41
Actian Vectorwise Simply Fast = High performance + Affordable Cost Christian RAZA SEMEA Sales Director [email protected] XpandIT / Big Data ecosystem / November 27 2013

Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

Embed Size (px)

DESCRIPTION

We live in the Age of Data. Now more than ever, it is crucial that organizations can connect, analyze and act on the vast amounts of data that surrounds them in order to succeed long-term. This session will discuss the Age of Data and how companies can deploy technology such as Actian ParAccel SMP, a fast analytic database platform that runs on standard hardware, in order to run sophisticated, unconstrained analytics on massive amounts of data (structured, unstructured, Hadoop etc) and turn their data into business value. Christian Raza - Director of Sales SEMEA, @Actian Corporation Actian presentation during the Pentaho & Big Data Ecosystem - Live Seminar 2013

Citation preview

Page 1: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

Actian Vectorwise Simply Fast = High performance + Affordable Cost

Christian RAZA

SEMEA Sales Director

[email protected]

XpandIT / Big Data ecosystem / November 27 2013

Page 2: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

Our Vision

2 © 2013 Actian Corporation

Page 3: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

3 3 Confidential © 2013 Actian Corporation

The Age Of Data

Page 4: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

• “Data are becoming the new raw material of

business.“

Craig Mundie, Microsoft

• “Data tap has been turned on and will never

be turned off.”

Mike Hoskins, Actian CTO, 2013

• “You can have data without information, but

you cannot have information without data.”

Daniel Keys Moran, Writer

• “It is a capital mistake to theorize

before one has data.”

Sherlock Holmes, Sir Arthur Conan Doyle

• “Torture the data,

and it will confess to anything.”

Ronald Coase, Economics, Nobel Prize

4 4 Confidential © 2013 Actian Corporation

The Age of Data

Page 5: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

5

Actian Big Data Platform

© 2013 Actian Corporation

• “Data is the new oil.“

Clive Humby,

ANA Senior marketer’s summit

• “Information is the oil of the

21st century, and analytics is

the combustion engine.”

Peter Sondergaard,

SVP at Gartner

Page 6: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

Big Data pipeline

6

Actian Big Data Platform

© 2013 Actian Corporation

Open Data

Shared Data

Operational Data Big Data

Reservoir

Analytic

Applications

Graph Based

Applications

Search Based

Applications

Reporting

Applications

Page 7: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

7

Extreme Reporting

© 2013 Actian Corporation

Page 8: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

8

Extreme Analytic

© 2013 Actian Corporation

Page 9: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

Actian Vectorwise

9 9 © 2013 Actian Corporation

Page 10: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

10

Why Vectorwise?

© 2013 Actian Corporation

Page 11: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

11

Vectorwise: started with an academic project

© 2013 Actian Corporation

first

colomn store DBMS

1990 1993 2005 2010

Vectorwise MonetDB MonetDBX100

• Patent USPTO #20100235335 (Peter BONCZ): Column-store database architecture utilizing positional delta tree

• Partnership with academic

• multi-core parallelism

• just-in-time compilation of predicates

• non-intrusive query execution on compressed data

• cooperative Scans

Page 12: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

12

Vectorwise

© 2013 Actian Corporation

New generation

Colummn DBMS

In CPU acceleration Ingres modules

encapsulation

Page 13: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

13

Vectorwise: simplified technical integration

© 2013 Actian Corporation

Connectivity

• .Net, JDBC, ODBC, PHP, Perl…

• ETL: Stambia, Talend, SyncSort, Informatica, DataStage…

• BI: BiBoard, BO, Cognos, MicroStrategy, Hyperion, JasperSoft, Pentaho, SpagoBI, TableauSoftware, YellowFin,…

• Analytics: SAS, SPSS, RapidMiner, RevolutionAnalytics,…

Standard SQL

• No proprietary SQL

• Support de ANSI SQL-92

DBA tools

• Actian Director

• BackUp / Restore

• Utilitaires Load/Unload (copy, vwload)

Page 14: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

Benchmark

14 © 2013 Actian Corporation

Page 15: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

15

Benchmark TPCH : 3 world records

© 2013 Actian Corporation

Page 16: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

16

Benchmark TPCH: Vectorwise VS MS SQL Server

© 2013 Actian Corporation

TPCH 1TB MS SQL Server Ingres VectorWise

CPU Type:Intel Xeon Processor E7-8870

2.40GHzIntel Xeon E7-8837 2.67GHz

Server: IBM System x3850 X5 8P Dell PowerEdge R910

Total # of Processors: 8 4

Total # of Cores: 80 32

Database ManagerMicrosoft SQL Server 2008 R2

Enterprise EditionVectorWise 1.6

Operating SystemMicrosoft Windows Server 2008 R2

Enterprise EditionRedHat Enterprise Linux.6.1

Total Storage/Database Size Ratio: 7.00 2.34

Metric 173,962 QphH@1000GB 436,789 QphH@1000GB

Price/Performance 1.37 USD per QphH@1000GB .88 USD per QphH@1000GB

Availability Date 20-May-2011 30-Jun-2011

Page 17: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

17

Benchmark TPCH: Vectorwise VS Oracle

© 2013 Actian Corporation

TPCH 1TB Oracle Ingres VectorWise

CPU Type: SPARC64 VII+ 3000MHz Intel Xeon E7-8837 2.67GHz

Server: SPARC Enterprise M8000 Server Dell PowerEdge R910

Total # of Processors: 16 4

Total # of Cores: 64 32

Database ManagerOracle Database 11g R2 Enterprise

Edition with PartitioningVectorWise 1.6

Operating System Oracle Solaris 10 RedHat Enterprise Linux.6.1

Total Storage/Database Size Ratio: 11.20 2.34

Metric 209,533 QphH@1000GB 436,789 QphH@1000GB

Price/Performance 10.13 USD per QphH@1000GB .88 USD per QphH@1000GB

Availability Date 22-sept.-11 30-Jun-2011

Page 18: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

18

Benchmark TPCH: Vectorwise VS SybaseIQ

© 2013 Actian Corporation

TPCH 1TB Sybase IQ Ingres VectorWise

CPU Type: AMD Opteron 8439 SE 6-Core 2.8GHz Intel Xeon E7-8837 2.67GHz

Server: IBM Power 780 Model 9179-MHB Dell PowerEdge R910

Total # of Processors: 8 4

Total # of Cores: 48 32

Database ManagerSybase IQ Single Application Server

Edition v.15.1 ESD #1VectorWise 1.6

Operating System Red Hat Enterprise Linux 5.3 RedHat Enterprise Linux.6.1

Total Storage/Database Size Ratio: 15.18 2.34

Metric 102,375 QphH@1000GB 436,789 QphH@1000GB

Price/Performance 3.63 USD per QphH@1000GB .88 USD per QphH@1000GB

Availability Date 01-Feb-2011 30-Jun-2011

Page 19: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

Reporting / Dashboard

19 © 2013 Actian Corporation

Page 20: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

20

Customer Benchmark: Vectorwise VS Teradata

© 2013 Actian Corporation

o This performance benchmark was performed by Leroy Merlin France to evaluate DBMS during a massive workload (load test duration: 1 hour). o The target workload represents the reporting/dashboard activity of 10 000 sales force between 08:00 and 09:00 AM of. The peak workload was estimated at 600 concurrent users which gives 1200 queries per minute.

o Vectorwise VS Teradata benchmark:

• Vectorwise 2.5 / Linux / 8-Core / 32 Go RAM • Teradata 6680 with 2 active nodes and SSD (24-core / 192 GB RAM)

Page 21: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

21

Customer Benchmark: Vectorwise VS Teradata

© 2013 Actian Corporation

o Result #1:

• Vectorwise was faster than Teradata on all single query runs o Result #2 :

• Vectorwise was 2x faster than Teradata at 50 queries / minute o Result #3 :

• With Vectorwise between 1 and 30 queries / minute, queries are running as fast as in single run

oResult #4:

• At 70 concurrent users (140 queries / minute), Teradata was 100% full • Vectorwise reach the target of 600 concurrent users (1200 queries / minute) • Vectorwise was loaded up to 2000 queries / minute during 2:30 hours (35% queries were in time out)

Page 22: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

22

Customer Benchmark: Vectorwise VS Teradata

© 2013 Actian Corporation

Page 23: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

23

Booster application

© 2013 Actian Corporation

Page 24: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

24

Booster application

© 2013 Actian Corporation

Page 25: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

25

Booster application

© 2013 Actian Corporation

Page 26: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

26

Booster application

© 2013 Actian Corporation

Page 27: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

Big Data

27 © 2013 Actian Corporation

Page 28: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

28

Iscool Entertainment

© 2013 Actian Corporation

Page 29: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

29

Social Gaming

© 2013 Actian Corporation

Page 30: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

30

Big Data

© 2013 Actian Corporation

Page 31: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

31

Big Data Analysis

© 2013 Actian Corporation

Page 32: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

32

Technical architecture

© 2013 Actian Corporation

Page 33: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

Vectorwise users

33 © 2013 Actian Corporation

Page 34: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

Vectorwise users

34 © 2013 Actian Corporation

Page 35: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

Conclusion

35 © 2013 Actian Corporation

Page 36: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

The Age Of Data

Turning Data into Action

36 Confidential © 2013 Actian Corporation

“Data matures like wine, applications like fish.” - James Governor

Page 37: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

Actian Corporation

37 © 2013 Actian Corporation

Page 38: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

70’s

38 © 2013 Actian Corporation

Page 39: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

70’s

Tandem NonStop SQL

HP Allbase

Cullinet

Britton-Lee

Wang PACE

CA-Universe

Teradata

Informix

39 © 2013 Actian Corporation

Ted Codd

1970 1980 1990 1995

Michael Stonebraker

UC Berkeley

IBM

System R

Oracle

Ingres

Sybase

MS SQL

MySQL

PostgreSQL

Page 40: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

40

Actian: 30-year old start up

© 2013 Actian Corporation

• +30 years of DBMS experience UC Berkeley, Relational Technology Inc., Ingres Corporation, Ask Group, Computer Associates, Ingres Corporation

• Independent since 2006 Ingres became Actian in 2011

• +10 000 customers in 58 countries

• 150 M$ of revenue

• Profitable, +30 M$ of cash

• 500 employees

• 3 acquisitions between December 2012 and April 2014

Page 41: Unconstrained Analytics in the Age of Data – Delivering High-Performance Analytics with ParAccel SMP

Disclaimer

This document is for informational purposes only and is subject to change at any

time without notice. The information in this document is proprietary to Actian and

no part of this document may be reproduced, copied, or transmitted in any form or

for any purpose without the express prior written permission of Actian.

This document is not intended to be binding upon Actian to any particular course of

business, pricing, product strategy, and/or development. Actian assumes no

responsibility for errors or omissions in this document. Actian shall have no liability

for damages of any kind including without limitation direct, special, indirect, or

consequential damages that may result from the use of these materials. Actian

does not warrant the accuracy or completeness of the information, text, graphics,

links, or other items contained within this material. This document is provided

without a warranty of any kind, either express or implied, including but not limited to

the implied warranties of merchantability, fitness for a particular purpose, or non-

infringement.

41 Confidential © 2013 Actian Corporation