Enabling Real-Time Analytics for IoT

Rick Negrin, Director of Product Management, MemSQL

March 3, 2017

Enabling Real-Time Analytics for IoTBuilding Real-Time Data Pipelines with Kafka and MemSQL

The Rise of Real-Time Analytics

On-demand economy Internet of Things New technologies

RetailDelivery FinancialEnergy Auto and Transportation And more...

Industries that Need Real Time

Data Producers(simulating

sensor activity)

User Interface

Architecting for Real-Time Analytics

Database

...

Data Transformation

Message Queue

About MemSQL

6

7

Real-Time Architecture and Ecosystem

Real-Time Data Warehouse

Streaming IngestReal-time

data pipelines

Live DataMemory optimized

tables

Historical DataDisk optimized

tables

ApplicationsEmbed custom visualizations and reports for analytic applications

BI DashboardsTableau, Looker, Zoomdata,

MicroStrategy, and Qlik

On-Premises or CloudFreely deploy on- premises, Amazon AWS, Microsoft Azure, Google Cloud Platform and beyond

Containers and VMsDeploy on virtualization and container services like Docker,Mesosphere, VMware, Kubernetes, Cisco, Dell and HP

Historical DataHDFS, Amazon S3, SQL Server,

SAP, Oracle, MySQL

Real-Time DataOperational Appsand web services

Messaging andTransforms

Kafka and Spark

MemSQL is the SQL database capable of ingesting millions of events per secondwhile analyzing petabytes of data for industry-leading time to insight results

8

CommunityFree Forever

EnterpriseFree 30-day Trial

CloudManaged Service

▪ Unlimited capacity and scale

▪ Comprehensive SQL features

▪ 24/7 support▪ Enterprise functionality

▪ High availability▪ Cluster replication▪ Granular access

controls▪ Security features

▪ Fully managed▪ Enterprise functionality▪ Currently in

Pre-Release(GA in 2017)

MemSQL Editions

Data Producers(simulating

sensor activity)

User Interface

Architecting for Real-Time Analytics

MemSQL

...

Data Transformation

Message Queue

MemEx

MemEx: IoT Showcase Application

- Combines MemSQL, Apache Kafka, and Spark for global supply chain management

- Enables enterprises to predict throughput of supply warehouses

- Processes 2 million data points, based on 2,000 sensors across 1,000 warehouses

Live Demomemex.memcompute.commemex-ops.memcompute.com

Data Producers(simulating

sensor activity)

MemEx UI (OpenMaps)

MemEx Architecture

...

Data Transformation

Apache Spark

Spark MLlib Predictive Model

Raw Sensor 1 + Predictive Score 1

S1 P1

1

Q&A

Thank You