructed to be stored in normalized schema
– On retrieval original messages have to be reconstructed
– BLOB/CLOB data types are difficult to search and have their own performance challenges
– Time series on structured data
88

New use cases are driving technology development
• API: Time series
> For event based feeds customers need lossless delivery of streaming data.
> Customers require the ability to pull historical data for batch feeds.
• Files: Historical Files and Restatements
> Files are currently constructed from a point in time snapshot.
> The snapshot time is always the time of the most recent batch.
> For performance reasons the snapshot is pre-computed.
> Historical files require a snapshot for an arbitrary point in time, within reasonable
parameters (e.g. < 3 months).
89

MongoDB versus RDBMS
• Write Performance (across 4 shards)
• Read Performance
• > 70% reduction in storage utilization for highly structured data compared to existing RDBMS implementation
90
Product Batch RDBMS (mm:ss) MongoDB (mm:ss) Improvement
Corporate and Sovereign Bonds Pricing N1600 00:18.7 00:04.0 4.7x
Municipal Bonds Pricing N1600 01:29.9 00:17.0 5.3x
Structured Products Pricing N1600 01:54.0 00:12.0 9.5x
Data Range RDBMS (ms) MongoDB (ms) Improvement
1 week of EOD data 758 7 113.7x
1 year of EOD data 2,523 10 261.0x

Alignment with design goals
• Simplification and standardization
> MongoDB natively stores processes and stores JSON documents
• Robustness
> Due to schema less design, no schema generator is required
> MongoDB provides redundancy through ‘replicas’
– Replicas can be used to provide load balancing or locality for reads
– Support hot-hot operations across multiple data centers in multiple regions
• Transparency
> Data is not transformed or modified on storage
> Resource allocation and capacity management
91

Alignment with design goals (2)
• Scalability
> MongoDB scales reads and writes by adding ‘shards’
– Each shard holds part of the dataset and can execute queries on that part of the dataset in
parallel to other shards
– Each shard holds up to 2TB of data and can handle up to 6 Gbps of I/O
• Data Agility
> Support dictionary onboarding process through Compass
> Faster time to market as no schema generator is required
> Support for true run-time dictionary changes
92

Example: Dictionary data in an RDBMS
• Example CDS curve
> Retrieval of 1 curve with 8 curve points requires reading 9 rows from two tables and joining them.
• Difficult to scale when running on millions of rows and/or with large numbers of curve points
93
RDBMS requires structured data to be stored across multiple tables

Dictionary data in MongoDB
• Example CDS curve
> 1 curve is 1 document
94
MongoDB stores dictionary data as a single document

Dictionary data in MongoDB
95
Visualization of data distribution using MongoDB Compass

MongoDB Topology
96
WriteConcern = Majority
US Data Center
(Eligible Primary)
UK Data Center
(Eligible Primary)
EU Data Center
(Eligible Primary)
Shard 1
Replica 1
Shard 2
Replica 1
Shard 3
Replica 1
Shard 1
Replica 2
Shard 2
Replica 2
Shard 3
Replica 2
Shard 1
Replica 3
Shard 2
Replica 3
Shard 3
Replica 3
Shard 4
Replica 1
Shard 4
Replica 2
Shard 4
Replica 3
Shard 5
Replica 1
Shard 5
Replica 2
Shard 5
Replica 3
Shard 6
Replica 1
Shard 6
Replica 2
Shard 6
Replica 3
Shard 7
Replica 1
Shard 7
Replica 2
Shard 7
Replica 3
Shard 8
Replica 1
Shard 8
Replica 2
Shard 8
Replica 3
Shard 9
Replica 1
Shard 9
Replica 2
Shard 9
Replica 3
Shard 10
Replica 1
Shard 10
Replica 2
Shard 10
Replica 3

Summary
97

Service Improvements
• Faster delivery
• Improved business continuity
• Simple to scale
• Shorter time to market
98

Questions?
99

Disclaimer
The information contained in this presentation is confidential. Any unauthorised use, disclosure, reproduction or dissemination, in full or in part, in any media or by any
means, without the prior written permission of Markit Group Limited or any of its affiliates ("Markit") is strictly prohibited.
Opinions, statements, estimates and projections in this presentation (including other media) are solely those of the individual author(s) at the time of writing and do not
necessarily reflect the opinions of Markit. Neither Markit nor the author(s) has any obligation to update this presentation in the event that any content, opinion, statement,
estimate or projection (collectively, "information") changes or subsequently becomes inaccurate.
Markit makes no warranty, expressed or implied, as to the accuracy, completeness or timeliness of any information in this presentation, and shall not in any way be liable
to any recipient for any inaccuracies or omissions. Without limiting the foregoing, Markit shall have no liability whatsoever to any recipient, whether in contract, in tort
(including negligence), under warranty, under statute or otherwise, in respect of any loss or damage suffered by any recipient as a result of or in connection with any
information provided, or any course of action determined, by it or any third party, whether or not based on any information provided.
The inclusion of a link to an external website by Markit should not be understood to be an endorsement of that website or the site's owners (or their products/services).
Markit is not responsible for either the content or output of external websites.
Copyright ©2016, Markit Group Limited. All rights reserved and all intellectual property rights are retained by Markit.
100

Conclusion
1 MongoDB is reshaping the
DB Management Landscape
2 Time to Market, Developer
Productivity and TCO are
driving this Change
3 Engage with your local
MongoDB Team

Resources to Get Started
Spin up a cluster on the
Free Tier today
Download the Whitepaper

Overcoming Today's Data Challenges with MongoDB

More Related Content

What's hot (20)

Similar to Overcoming Today's Data Challenges with MongoDB (20)

More from MongoDB (20)

Overcoming Today's Data Challenges with MongoDB