Tapping into the Big Data Reservoir (CON7934)

Copyright © 2014 Oracle and/or its affiliates. All rights reserved. |
Oracle Data Integration:CON7934
Tapping into the Big Data Reservoir with All Data
Jeff Pollock
Vice President, Oracle Data Integration
1 Oracle OpenWorld 2014

Safe Harbor Statement
The following is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features or functionality described for Oracle’s products remains at the sole discretion of Oracle.
Oracle OpenWorld 2014 2

Today’s Agenda
3
Oracle Data Integration Solutions
Big Data Reservoir
•Next generation data platform architecture on Hadoop
Oracle Data Integration for Big Data Reservoir
•Take complete advantage of the modern Big Data platform and leave legacy ETL tools behind
Proven Results with Big Data
•Beyond theory, early adopters getting benefits NOW!
1
2
3
4
Oracle OpenWorld 2014

Oracle Data Integration Solutions and Proven Benefits
Improve Agility
•Deploy Projects Faster
•Reliable Real-Time Reduce Risk
•Popular, Proven Tools
•Open, Not Proprietary Reduce Costs
•Better Productivity
•Eliminate ETL Servers
Analytic Data Integration
•Big Data Integration & Governance
•Data Warehouse Integration
•Business Intelligence Applications
Enterprise Data Integration and Governance
•Enterprise Data Quality and Profiling
•Comprehensive, Heterogeneous Data Integration
•Business Glossary and Metadata Management
Business Continuity
•Active-Active for Maximum Availability
•Zero Downtime Migrations
•Data Consolidation / Application Modernization
24 x 7 x 365

Comprehensive Data Integration & Governance Capabilities
Real-Time Data Movement
–Low impact capture, stage in Hadoop
–Continuous data availability
Data Transformation
–Bulk data movement
–Pushdown data processing
Data Federation
–Virtualized Data Services
Data Quality & Verification
–Fix quality at the source
–Verify data consistency
Metadata Management
–Lineage and Impact Analysis
–Business Glossary Semantics
Data GovernanceFoundationOracle Data Integrator(Transformation) Enterprise Data Quality(Profile, Cleanse, Match and De-duplicate)
FastLoadOracle GoldenGate(Movement) Enterprise Metadata Management & Business Glossary(Business Glossary, Data Lineage, Impact Analysis and Data Provenance) Data Service Integrator(Federation) GoldenGateVeridata(Online Data Verification)
ELT Processingon Hadoop or SQL
Continuous Availability

Data GovernanceFoundation
Differentiated Technical Approach
Dynamic Data Movement
–Real-time CDC is by default, not ETL
–Least invasive on sources
–Proven best performance
–Integrated Oracle capture/apply
No ETL Engines
–Take the processing to the data; don’t move the data to the process
–Leverage your data engines for the workloads (Hadoop or SQL)
Most Heterogeneous
–Leverage open source Hadoop, not proprietary distributions
–Hadoop is the Hub, not ETL tools
–Open metadata standardsOracle Data Integrator(Transformation) Enterprise Data Quality(Profile, Cleanse, Match and De-duplicate)
FastLoadOracle GoldenGate(Movement) Enterprise Metadata Management & Business Glossary(Business Glossary, Data Lineage, Impact Analysis and Data Provenance) Data Service Integrator(Federation) GoldenGateVeridata(Online Data Verification)
ELT Processingon Hadoop or SQL
Continuous Availability

Today’s Agenda
7
1
2
3
4
Big Data Reservoir

Why the word “Reservoir?”
8
https://blogs.oracle.com/bigdata/entry/big_data_and_analytic_top

True Hadoop Opportunity: Big Data Reservoir
9
Deep DataStorage
Data Preparation
Data Discovery
Data staged / merged in
Hadoop to provide single place to explore/discover data
External data staging and long running batch jobs run in Hadoop to make the most of the DB
Store more raw detail data for
less cost, while keeping
aggregates in the DB
DW
Support for Exploratory Analytics without time consuming data modeling
Lower cost data staging and data preparation
Lower cost storage for questionable business data
Data Staging & Preparation
New Data Discovery
Detailed, Deep Data

Copyright © 2014 Oracle and/or its affiliates. All rights reserved. | 10
Reports & Dashboards
Query Planning
Data Integration
Data Modelling
Database Mgmt.
Data Visualization
Query Construction
Data Enrichment
Data Preparation
Data Exploration
Data Acquisition
Operational Responsibilities
Data Science & Discovery
Operational Data Flow and Staffing Models
Data Scientists
DBAs, Developers, Data Stewards, Analysts

Logical Architecture –Seamless Data Integration is Crucial
11
Virtualisation & Query Federation
Enterprise Performance Management
Pre-built & Ad-hoc BI Assets
InformationServices
Data Ingestion
Information Interpretation
Access & Performance Layer
Foundation Data Layer
Raw Data Reservoir
Data Science
Data Engines & Poly-structured sources
Content
Docs
Web & Social Media
SMS
Structured
DataSources
•Operational Data
•COTS Data
•Streaming & BAM
Immutable raw data reservoir
Raw data at rest is not interpreted
Immutable modelled data. Business Process Neutral form. Abstracted from business process changes
Past, current and future interpretation of enterprise data. Structured to support agile access & navigation
Discovery Lab Sandboxes
Rapid Development Sandboxes
Project based data stores to support specific discovery objectives
Project based data stored to facilitate rapid content / presentation delivery
Data Sources
Master & ReferenceData Sources
DataIntegration & Governance
DataIntegration & Governance
DI&G
DI&G
DI&G
DI&G

Concrete Business Value with Big Data Reservoir
Lower TCO for the Data Warehouse
LoBFaster Access to Analytic Data
New Types of Analytics for All Data
•Control the costs of the Data Warehouse
•Massive value multipliers for Teradata and Netezzacustomers
•Put an end to the annual upgrade cycle
•Give analytics to the business earlier in the data lifecycle
•Avoid up front modelling overhead for Discovery
•Empower IT to focus on highest value analytics
•Run BI queries faster
•Support Exploratory Analytics directly from Hadoop
•Run Streaming Analytics from OEP, Storm, Flume etc.
•Drive new business solutions (telematics data, machine data, log data, unstructured data) COSTSPEEDVALUE

Top US AutomakerOracle Data Integration for RealtimeData Delivery to Hadoop Reservoir
Petabyte Scale

Today’s Agenda
14
1
2
3
4
Big Data Reservoir

Oracle Data Integration –Powerful Big Data Solutions
15
Commodity Data Reservoir
Leverage Oracle Data Integration with a wide array of databases or data warehouse appliances
Support Hadoop distributions on commodity hardware
Oracle Engineered Systems
Deeply integrated with Oracle Big Data Appliance and Exadata
Take advantage of Infinibandperformance, Oracle Big Data SQL, Columnar Compression, and all integrated Loader technologies
Streaming Big Data
Integrate realtimetransactional databases with streaming analytics
Filter, join and transform data while it is in motion, make business decisions while data is in memory

Heterogeneous Reservoir with Oracle Data Integration
16
Flume
Hive on MR, Tez, Spark
Logs
OLTP DB
SQOOP
OGG
Pig on MR, Tez, SparkODI
SQOOP
Any DWOGG
Spark
OozieOEDQOEMM
Data Validation & Cleansing
Metadata Mgmt
& Lineage
API/File
Hive/HCat, HDFS,HBase
Hive/HCat, HDFS,HBase
NoSQL
Flume

Copyright © 2014 Oracle and/or its affiliates. All rights reserved. | Oracle OpenWorld 2014 17
European Energy Co. Oracle Data Integration for Data Staging and Transformingin Hortonworks
Real-Time to Hive

Load to Oracle OLH/OSCH
Red Stack Reservoir with Oracle Data Integration
18
TransformHiveODI
Hive/HDFS
FederateHive/HDFS to Oracle
Big Data SQL
Oracle DB OLTP
Load from OracleCopyToBDA
Hive/HDFS
Federate Oracle to HiveQuery Provider for Hadoop
OGGOGG
Hive/HDFS

Engineered System for Big Data from Oracle
19
DISK
PCI FLASH
DRAM
Warm Data
Hottest Data
Active Data
•Engineered data platform
•ODI Data Transformation at the speed of DRAM or the scale of Hadoop
•Utilize each data tier for specialized algorithms & compression
•Speedof DRAM
•I/Osof Flash
•Costof Disk
•Scaleof Hadoop
Hadoop DISKS
Deep DataOracle Data IntegratorOracle GoldenGate
Fully exploit Big Data SQL, In-Memory and No-SQL Advancements from Oracle

Top European BankOracle Data Integration MapReduceData Transformations in Big Data Appliance
Massively Parallel

Streaming Reservoir with NoSQL and DIS
21
Transform(Hive, Pig/Oozie, Spark) ODI
FederateHive/HDFSBig Data SQL
OracleNoSQL
Hive/HDFS
OGGOGG
Hive/HDFS
Any DB
Sensors & Events
Hive/HDFSOEP
Load to Oracle OLH/OSCH

US Digital TV ProviderOracle Data Integration with Hadoop & Kafka
100m Tx/Hr

Today’s Agenda
23
1
2
3
4
Big Data Reservoir

Copyright © 2014 Oracle and/or its affiliates. All rights reserved. | Oracle OpenWorld 2014 24
The 90’s Are Calling: Don’t Custom Code Data Integration!!
Hey big data coders!
Yes, all you out there writing your data load programs in Scala, PigLatin, HiveQLor Java MR….
Custom coded data loading is BAD, stay away!
Been there, done that with C++, Pipes and Pro*C
Debugging kills, live data is always bad, downtime is a major bummer, projects can’t scale to large teams…
When you are past Discovery and in to Operations, use enterprise tools for their reliability and reach into existing IT systems.

Oracle Does Big Data Integration Better
25
Dynamic Data Movement
–CDC is by default, not an add-on
–Least invasive on sources
–Proven best performance
–Native Oracle capture/apply
NoETLEngine
–Take the processing to the data; don’t move the data to the process
–Leverage your data engines for the workloads (Hadoop or SQL)
Most Heterogeneous
–Leverage open source Hadoop, not proprietary distributions
–Hadoop is the Hub, not ETL tools
–Open metadata standards
vs.
Batch Data Movement
–Typical ETL vendors all default to batch data movement in their reference architectures
–Some can “talk the talk” but their CDC tech can’t touch Oracle GoldenGatescale/performance
ETL Engine Must Scale Alongside Hadoop
–Carefully watch how ETL engines scale out; parallelism runs via the Engine –more H/W to buy
–Map out the physical deployment architecture, compare to GG&ODI, the benefits will be clear
Proprietary Vendor Lock-in
–One popular ETL vendor puts their engines at the center of the architecture, not Hadoop
–The mainframe of ETL vendors has proprietary features that mainly run in their own distro
–“Fake free” ETL vendors sell proprietary add-ons
vs.
vs.

Oracle Does Big Data Better: Dynamic Data Movement
HDFS (Files)
HBase(NoSQL)
Hive / Hive Streaming (SQL)
Flume & Storm (Streaming)
Kafka (MPP Pub/Sub)
Spark Streaming (Machine Learning)
Capture Database Transactions and Deliver to Big Data in Real-Time
Capture
Trail
Route
Deliver
Pump GoldenGate

Oracle Does Big Data Better: Invented Pushdown Processing
27
ORCL Investments in ELT/Pushdown Tech
ScriptedSQL
StoredProcs
WarehouseBuilder
DataIntegrator(Heterogeneous)
ODI forColumnarDBs
ODI forIn-MemoryDBs
ODI forEngineeredSystems
ODI forHadoopNoSQL
ODI forHadoopPig & Oozie
ODI forSpark
ODI for …
1990’s
Eon of Scripts and PL-SQL
Era of Native SQL
Big Data Revolution
Oracle’s tool maturity and operational know-how for E-LT is unmatched
10x bigger footprint with E-LT than next closest competitor using “pushdown”
Simple and easy way to blend Hadoop and SQL E-LT execution from one tool
ODI forHadoopHive

Oracle Does Big Data Better: NoETLApproach
28
OneLogical Design:
ManyEngine Alternatives:
Data Engines:
Examples:
Engine I/O:
Best Use:
SQL/ OLTP Database
•Oracle DBMS
•Any OLTP DBMS
•DW Appliances
SSD / Diskbased
High volumes of transformations on relational data
MapReduce
•Hive / MR2
•Pig / Oozie/ MR2
SSD / Disk based
Huge batch-like transformations on any data types
In Memory(SQL / BigData)
•Oracle InMemory
•Hive / Tez/ YARN
•Spark / YARN
•ClouderaImpala
D/RAM;with various built in spill to disk approaches
Highlyinteractive data transformation patterns
StreamingBig Data
•Storm/ YARN
•Oracle Event Processor (OEP)
D/RAM; “always on” data pipeline
Verylow latencytransformations
Modern design studio for simple map development
Team-based GUI Tooling for work on Enterprise projects
Integrated lifecycle and metadata management
Automated support for Changed Data Capture
SEPARATE ETL ENGINE NOT REQUIRED!
Data Integrator

Oracle Does Big Data Better: Most Open & Heterogeneous
Hadoop HBase
Hadoop Hive/Flume
HP Enscribe
HP NonStop
HP Neoview
Hypersonic SQL
IBM DB2 iSeries
IBM DB2 UDB
IBM DB2 z Series
IBM Informix
IBM Netezza
JMS / MQ
Microsoft Access
Microsoft SQLServer
MySQL
Pivotal Greenplum
PostgreSQL
Salesforce.com
SAP BW / BI
SAP ERP / ECC
SAS
SQL/MP
SQL/MX
Sybase ASE
Sybase IQ
Teradata
Adaptive
Altova
Apache Hcatalog
Apache Hive/HQL
Borland
CA ERwin
ClouderaImpala
COBOL Copybook
DataStax
Embarcadero
EMC ProActivity
GentleWare
Google BigQuery
Grandite
HadaptHive
HortonworksHive
IBM Cognos
IBM DB2
IBM DataStage
IBM Discovery
IBM Federation Server
IBM Lotus Notes
IBM Netezza
IBM Rational Rose
IBM Rational Architect
InformaticaMetadata Mgr.
InformaticaPowerCenter
CoSORT
ISO SQL Standard (DDL)
MapRHadoop Hive
MicroFocus
Microsoft Access
Microsoft Office Excel
Microsoft Visio
Microsoft SQL Server
Microsoft SSIS
Microsoft Visual Studio
Microstrategy
Magic Draw
OMG CWM Standard
OMG UML Standard
Oracle BI Answers
Oracle BI Enterprise Edition
Oracle BI Server
Oracle DAC
Oracle Data Integrator
Oracle Data Modeler
Oracle Database
Oracle Designer
Oracle Hyperion Applications
Oracle Hyperion Essbase
Oracle Warehouse Builder
Pivotal Greenplum
PostgreSQL
QlikView
SAP BO Crystal Reports
SAP BO Designer
SAP BO Desktop Intelligence
SAP BO Repository
SAP BO Data Integrator
SAP BO Data Steward
SAP Master Data Management
SAP Sybase PowerDesigner
SAP Sybase ASE Database
SAS Data Integration Studio
SAS BI Server
SAS Information Map
SAS Metadata Management
SAS OLAP Server
Select
SparxArchitect
Syncsort
Tableau
Talend
Teradata
Tigris
Visible
W3C DTD & XSD Schema
Operational Integration (Movement / Transformation)
Metadata Harvesting (Glossary, Lineage & Impact Analysis)
Oracle Database
Oracle Exadata
Oracle Big Data Appliance
Oracle TimesTen
Oracle OLAP
Oracle Business Intelligence
Oracle BI Applications
Oracle E-Business Suite
Oracle JD Edwards Enterprise One
Oracle JD EdwardsWorld
Oracle Fusion Applications
Oracle Governance Risk and Compliance
Oracle Fusion AIA
OracleRetail Applications
Oracle Agile BI/ DW
OracleAgile PLM for Process
OracleiFlexFlexCUBE
Oracle iFlexMantas
Oracle HyperionApplications
Oracle PeopleSoft
Oracle Siebel CRM / OnDemand
Oracle Communications
Oracle WebLogic Server
Oracle Coherence Data Grid
Oracle SOA Suite
Oracle Enterprise Service Bus
+ open APIs and standards based meta-model

Oracle Does Big Data Better: Clear Business Benefits
30
Proven Technology
Better Architecture
Best for Oracle
•Unlike custom coding, a tools based approach is proven to result in lower cost long term operations
•Oracle GoldenGateis industry standard for Data Replication
•Oracle invented E-LT Pushdown processing and is 10x more widely deployed than competitors
•Oracle GoldenGateprovides the most scalable, native integration for database replication
•Oracle Data Integrator provides ultimate scalability and choice for Hadoop data transformations
•Consistent agent-based architecture avoids having multiple, incompatible engines (eg; old style ETL tools)
•Exadata–OGG and ODI are deeply integrated and are the only Replication and ETL processes certified to run on the appliance
•Big Data Appliance –deeply integrated technology part of core reference architecture
•Big Data Connectors –ODI included with core connector technologies for Hadoop
RISK
SCALE
COMPLETE
Heterogeneous Access

Copyright © 2014,Oracle and/or its affiliates. All rights reserved. |
Join the Community
#OOW14 #ODI12c #GoldenGate12c #EDQ12c
Oracle Data Integration blogblogs.oracle.com/dataintegration
Connect with Oracle on Social Media
OR connect via the web
Oracle Data Integration Home Pageoracle.com/goto/dataintegration

2014 Oracle Excellence Award Ceremony for Fusion Middleware Innovation
ORACLE FUSION MIDDLEWARE: CELEBRATE THIS YEAR'S MOST INNOVATIVE CUSTOMER SOLUTIONS
Tuesday, September 30, 2014 5:00-5:45pm YBCA Theater (next to MosconeNorth)
Session ID: CON7029

Oracle Fusion Middleware
The Cloud Platform for Digital Business
Cloud
On-Premise
DIGITAL ENGAGEMENT APPLICATION & DATA INTEGRATION
IDENTITYMANAGEMENT
SYSTEMS
MANAGEMENT
APPLICATION INFRASTRUCTURE & TOOLS
BUSINESS PROCESS MANAGEMENT
BUSINESS ANALYTICS
CONTENT & COLLABORATION
Web
Mobile
Social
Internet of Things

Questions and Answers

Tapping into the Big Data Reservoir (CON7934)

Tapping into the Big Data Reservoir (CON7934)

Recommended

More Related Content

What's hot (20)

Similar to Tapping into the Big Data Reservoir (CON7934) (20)

More from Jeffrey T. Pollock (13)

Recently uploaded (20)

Tapping into the Big Data Reservoir (CON7934)