In Memory Technology HANA
-
Upload
kishore-palakurthi -
Category
Documents
-
view
35 -
download
5
description
Transcript of In Memory Technology HANA
In-Memory Computing:
Realized Business Benefit and the Road Ahead
Noam Berda, Business Development Manager, Business Analytics & Technology
SAP Asia Pacific Japan
© 2011 SAP AG. All rights reserved. 3
Innovation Drives Success
Technology Innovations Enable Businesses to Become…
More flexible and
quick to action with
proper insight
Fiscally and operationally
efficient
Empowered at the
business user level to
make smart decisions
and act on these
demands
© 2011 SAP AG. All rights reserved. 4
In-Memory Computing
In-Memory Computing
Technology that allows the processing of
massive quantities of real time data
in the main memory of the server
to provide immediate results from
analyses and transactions
© 2011 SAP AG. All rights reserved. 5
Poll Question
By 2012 what would be the percentage of organizations (global 1000) that loads
data into in-memory technology for BI performance optimization?
• 5%
• 40%
• 70%
© 2011 SAP AG. All rights reserved. 6
Gartner
AGENDA
© 2010 SAP AG. All rights reserved. / Page 7
1. SAP’s In-Memory Computing Technology
2. How do you benefit?
3. SAP’s In-Memory Computing Offerings
4. Example in-memory computing scenario
© 2011 SAP AG. All rights reserved. 8
Don’t Bet on Database Performance
Reality Dictated by Physics
Improvement20101990
216Addressable
Memory
250x5MB/$
0.02MB/$
Memory
143x7.15MIPS/$
0.05MIPS/$
CPU
Technology Drivers
130MBPS
5MBPS
Disk
Data Transfer 25x
100x10Gbps
100Mbps
Network Speed
264 248x
Performance will continue to be an issue for analyzing large amounts
of information
© 2011 SAP AG. All rights reserved. 10
Discrete
The Inflection PointIn-Memory Computing
Hardware
Multi-core architecture
Massive parallel scaling
64-bit address space
Upto 2TB main memory
100GB/s data throughput
Row and
Column Store
Compression
Partitioning
Virtually unlimited size
Fast prefetch
Volatile and/or persistent
No Aggregate
Tables
Insert Only
on Delta
Software Today Tomorrow
10X compression
Massively parallel
processing
Cache
Disk
++
+ +
+
MemoryIn-Memory
Disk
Integrated
© 2011 SAP AG. All rights reserved. 11
SAP In Memory Solutions Available Today
SAP Applications Powered by In-Memory Technology
Today
SAP NetWeaver BW Accelerator (BWA)
SAP BusinessObjects Explorer, accelerated
SAP CRM Customer Segmentation
SAP Business ByDesign analytics
SAP Advanced Planner and Optimizer
SAP NetWeaver Enterprise Search
© 2011 SAP AG. All rights reserved. 12
Evolution of In-Memory at SAP
SAP NetWeaver®
BI Accelerator 7.0
2005
SAP NetWeaverBW Accelerator 7.20
2008
SAP BusinessObjectsExplorer Accelerated
2009
HANA 1.0
2010
HANA 1.5
2011
today
* Dates relates to RTC (release to customer)
© 2011 SAP AG. All rights reserved. 13
In-Memory Analytics in Action
Retail Customer – Acceleration of Complex Reports
Average improvement 96.7% improvement in
reporting time
Loading 1.1 million records/day with 36
months of history
Tetra Pak External
Global Information Management
Chris Rowley / 31 July 2008
14
From Department to Enterprise BI Business user adoption is the measure of success
2008 BI User Adoption
Tetra Pak External
Global Information Management
Chris Rowley / 31 July 2008
15
Successful Selling Strategies
© 2011 SAP AG. All rights reserved. 16
Preconfigured Analytical Appliance
■ In-Memory software + hardware
(HP, IBM, Fujitsu, Cisco)
In-Memory Computing Engine Software
■ Data Modeling and Data Management
■ Real-time Data Replication Data Services for SAP
Business Suite, SAP BW and 3rd Party Systems
Capabilities Enabled
■ Analyze information in real-time at
unprecedented speeds on large volumes of non-
aggregated data
■ Create flexible analytic models based on real-
time and historic business data
■ Foundation for new category of applications
(e.g., planning, simulation) to significantly
outperform current applications in category
■ Minimizes data duplication
SAP High-Performance Analytic Appliance (SAP HANA)Architecture
BICS SQL MDXSQL
Modeling
Studio
Real–Time
Replication
Services
Data
Services
SAP HANA
SAP BusinessObjects Other Applications
SAP NetWeaver
BW
SAP Business
Suite3rd Party
In-Memory Computing Engine
Calculation and
Planning Engine
© 2011 SAP AG. All rights reserved. 17
Technical Overview (1 of 2)
Calculation models – Extreme Performance and Flexibility with Calculations on the fly
Calculation Engine
Calculation Model
Distributed Execution Engine
Row Store Column Store
SQL MDXSQL
Script
Plan
Modelother
Compile & Optimize
Physical Execution Plan
Logical Exection Plan
Parse
SAP in-memory computing engine
Calculation Model
Artifacts from domain specific languages
(SQL, MDX, etc) get translated into a
common representation (calc model)
A calc model is a directed graph
representing the flow of data from input to
output passing various operations
A calc model can be generated on the fly
based on an expression provided by a client
(Excel providing an MDX)
A calc model can also define a
parameterized calculation schema for highly
optimized reuse
A calc model supports scripted operations
© 2011 SAP AG. All rights reserved. 18This presentation and SAP‘s strategy and possible future developments are subject to change and may be changed by SAP at any time for any reason without notice. This document is
provided without a warranty of any kind, either express or implied, including but not limited to, the implied warranties of m erchantability, fitness for a particular purpose, or non-infringement
ABAP AS
App
DB
ABAP AS Next Generation
Next Generation Apps
SAP HANA
Data in
memoryRuntime
Procedure
code
Program
code
compile
& deploy
Fast data
transfer
Technical Overview (2 of 2)
Applications – Tight coupling between application server and SAP HANA
Today Mid-Term (Plan)Tight Coupling
With large data volumes,
reading information
becomes a bottleneck
Next generation
applications will delegate
data intense operations
The runtime environment
executes complex
processes in memory
In memory computing
returns results by pointing
apps to a location in
shared memory
AGENDA
© 2010 SAP AG. All rights reserved. / Page 19
1. SAP’s In-Memory Computing Technology
2. How do you benefit?
3. SAP’s In-Memory Computing Offerings
4. Example in-memory computing scenario
© 2011 SAP AG. All rights reserved. 20
154,000customers
SAP HANA in Action at a CPG Company
Dunning Process Acceleration
1.8M 1,00070,000rows of data
B2B customers
collection notices 13 seconds
77 minutes
Standard System In-Memory System
356xfaster
© 2011 SAP AG. All rights reserved. 21
SAP HANA for Data Intensive Point of Sale Analysis
Large CPG company wants to
analyze all their POS of data to
predict demand
Target -stock shelves with 48 hour
turn-around
Data Set is 460 Billion records (40
Terabytes)
Unable to analyze data using
current database platform
10 HANA blades with 500GB per
blade & 2TB SSD Storage
HW Cost = $532K
SAP BusinessObjects Explorer
120 TB in traditional system =
40TB Hana
20x Faster Analysis with 200x
Better Price/Performance
Moved from 5 days down to 2
days for shelf turnaround
Eliminates out of stock
scenarios during promotions
Challenge ResultsSolution
© 2011 SAP AG. All rights reserved. 22
NRI Japan
Real Time Traffic Analysis by geo and time- Collecting location data of over 12,000 contracted taxis (15m rec/ day).
- World first probe data-based real-time navigation service.
- Drive less be more efficient.
- Reduce hazards and traffic congestion.
Why SAP HANA?- Existing DB cannot handle such high volume of analysis in real-time.
- HANA can deliver real-time analysis on large amounts of data.
APJ HANA Win Story
AGENDA
© 2010 SAP AG. All rights reserved. / Page 23
1. How do you benefit?
2. SAP’s In-Memory Computing Technology
3. SAP’s In-Memory Computing Offerings
4. Example in-memory computing scenario
© 2011 SAP AG. All rights reserved. 24
SAP ECC
4.6c / 4.7 / ECC 6.0(or CRM, SRM, SCM)
Traditional DB
Oracle, DB2, SQL
Server, MaxDB
EDW
SAP BW / Custom
Traditional
DBBWA
BI
D
D
D
Data
Marts BI
ETL
Update
Daily
Today’s System – After Event Analysis
© 2011 SAP AG. All rights reserved. 25
SAP ECC
4.6c / 4.7 / ECC 6.0(or CRM, SRM, SCM)
Traditional DB
Oracle, DB2, SQL
Server, MaxDB
EDW
SAP BW / Custom
Traditional
DBBWA
BI
D
D
D
Data
Marts BI
ETL
Update
Daily
HANA 1.0
~10x compression
100x faster analysisNear Real-time
BI
Step 1. Install and Run HANA in parallel with ECC
Q4
2010
Any Data
Real-Time Operational Analytics
© 2011 SAP AG. All rights reserved. 26
SAP ECC
4.6c / 4.7 / ECC 6.0(or CRM, SRM, SCM)
Traditional DB
Oracle, DB2, SQL
Server, MaxDB
SAP BWBI
H
Data
Mart BI
HANA 1.5
~10x compression
100x faster analysisNear
Real-time
BI
Step 1. Install and Run HANA in parallel with ECC
Step 2. Use HANA as in-memory EDW and Datamart for All Data
Any Data
SAP
Non-SAP
Accelerated
Accelerated
2011In-Memory EDW/Data Mart + Accelerated BI
© 2011 SAP AG. All rights reserved. 27
SAP ECC
4.6c / 4.7 / ECC 6.0(or CRM, SRM, SCM)
Traditional DB
Oracle, DB2, SQL
Server, MaxDB
SAP BWBI
H
Data
Mart BI
HANA 1.5
~10x compression
100x faster analysisNear
Real-time
BI
Step 1. Install and Run HANA in parallel with ECC
Step 2. Use HANA as in-memory EDW and Datamart for All Data
Step 3. Deliver new In-Memory Applications (e.g. BPC, Demand Planning)
Any Data
SAP
Non-SAP
Accelerated
Accelerated
Applications
2011In-Memory Applications – Redefine Planning
© 2011 SAP AG. All rights reserved. 28
SAP ECC
4.6c / 4.7 / ECC 6.0(or CRM, SRM, SCM)
SAP BWBI
H
Data
Mart BI
HANA 2.0
~10x compression
100x faster analysis
BI
Step 1. Install and Run HANA in parallel with ECC
Step 2. Use HANA as in-memory EDW and Datamart for All Data
Step 3. Deliver new In-Memory Applications (e.g. BPC, Demand Planning, etc.)
Step 4. Zero Latency Replication – Real-Real Time
Step 5. Eliminate All 3rd Party DBs – Run In-Memory
Any
Data
SAP
Non-SAP
Accelerated
Accelerated
Applications
FutureIn-Memory ECC
© 2011 SAP AG. All rights reserved. 29
HANA 1.0 – Scenarios
1. SAP Environment 3. non-SAP Environment
HANA
SAP ECC
4.6c / 4.7 /
ECC 6.0
(or CRM,
SRM, SCM)
Data Replicator
or
ETL
EDW 1
EDW 2
EDW 2
ETL
2. Mix Scenario
BW
SAP Business Objects BI 4.0
In-memory
Business Scenario
© 2011 SAP AG. All rights reserved. 30
HANA Certified HW Providers
Server Memory
Configuration
Data Volume(x5 compression)
# Rows(500 byte / row)
128 GB 320 GB 687 million
256 GB 640 GB 1.3 billion
512 GB 1280 GB 2.7 billion
1 TB 2.5 TB 5.5 billion
Clustered
Environment
(MPP)
unlimited unlimited
AGENDA
© 2010 SAP AG. All rights reserved. / Page 31
1. How do you benefit?
2. SAP’s In-Memory Computing Technology
3. SAP’s In-Memory Computing Offerings
4. Example in-memory computing scenario
© 2011 SAP AG. All rights reserved. 32
Telco - Customer Behavior Analytics
Telco challenges today are to measure and understand customer usage behaviors
• Focused marketing
• Conversion to post paid
• Increase prepaid spend
• Understanding customer behavior
• Fraud
• Real time data volumes
• Analysis speed
=
© 2011 SAP AG. All rights reserved. 33
Analyze Prepaid Customer Behavior5
$
10
$
5$
3 days 5 days
Customer Id Topup spend Date
1 5$ 8 days ago
1 10$ 5 days ago
1 5$ Today
• Total Topup spend - 20$
• Total days - 8 days
• Average spend - 2.5$ / day
© 2011 SAP AG. All rights reserved. 34
Analyze Prepaid Customer Behavior5
$
10
$
5$
3 days 5 days
1 day
2 days
2 days
3 days
• Detailed analysis of customer usage behavior patterns
• Focus marketing campaigns (Geography, Usage behavior etc..)
• Accelerate usage time = Customer spend money faster
• Reduce dead time = Customer always spend
• Link with dealer performance (Lower -> adjust incentives)
© 2011 SAP AG. All rights reserved. 35
Analyze Prepaid Customer Behavior
Demo
© 2011 SAP AG. All rights reserved. 36
Analyze Prepaid Customer Behavior
Screen Shoots 1
© 2011 SAP AG. All rights reserved. 37
Analyze Prepaid Customer Behavior
Screen Shoots 2
© 2011 SAP AG. All rights reserved. 38
Analyze Prepaid Customer Behavior
Balance
(OLTP)
Topup history
(OLAP)
Mill
ion
Re
co
rds
Inse
rt \
Se
lect \
Upd
ate
Bill
ion
Re
co
rds
Inse
rt O
nly
\A
gg
reg
atio
n
Billing /
EDW
1. Read Current Balance
2. Calculate usage / dead time
3. Write To History
4. Update Balance
© 2011 SAP AG. All rights reserved. 39
Why HANA For Customer Behavior Analytics
Column / Row ( OLAP / OLTP)
technology in same DB / Transactions
• Develop new type of scenarios
• Reduce solution complexity
• Low TCO
In-Memory • Fast Query
• Fast DB operations
• No pre aggregation
• Scalable
• Easily developed
Compression • Reduce TCO
• Small HW footprint
© 2011 SAP AG. All rights reserved. 40
The Future - HANA Content + Applications
Strategic Workforce
Planning
Smart Grid Analytics
COPA case (includes financial line item reporting)
Inventory Movement
Billing Management
Smart Grid Analytics
SAP BW Powered by HANA
SAP BPC Powered by HANA
Order to Cash analysis
Work Force Planning
High Tech: Operational Reporting for Operations and Finance, Sales Pipeline Analyzer
Banking: Bank Analyzer
20+ SAP Projects - H2 2011 for Ramp-up & GA Early 2012
© 2011 SAP AG. All rights reserved. 41
Customer Scenario Data Source Benefits
Chemicals
(DE)
profitability analysis ERP 8TB ERP 6.0 EhP4
on Oracle/AIX
query performance,
load time/latency
Manufacturing
(DE)
material disposition,
production line utilization
ERP 1TB
BW 2TB
ECC 6.0
on DB2/AIX
BW 7.0
on DB2/AIX
reduce extraction impact on
ERP,
query performance
Apparel
(US)
demand planning,
order planning
ERP 2TB R/3 4.6c on Oracle
BW 7.0
Teradata
load time/latency (currently 24h
with Teradata, target is 15 min)
CPG
(US)
multiple ABAP reports
(MB51 material document list,
FAGLL03 GL line item display, KE27
periodic valuation), CPG warehouse
scorecard, trade latest estimate,
customer service dashboard
ERP 10-200 mio.
records of relevant
data
ERP 6.0 EhP4
on Oracle/Linux
BW
query performance
Manufacturing
(CH)
customer contact listing,
open quotes/open orders
ERP
200 m. orders, 9 m.
contacts
ERP 6.0
on Oracle/Linux
impact on ERP,
query performance
SAP GFO
(DE)
sales pipeline analysis CRM 2,7TB CRM load time/latency, performance
CPG
(US)
profitability analysis ERP 6.0
on DB2 on AIX
query performance,
load time/latency
SAP HANA – Example Pilot Customers
© 2011 SAP AG. All rights reserved. 43
SAP Applications using In-Memory Technology
Next wave of technology innovation
Combined in-memory analytics & transactional applications
Available today, delivered without disruption
Continuous real-time link between insight, foresight and action
Plan Smarter
…run faster …perform better
To Empower Your Organization…
…plan smarter
© 2011 SAP AG. All rights reserved. 44
3 Steps for HANA Project
Step 1
Introduction to HANA
Duration: 2 hours
Agenda:
overview presentation about
HANA and SAP in-memory
strategy.
Target Audience:
• Line of business managers
• Reporting managers
• Solution architects
• Data warehouse team
Step 2
Scenario Workshop
Duration: 1-2 days
Agenda:
Identify potential HANA
scenarios within customer
landscape and provide an
overview on BI4.0
Target Audience:
• Line of business managers
• Reporting managers
• Solution architects
• Data warehouse team
Step 3
Scenario Development
Duration: 60 days
Agenda:
Develop potential scenario on
HANA and BI4.0
© 2011 SAP AG. All rights reserved. 45
More information ...
Visit
http://www.sap.com/platform/in-memory-computing/index.epx
SAP Developer Network
http://www.sdn.sap.com/irj/sdn
Hasso Plattner Institute, Potsdam
http://www.hpi.uni-potsdam.de/
© 2011 SAP AG. All rights reserved. 46
Questions?