Slides: Relational to NoSQL Migration

Relational (RDBMS) to NoSQL
Migration
Ankit Patel | DataStax | Principal Strategy Architect

2 © 2020 Datastax, Inc. All rights reserved.
“We cannot solve our problems
with the same thinking we
used when we created them.”
- Albert Einstein

The Digital Era - The Need to Modernize
Digital Data-Driven AI Enabled

The Modern Era
SAD (Silos Affects Delivery) Speed of Data Matters!
4
Data access
Legacy
processes
Lack of data
analytical skills
Resistance
to change
© 2020 Datastax, Inc. All rights reserved.
Source: https://www.pinterest.com/pin/573716440029920090/

NoSQL - The Future
What is a NoSQL (Not-only-SQL) Database?
• Non Relational Database - supports
ability to access data using other
forms besides Structured Query
Language (SQL)
• Designed to be used by Cloud
Applications’ need to handle massive
amounts of Data in real-time
• Provides ability to overcome scale,
performance, data storage, data
model, and data distribution
limitations

NoSQL vs RDBMS….
C When to use NoSQL? When to use RDBMS?
Applications Decentralized (scalable)
microservice applications
Centralized monolithic
applications
Availability 100% availability,
zero-downtime
Moderate to high
Data Low latency
structured/semi/unstructured
data @ high velocity
Structured data @ moderate
velocity & latency
Transactions Simple transactions & queries Complex nested transactions &
joins
Scalability
(Reads/Writes)
Horizontal (Linear) scaling Vertical scaling

Cassandra: The Best NoSQL Database of Choice
Active-everywhere,
masterless, scales linearly
Best NoSQL database for
cloud-native and microservices
#1 choice of world’s largest
consumer internet applications
Zero Lock-in Global ScaleZero Downtime
If you use a website or a smartphone today,
you’re touching a Cassandra backend system.
Source: https://sdtimes.com/data/apache-cassandra-4-0-beta-now-available/

Cassandra: Cloud Native NoSQL Database
Why?
With Cassandra masterless architecture,
easily achieving 100% uptime across
on-prem, single cloud, hybrid, and/or
multi-cloud deployments is engraved in
the technology.
Experiences, Microservices
& Insights
ON PREM

● CQL – Cassandra Query Language
● Similar to syntax compared to SQL
● Standard way to communicate to DSE C* cluster for
reading/writing data.
● Feature rich language that allow you to manage the cluster
(managing schema/permissions, managing roles, JSON support,
UDF/UDA support…)
● Example Read: select * from keyspace.table where
partition_key=<value>;
● Example Writing Data: insert into keyspace.table
(partition_key,clustering_key,value1) values (‘A’,’B’,’C’);
Cassandra: What is CQL?
9

● Similar to schema in RDBMS
● Container for multiple tables
● Replication Strategy is set at the keyspace level (Example:
SimpleStrategy, NetworkTopologyStrategy)
● Replication Factor deﬁned at the keyspace level
● DURABLE_WRITES is set at the keyspace level. Setting to false
will bypass the commit log.
● Example to create keyspace:
CREATE KEYSPACE test WITH replication = {'class':
NetworkTologyStrategy', 'DC1': '1'} AND durable_writes = true;
Cassandra: What is a Keyspace?
10

● Same as RDMBS table
● Contains a primary key
● Always has partition key as part of primary key
● Optionally can deﬁne a clustering key (ordering can be deﬁned)
● Both partition and clustering key can be composed of multi-column
● A of parameters can be adjusted at the table level (compaction,
compression, gc_grace_seconds, time to live, etc..)
Cassandra: What is a Table?
11

CREATE TABLE test.sample_table (
par_key1 uuid,
par_key2 uuid,
clust_key1 timestamp,
clust_key2 int,
value1 text,
value2 double,
PRIMARY KEY ((par_key1, par_key2), clust_key1, clust_key2)
) WITH CLUSTERING ORDER BY (clust_key1 DESC, clust_key2
ASC)
Cassandra: Example Create Table
12

● Replication factor determines how many copies of your data are
stored in the Cassandra Cluster.
● Each copy is stored in a diﬀerent node.
● Replication Factor can be deﬁned by datacenters that you’ve setup
● This is a parameter set at the keyspace level within the cluster.
Cassandra: What is Replication Factor
13

● This parameter is set by the client on individual queries
● This parameter combined with replication factor can help you achieve
the consistency requirement the speciﬁc use case is looking for.
● Some of the diﬀerent values are
ONE
LOCAL_ONE
QUORUM
EACH_QUORUM
LOCAL_QUORUM
ALL
Cassandra: What is Consistency Level
14

Cassandra - Read/Write in Action
Replication - 3 per DC
Consistency - Per Read/Write
Request from Client
Application - Active/Active
Deployment across DC for
Read/Write
APP
ON-PREM AWS AZURE
APP APP

● Structured Data is the norm for both
● Re-evaluate the need for ACID transactions with
Lightweight-transactions (LWT) in Cassandra
● Take advantage of Cassandra Performance
○ Move Joins to Application Stack
○ Denormalization & Data Duplication is eﬃcient
○ Choose type of Index wisely based on Latency/TPS
requirements
● Thoroughly plan the Data Model in Cassandra
How can My Enterprise get from an RDBMS Based
Design to Cassandra Based Architecture?
16

ERD to Query Based
ERD Based Design Query Based Design
© 2020 Datastax, Inc. All rights reserved.17

5 Steps to Query Based Design
Design a Mental Model of
Access Patterns
Examples:
Medical History: Read
Surgeries, Read Allergies,
Read Health Conditions
Doctor Visit: Read Notes,
Read Prescriptions, Read
Vitals
Decide the application
access patterns to various
entities to deliver business
functionality.
Examples:
Medical History Queries
Doctor Visit Queries
Deﬁne the structure of the
data elements based on
query based design
Example: Read
Prescriptions (patient,
date, drug, dosage, etc..)
Make optimizations to
access the data
Example: Create index to
Read Prescription by drug
type or prescribing Doctor.
Build Cassandra table
schema based on logical
model & optimizations
Example: Table
prescriptions with primary
key patient, date and
index on doctor & drug
type
Application
Conceptual
Model
Logical
Model
Optimizations
Physical
Model

DataStax Enterprise: Cassandra Data Platform
Kubernetes Operator (Cloud-Native Automation + Elasticity)
Developer and DevOps APIs (K8S, CQL, REST, GraphQL, gRPC)
Operational Reliability (Advanced Performance, Enterprise Security, Monitoring)
AI-Scale Experiences, Microservices and Insights
Apache Cassandra NoSQL Database (100% Uptime, Zero-Lock-In, Global Scale)
TRUSTED
ACCELERATED
STRATEGIC
OUTCOMES
FOUNDATIONAL
Operational
Analytics
(Spark, Pipelines,
Streaming)
Enhanced
Search
(Enhance Any Query)
Extensible
Integration
(Kafka, Elastic,
Bulk Loading)
Graph
Engine
(Relate Data Across
Partitions)
Multi-Model
Data
(All Data Styles)
Tools
Thought Leadership
Enterprise Support
Partnerships
OSS Commitment

DataStax Astra: Cassandra Made Easy in the Cloud
Cloud-native
Database-as-a-Service built
on Apache Cassandra
Eliminate the overhead
to install, operate, and
scale Cassandra
Out-of-the-box REST
and GraphQL endpoints
and browser CQL shell
Powered by our
open-source Kubernetes
Operator for Cassandra
Deploy on AWS or GCP and
keep compatibility with
open-source Cassandra
Launch a database in the
cloud with a few clicks,
no credit card required
Cassandra-as-a-Service No Operations Powerful APIs
Cloud Native Zero Lock-in 10 Gig Free Tier

21
Use Case #1 - C&S Wholesale Grocers - Supply
Chain
● Delivers over 140,000 food and non-food items to from over 50
warehouse locations
● Operates over 18 million square feet of storage
● Some of C&S’s customers are Safeway, Target, Stop & Shop
● Traditional solutions slowing down distribution eﬃciency &
impeding innovation
● Business growth leading to Technology Innovation

22
Use Case #1 - C&S - The Challenge
● Supply Chain Process in local RDBMS to warehouse
● Business need to consolidate warehouse data for ease of
management via mobile app
● The transaction volumes were in the thousands per several
seconds
● Needed real-time view of all the working parts of the
manufacturing operations. Warehouse → locations → pallet
● Data Platform capable of operational analytics

23
Use Case #1 - C&S - Why Cassandra?
● Scalable
● High Transaction Volume
● Low Latency
● High Availability - Warehouse operations 24/7
● Ease of Development for Microservices & Mobile App
● Multi-DC Deployment Capability
● Ease of Operational Analytics

24
Use Case #1 - C&S - Business Beneﬁts
● 5 year ROI projection to save multi-millions
● Able to optimize management capabilities of consolidated
warehouse operations
● Achieve remarkable eﬃciency in data pipeline
● Transactions - Read/Write Thousands in seconds
● Supports 300+ Users processing ~ 300k records in 5 mins

25
Use Case #1 - C&S - The Architecture

26
C&S - Case Study
We needed an application that
was entirely reliable and not
vulnerable to unplanned outages
because our warehouses are
pretty much 24/7...
https://www.datastax.com/resources/case-study/cs-
wholesale-achieving-seamless-supply-chain-master
y-datastax-enterprise

27
Use Case #2 - Financial Services - Mobile Banking
● Very competitive retail banking market
● Need to keep up with demand growth in digital banking
● Have high customer satisfaction rates
● Achieve eﬃcient DR & Business Continuity Plans

28
Use Case #2 - Financial Services - The Challenge
● # of Transactions in RDBMS was not easily scalable
● DR was not easy
● Achieving Latency metrics was harder as volumes increased
● Downtime or poor experience would translate to customer churn

29
Use Case #2 - Financial Services - Why
Cassandra?
● Deploy 3 DC Cluster
● Microservices Architecture
● Scale Application Stack w/ Database
● Achieve low latency SLA (<20ms on avg)
● DR Strategy was solid w/ High Availability
● Capable of processing billions of transactions per month

• Customer 360/SVOC
• Omnichannel & Global
Payments
• IoT/Time Series/eCommerce
Data (sensors, tick data, user
interactions, shopping cart)
• Fraud Detection
• Online/Mobile Banking
• Inventory Management
30
Some Other Common Use Cases
• Recommendations (products
& services)
• Regulatory Compliance
• Alerts & Monitoring (Credit
card transactions)
• Global Payments
• Portfolio Management
• Loan Authorization
• Authentication (Mobile
Logins)

Thank You!
Ankit Patel
Principal Strategy Architect @ DataStax
https://www.linkedin.com/in/ankit-p-patel

Slides: Relational to NoSQL Migration

More Related Content

What's hot (20)

Similar to Slides: Relational to NoSQL Migration (20)

More from DATAVERSITY (20)

Recently uploaded (20)

Slides: Relational to NoSQL Migration