22 - Justin - T - Build Modern - applications-DB

Build modern
applications with
purpose-built databases
Justin Thomas
AWS Principal Neptune Specialist
© 2021, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Agenda
• New data requirements for modern applications

• Why purpose-built databases?
• Picking the right tool for the job
• Resources
New data requirements
for modern applications
Rapid expansion of data requirements
Explosion of data Microservices change data Accelerated rate of change

and analytics requirements driven by DevOps
Dev Ops
Data grows 10x every 5 years Microservices architecture Transition from IT to

driven by network- decreases need for “one size fits all’ DevOps increases rate of
connected smart devices databases and increases need for change
real-time monitoring and analytics
Building modern applications with purpose-built
databases
Relational Key-value Document In-memory Graph Time-series Ledger Wide Column
Referential High Store documents Query by key Quickly and easily Collect, store, Complete, Scalable, highly
integrity, ACID throughput, and quickly with create and and process immutable, and available, and
transactions, Low latency access querying microsecond navigate data sequenced verifiable history managed Apache
schema- reads and writes, on any attribute latency relationships by time of all changes to Cassandra-
on-write endless scale between data application data compatible service
AWS
Service(s) ElastiCache &
Aurora RDS DynamoDB DocumentDB Neptune Timestream QLDB Keyspaces
MemoryDB Managed Cassandra
Lift and shift, Real-time Content Caching, session Fraud detection, IoT Systems Build low-latency
Common
Use ERP, CRM, bidding, management, management, social applications, of record, applications,
Cases finance shopping cart, personalization, gaming networking, event tracking supply chain, leverage open
social, product mobile leaderboards, recommendation health care, source, migrate
catalog, geospatial engine registrations, Cassandra to the
customer applications financial cloud
preferences
The Walt Disney Company leverages AWS’s proven global
infrastructure to improve performance and reliability for
Disney+, as one of the largest online streaming services
continues to scale globally.
“
AWS has been our preferred cloud provider for
years, and its proven global infrastructure and “
• Preferred cloud provider: The Walt Disney Company relies
on AWS as its preferred public cloud infrastructure provider
to support the explosive growth of Disney+, which quickly
expansive suite of services has contributed reached its five-year goal of registering 100 million
subscribers in 59 countries, 9 months after launch
meaningfully to the incredible success of Disney+.
• Purpose-built: Disney+ uses more than 50 AWS technologies,
—Joe Inzerillo like machine learning, database, storage, content delivery,
Executive Vice President & CTO, Direct-to-Consumer serverless, and analytics, including Amazon Kinesis,
The Walt Disney Company Amazon DynamoDB, and Amazon Timestream, for
quality streaming experiences
• Performance at scale:” Disney+ handled three billon

requests for Disney+ content in the first three days post
launch. It also successfully managed huge peaks in demand
for premium content, like “Hamilton,” “Mulan,” and “The
Mandalorian
Picking the right
tool for the job
Relational data
Patient
* Patient ID
First name
Divide data among tables Last name
Gender
Highly structured DOB *
Visit
Visit ID
Relationships established via * Doctor ID * Patient ID
* Hospital ID
keys enforced by the system Date
Doctor
Data accuracy and consistency * Doctor ID * Treatment ID
First name
Last name
Medical specialty Medical Treatment
* Hospital affiliation * Treatment ID
Procedure
How performed
Hospital Adverse outcome
* Hospital ID
Contraindication
Name
Address
Rating
Amazon Aurora
MySQL and PostgreSQL-compatible relational database built for the cloud
Performance Availability Highly Fully

& scalability & durability secure managed
5x throughput of standard Fault-tolerant, self-healing Network isolation, Managed by Amazon RDS:
MySQL and 3x of standard storage; 6 copies of encryption at On your part, no server provisioning,
PostgreSQL; scale out up data across 3 AZs; rest / in transit software patching, setup,
to 15 read replicas continuous backup to Amazon S3 configuration, or backups
A leader in 3D design, engineering, and entertainment
software, Autodesk makes software for people who make
things
Challenge:
Their ACM (Access Control Management) app connected to a
single RDS MySQL database instance, limiting the available
scaling options
Solution:
They chose Amazon RDS for MySQL first, and migrated to
Aurora for improved scalability, availability, and performance
Result:
• Aurora provides up to five times the throughput of
standard MySQL databases
• Aurora scales up to 64 TB, with no need to scale manually
• Up to 15 low-latency Aurora Replicas improve availability
Key–value data
PUT {
Simple key–value pairs
TableName:"Gamers",
Partitioned by keys Item: {
Gamers
Resilient to failure "GamerTag":"Hammer57",
Primary key Attributes
High-throughput, low- "Level":21,
Gamer tag Level Points High score Plays
"Points":4050,
latency reads and writes Hammer57 21 4050 483610 1722
"Score":483610,
Consistent performance "Plays":1722
FluffyDuffy 5 1123 10863 43
at scale } }
Lol777313 14 3075 380500 1307
Jam22Jam 20 3986 478658 1694
ButterZZ_55 7 1530 12547 66
… … … … …
GET {
TableName:"Gamers",
Key: {
"GamerTag":"Hammer57“,
“ProjectionExpression“:”Points”
} }
Amazon DynamoDB
Fast and flexible key value database service for any scale
Performance Serverless architecture Enterprise Global replication

at scale security
Consistent, single-digit No hardware provisioning, Encrypts all data by Build global applications
millisecond response times at software patching, or upgrades; default and fully integrates with fast access to local data
any scale; build applications scales up or down automatically; with AWS Identity and by easily replicating tables
with virtually unlimited continuously backs up your data Access Management for across multiple
throughput robust security AWS Regions
Dropbox saves millions by building
scalable metadata store on AWS
Challenge:
Dropbox had a capacity crunch in its on-premises metadata
store due to fast data growth in some partitions. It had three
options: double its on-premises storage, delete swaths of
metadata, or find a new solution.
Solution:
Dropbox turned to Amazon DynamoDB and Amazon Simple
Storage Service for its new managed storage system, which
saved the company millions of dollars and has room for
virtually unlimited user metadata.
Results:
• Launched metadata storage system on AWS in 1 year
• Cut cost per user gigabyte by a factor of 5.5
• Ingests data at 4,000–6,000 queries per second
Amazon
© 2021, Amazon
Amazon Web Services, Inc. or its affiliates. All rights reserved.
DynamoDB S3
Wide column: Apache Cassandra
• Open-source, wide-column data store

• Large scale applications that require fast read and write performance
• Use cases:
• User profiles
• Device metadata
• Time-series data
• Transaction logging
• Cassandra Query Language (CQL)
Amazon Keyspaces (for Apache Cassandra)
Scalable, highly available, and managed Apache Cassandra-compatible database
service
Single-digit
Apache Cassandra- No servers to millisecond Highly available and
compatible manage performance at scale secure
Use the same Cassandra No need to provision, Scale tables up and down 99.99% availability SLA
drivers and tools configure, and operate automatically within an AWS Region
large Cassandra
clusters Virtually unlimited throughput Data encrypted at rest;
and storage integrated with IAM
Check-In: A PwC product
Pseudonymized
contact tracing
PwC
Links to
employees only
to determine
potential
exposure
Mobile application Backend tech Dashboard
Users either install a mobile app Associations between signals Authorized users input name of
on their phones or carry a observed by anonymized impacted employee; ACT
hardware device such as a phones enables proximity instantly provides a list of
beacon tracing between employees potentially at-risk colleagues
Why use a document database?
The JSON document model maps naturally to application data
Each document can have a different data structure and

is independent of other documents
Index on any key in a document, and run ad hoc and

aggregation queries across your data set
Amazon DocumentDB
Fast, scalable, and fully managed MongoDB-compatible database service
Fast Scalable Fully managed MongoDB compatible
Millions of requests per second Separation of compute and Managed by AWS: Compatible with MongoDB 3.6
with millisecond latency storage scales both no hardware provisioning; and 4.0; use the same SDKs,
independently; scale out to auto patching, quick setup, tools, and applications with
15 read replicas in minutes secure, and automatic Amazon DocumentDB. Migrate
backups workloads with AWS DMS.
Purpose-built document database

engineered for the cloud
FINRA better manages millions of documents with DocumentDB
FINRA processes 4 million filings and filing

attachments per year. The regulator’s existing
solution to store millions of documents as
XML in a relational database was consuming
too much storage, required custom tooling,
and was difficult to manage.
FINRA chose Amazon DocumentDB (with

MongoDB compatibility) as a managed JSON
document store, making it simpler to query
“
Amazon and index regulatory documents, reduce
One of the reasons that we picked a managed service is we DocumentDB development cycles, and extend usability of
are big in requirements around security, audit, versioning,
data.
records management. All of these are very important to us,
so by going to a managed service like using
DocumentDB…we get all of these things free for us so that
we do not have to worry about building a lot of this by Unlike in a relational DB, the storage and
”
ourselves. compute are decoupled in DocumentDB,
which enables FINRA’s databases to scale
- Ranga Rajagopal, Senior Director, Enterprise Data Platforms, automatically, independently and quickly,
FINRA regardless of the size of their data.
In-memory databases: usage patterns
Caching Real-time Gaming Geospatial Media

analytics store leaderboards streaming store
Session Chat apps Job Machine learning

store pub/sub queue real-time model scoring
The Pokémon Company migrates
to AWS purpose-built databases
Challenge:
The Pokémon Company International wanted to address the
complexity of managing NoSQL and memory-caching systems
for a database of more than 300 million users.
Solution:
The company uses Amazon Aurora as its main user database,
Amazon DynamoDB to reduce bot login attempts, and Amazon
ElastiCache to enable dynamic caching of user logins.
Results:
• Cuts monthly costs by tens of thousands of dollars
• Reduces number of nodes from 300 to 30
• Experiences zero hours of downtime or performance
degradation after migration
Amazon Amazon Amazon

Aurora DynamoDB ElastiCache
What is a Graph?
• Graph databases are purpose-built to store and navigate relationships.
• Nodes represent real-world objects
• Edges store relationships between objects
• Properties and Labels can be added to both Nodes and Edges
Node Node
Edge
A Social Network is a Graph
• Nodes: People
• Edges: Friend Relationships
• Edges can have direction
Graph Query languages- Designed to move
through data
Graphs query languages are optimized to use connections to move through a network
Jack Annie Sam

Friend Friend
Relational queries work by combining sets of data, via joins or recursive CTE
Person Friends Friends of a Friend
Perso
Perso Name Person Person Person
Friend Person Person Person n Friend Person Person Person
nID Name Name Name
ID ID1 Name1 ID2 Name ID ID1 ID2 ID3
P1 Jack X 2 = 1 2 3
E1 P1 Jack P2 Annie E4 P1 Jack P2 Annie P3 Sam

P2 Annie
E2 P2 Annie P3 Sam
P3 Sam
Why use a graph?
Social Knowledge Fraud Life Network & IT

Recommendations operations
networking graphs detection Sciences
With connected datasets

• Where relationships between data matter just as much as the data itself
• When results are dependent on the basis of the strength, weight, or quality of relationships
Amazon Neptune
Fast, reliable graph database built for the cloud
OPEN FAST RELIABLE EASY
Supports Apache Query billions of 6 replicas of your data Build powerful

TinkerPop, relationships with across 3 AZs with full queries easily with
openCypher, W3C millisecond latency backup and restore Gremlin and SPARQL
RDF graph models
Siemens is a global powerhouse focusing on the areas
of electrification, automation, and digitalization
Challenge:
They were faced with isolated data silos from different
departments that resulted in data inaccessibility,
inefficient workflows, and low data quality
Solution:
• Industrial knowledge graphs for capturing
Siemens Domain Knowledge
• Providing knowledge graphs as a service
Ledgers and audit tables have
been around for a while
Banking and finance Manufacturing Ownership

Keeping track of transactions, Recording components Maintaining records of asset
trades, and accounts used in manufacturing ownership or provenance
Amazon Quantum Ledger Database
Fully managed ledger database
Track and verify history of all changes made to your application’s data
Immutable and transparent Cryptographically Highly scalable Easy to use

verifiable
Append-only, immutable All changes are Executes 2 – 3X as many Flexible document model,
journal tracks history of all cryptographically transactions than ledgers in query with familiar SQL-
changes which cannot be chained and verifiable common blockchain like interface
deleted or modified. Get frameworks
full visibility into entire
data lineage
BMW
Digital Vehicle Passport
Challenge
BMW needs to track trusted, verifiable
automotive data in order to get full
transparency on transactions across
multiple entities.
Solution
Started building a BMW Digital Vehicle
Passport App that provides a transparent and
complete history of vehicle data such as
fueling, inspection, oil changes, diagnostics,
repairs, tire changes, and sales across multiple
partners. Amazon QLDB is at the core of this
solution and provides BMW the verified data
with a centralized trust
Time series data
Time-series data is
a sequence of data
points recorded over
a time interval for
measuring events that
change over time
Amazon Timestream
Fast, scalable, fully managed time series database
1,000x faster and 1/10th the Trillions of

cost of relational databases daily events Time-series analytics Serverless
Collect data at the rate of Adaptive query processing Built-in functions for Automated setup,
millions of inserts per engine maintains steady, interpolation, smoothing, configuration, server
second (10M/second) predictable performance and approximation provisioning, software patching
Customer use case
Trimble Inc., is a leading technology

provider of productivity solutions for the
construction, agriculture, geospatial, and
transportation industries.
“Whenever possible, the Trimble Cloud Platform

team tries to leverage AWS’ managed service
offerings. We are excited to now use Amazon
Timestream as a serverless time series database
supporting our IoT monitoring solution.
Timestream is purpose-built for our IoT-
generated time series data, and will allow us to
reduce management overhead, improve
performance, and reduce cost for our existing
monitoring system.” David Kohler, Engineering
Director – Trimble
Conclusion
Our approach
Architect services Offer a portfolio Help you innovate Provide services that
ground-up for the of purpose-built faster through help you migrate existing
cloud and for the services, optimized managed services apps and databases
explosion of data for your workloads to the cloud
Accelerate modernization with continuous learning
Training created by the experts at AWS
Free digital courses, including: Earn an industry-recognized credential:

Architecting Serverless Solutions AWS Certified Developer – Associate
Getting started with DevOps on AWS AWS Certified DevOps – Professional
Hands-on classroom training Create a self-paced learning roadmap

(available virtually) including: AWS Developer Ramp-up guide
Running Containers on Amazon Elastic AWS DevOps Ramp-up guide
Kubernetes Service (Amazon EKS)
Advanced Developing on AWS
Learn more about

Take Developer
training for you and your
and DevOps
team
training today
Thank you!

22 - Justin - T - Build Modern - applications-DB

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

22 - Justin - T - Build Modern - applications-DB

Uploaded by

Copyright:

Available Formats

Build modern

• New data requirements for modern applications

Explosion of data Microservices change data Accelerated rate of change

Data grows 10x every 5 years Microservices architecture Transition from IT to

Relational Key-value Document In-memory Graph Time-series Ledger Wide Column

• Performance at scale:” Disney+ handled three billon

Relationships established via * Doctor ID * Patient ID

MySQL and PostgreSQL-compatible relational database built for the cloud

Performance Availability Highly Fully

Performance Serverless architecture Enterprise Global replication

• Open-source, wide-column data store

• Cassandra Query Language (CQL)

Mobile application Backend tech Dashboard

The JSON document model maps naturally to application data

Each document can have a different data structure and

Index on any key in a document, and run ad hoc and

Fast Scalable Fully managed MongoDB compatible

Purpose-built document database

FINRA processes 4 million filings and filing

FINRA chose Amazon DocumentDB (with

Caching Real-time Gaming Geospatial Media

Session Chat apps Job Machine learning

Amazon Amazon Amazon

• Graph databases are purpose-built to store and navigate relationships.

• Nodes represent real-world objects

• Edges store relationships between objects

• Properties and Labels can be added to both Nodes and Edges

• Edges: Friend Relationships

• Edges can have direction

Jack Annie Sam

E1 P1 Jack P2 Annie E4 P1 Jack P2 Annie P3 Sam

Social Knowledge Fraud Life Network & IT

With connected datasets

OPEN FAST RELIABLE EASY

Supports Apache Query billions of 6 replicas of your data Build powerful

Banking and finance Manufacturing Ownership

Immutable and transparent Cryptographically Highly scalable Easy to use

1,000x faster and 1/10th the Trillions of

Trimble Inc., is a leading technology

“Whenever possible, the Trimble Cloud Platform

Free digital courses, including: Earn an industry-recognized credential:

Hands-on classroom training Create a self-paced learning roadmap

Learn more about

You might also like