Sei sulla pagina 1di 49

SAP BW Archiving and Data

Aging Strategies

Ron Silberstein Mike Eacrett


Platinum Consultants
SAP Netweaver RIG US – Business Intelligence
SAP Labs, LLC
Agenda

Information Lifecycle Management

Information Lifecycle Management

Data Aging Strategies in a


Heterogeneous Application Environment
SAP ERP Components
SAP Business Information
Warehouse

Summary

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 2
Agenda

Information Lifecycle Management

Information Lifecycle Management

Data Aging Strategies in a


Heterogeneous Application Environment
SAP ERP Components
SAP Business Information
Warehouse

Summary

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 3
Increasing Volume of Databases

Meta Group

“70% of all enterprise data currently resides in


databases”

Gartner Group

“Multi-Terabyte databases are a reality today – and will


grow to hundreds of TB by 2006”

“Database applications – like SAP – are a key driver of


storage growth”

“…growing at 64% rate”

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 4
Distribution of Activity vs. Age of Data

Frequency
of access

reads

updates

Age of data

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 5
Distribution of the memory costs

“ Hard disk costs do not even represent a quarter of the memory costs “
(Giga Information Group)
Misc
(Purchasing,
training)
10%

Personel Hard disk


45% 23%

Storage-
Environment Mangement
(Electricity, Space)
(Soft- & Hardware)
3% 19%
“ Administrative expense for 1 Terabyte of memory are appropriate for
five to seven times more higher than the memory costs themselves “
(Dataquest/Gartner)

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 6
Data Lifecycle vs. Storage Media Types/Costs

Cost Performance
Access Frequency RAM

DASD

Virtual
high Tape
Access Centric
Tape
Capacity Centric
Tape
low
time
days months years
Online- Near-line Offline
Storage Storage Archive

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 7
How to combine all this in a Data Aging Scenario?

Online Offline/
Query
Near-line File System
Read

Database

Data Objects

Storage
Management
Application System
Data

Archive Files

Move
Save

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 8
Benefits of a sound Data Archiving Strategy

Value proposition
for deploying Data Archiving

1 Availability

2
Data availability: faster data
loads, rollups, change runs, etc.
System availability: Faster and Resource consumption
simpler software and release Reduction of the hardware costs
mgmt and upgrades. Reduced for hard disks, main memory and
backup & recovery times. CPU, and the costs of the

3
system administration

Performance
Faster Query response times.
Faster Data load times.

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 9
Agenda

Information Lifecycle Management

Information Lifecycle Management

Data Aging Strategies in a


Heterogeneous Application Environment
SAP ERP Components
SAP Business Information
Warehouse

Summary

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 10
BW Architecture: Layers & accessibility

Operational Data Store Data Warehouse Multidimensional Model


Operational Reporting Non volatile Multidimensional analysis
Near Real-Time / Volatile Granular Aggregated view
Granular Historical foundation Integrated
Built with ODS Objects Integrated
Built with InfoCubes
Built with ODS
Objects

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 11
Transaction Data Processing into the BW ADM Layer
SAP BW InfoCube
Material dimension
Customer dimension
4712
InfoCubes A
4713

Amount
Customer Material Time Company Information
ODS ODS
Currency
Object
Objects A 4712 200211 100
A 4713 200211 150
BW Architected Data Mart Layer
Time dimension
other InfoCubes
other InfoCubes 200211

monthly weekly daily

SAP ERP daily Process


Data

Master Amount
Master data
data
Customer Time Doc No Pos Material Local
Currency
Documents
Documents A 20021107-10am 1 10 4711 100 - New booking
Change A 20021107-3pm 1 10 4712 200 - Correction booking
Change Docs
Docs
A 20021107-4pm 2 10 4713 300 - New booking

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 12
Transaction Data Processing with BW EDW Layer
SAP BW Customer dimension InfoCube Material dimension
A 4712
4713
InfoCubes Amount
Customer Material Time Company
Currency Information
A 4712 200211 100
BW Architected Data Mart Layer A 4713 200211 150
Time dimension
other InfoCubes
other InfoCubes 200211

monthly weekly daily


BW Enterprise Information
Amount Amount
Data Warehouse Base
Customer Time DocNo Pos Material Local Company
Layer Currency Currency
A 20021107-10am 1 10 4711 100 50
ODS A 20021107-3pm 1 10 4712 200 100
A 20021107-4pm 2 10 4713 300 150
ObjectODS
Objects

daily Process
SAP ERP
Amount Data
Customer Time DocNo Pos Material Local
Master
Master data
data Currency
A 20021107-10am 1 10 4711 100 - New booking
Documents
Documents A 20021107-3pm 1 10 4712 200 - Correction booking
A 20021107-4pm 2 10 4713 300 - New booking
Change
Change Docs
Docs

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 13
Impact of Information Base EDW Layer

Pros Cons
Designed for the future Increasing Data Volume
Foundation of data for Data redundancy
future development Increased data management
Easier delta load control and coordination
Reduce redundancy in very Increased DW administration
granular data Upfront design considerations
Greater control of data
distribution
Fine tuning of data availability
in InfoProviders
Archiving of base data possible

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 14
Bill Inmon’s: Enterprise Data Warehousing

SAP BW DSS Applications


Departmental Data Marts Powered by SAP NetWeaver
Acctg Finance
ERP
Marketing Sales
ERP
ERP
CRM
ETL
Staging Area

Changed
eComm.
Data
EDW Bus. Int.

Exploration
Global warehouse/
ODS data mining
ERP
Corporate Oper.
Applications
Mart cross media
local Granularity
Storage mgr
ODS Manager

Session Near line


Analysis
Storage
Dialogue
Manager
Cookie Archives
Cognition Web Logs
Preformatted
Internet dialogues Source:Bill Inmon
© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 15
Agenda

Information Lifecycle Management

Information Lifecycle Management

Data Aging Strategies in a


Heterogeneous Application Environment
SAP ERP Components
SAP Business Information
Warehouse

Summary

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 16
Data Archiving and Storage within ERP Environment

mySAP System External Storage System

ArchiveLink Storage System


Database

Data Objects File System HSM-


System
Application data
Alternative
Storage

SAP Third Parties (optional)

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 17
Archive Development Kit – ADK

Development tool for data archiving solutions


From SAP
Integration of archiving functionality into SAP applications
From customers
Archiving solutions for customer tables and enhancements

Application
data

Database mySAP
Applications

ADK Code page adjustments, alphanumeric format, structural


changes, compression, file handling, job control, …

Archive Files

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 18
The Archiving object

Definition of the data associated to logical work units

All programs that are necessary for archiving,


e.g. Testing , Reading, Writing and Delete programs

Customizing of data selection Archiving Object


and processing

Customizing Data

Program

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 19
Data reduction in the ERP field!

Is the data Yes Can Data be Yes Consolidate


still needed? consolidated? Data!

No No Yes

Access to Can the data be archived?


Productive
data switched
off!
No Yes

Data remains Archive


in the
Data!
database!

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 20
Accessing Archived Data

Database Search DB + Archive


Business processes Display work units

DRB Application
SAP System Archive files

AS
100%
ADK

Index / Archive info structure

Indexing Technical display for all Archive object


Search on archived data Display of Object lists

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 21
Archive Information System

Flexible creation of Archive data indices


Freely configurable info. structures are stored in transparent DB
tables
Manually built with selected archive files
Automatically built during the delete phase of a archiving run

Search of archived data

Technical display for all archiving objects

Display of object lists


Based on info. structures
Ad-hoc display of selected archive files
Creation of display variants

Generic tools for customer developments

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 22
Definition of Data Objects and Influence on Indexing

ODS Object Archive File Archive index

Key Data Object key


YEAR YEAR
DOCNO CUSTOMER

Data Object 1
YEAR DOCNO CUSTOMER AMOUNT YEAR CUSTOMER FILE OFFSET
YEAR DOCNO CUSTOMER AMOUNT 1995 10000450 4710 100.00 1995 4710 ... ...
... 1995 10000452 4710 200.00 1995 4711 ... ...
1995 10000450 4710 100.00 1995 10000455 4710 400.00 1995 4712 ... ...
1995 10000451 4711 250.00
1995 10000452 4710 200.00 Data Object 2
1995 10000453 4712 150.00 YEAR DOCNO CUSTOMER AMOUNT
1995 10000454 4711 300.00 1995 10000451 4711 250.00
1995 10000455 4710 400.00 1995 10000454 4711 300.00
...
Data Object 3
YEAR DOCNO CUSTOMER AMOUNT
1995 10000451 4712 150.00
...

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 23
Document Relationship Browser (DRB)

User overview
• Archive Explorer
• Transaction ALO1 Document
Relationship
• User role SAP_DRB
• Application transactions
Browser

Display

Database

Display of company process


data from the archive files and Application Infra- Archive
data structure files
the underlying DB structures

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 24
Agenda

Information Lifecycle Management

Information Lifecycle Management

Data Aging Strategies in a


Heterogeneous Application Environment
SAP ERP Components
SAP Business Information
Warehouse

Summary

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 25
Typical Data Warehouse Problems

End-User Challenges
Making timely, informed business decisions
- Users cannot wait for historical data to be restored
- Transparent access to data for regular reporting and ad-hoc
analysis

IT Management Challenges
Meeting end-user data demand while managing cost
- High costs of adding/managing online disk storage
- High costs of backup and recovery – especially when data is
infrequently accessed
- Data protection and availability

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 26
Motivation for a Data Aging Strategy : Benefits

Costs
offline vs. online storage costs
System usage overhead – CPU, Memory, etc
Control of system growth
System Availability vs. costs
Data availability – faster rollups, change runs, etc
System availability – less downtime for backups, upgrades,
etc
Performance vs. costs
Faster load times
Faster query times
See also "Scalability with SAP® Business Information Warehouse“
at http://www.sap.com/solutions/bi/brochures/
Legal Requirements

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 27
SAP BW Customer motivation to archive

Customer requirements (derived from ASUG):


Data objects relevant for archiving
InfoCubes
ODS Objects
PSA
Master data

Functionality
Both Archiving and Data Deletion (without
archiving)
Select data based on any criteria
Automatically scheduled on a periodic basis
Restoring of archived data

Data retention time


3 to 5 years in InfoCube and ODS Objects

Consistent archiving processes


© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 28
How to avoid high Data Volumes in a DW Environment?

Can the
Is Data Yes Data be Yes Aggregate
still needed? aggregated? Data!

Maybe No Yes

Can the data be archived?

No Yes

We need to address Data Archive


the no longer needed remains in the
Data!
AND the maybe scenario! database!

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 29
Data Aging Strategy Implementation

Data aging is a strategy for managing data over time, balancing


data access requirements with TCO.
Each data aging Strategy is uniquely determined by the customer’s
data and the business value of accessing the data.

Which tools should I consider to use?

Online Database Near line BW data


Storage Storage archiving

Frequently read
/updated data

Infrequently
read data

Very rarely
read data

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 30
Where is Archiving and Near-line Storage applicable?
Access frequency

Archiving (BW 3.X)


Decreasing TCO For analysis, archived data must be
reloaded first again into the BW
reads
database
Archiving with SAP BW
Reduction in costs of data retention
on alternative media

Data age

Access frequency

Decreasing TCO
Near-line Storage (BW 4.X*)
Direct accesses to data to alternative
reads
storage media for Queries
Near Line Storage Performance and data retention costs
for access aged data can be minimized

Data age * Pilot project possible for BW 3.X


© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 31
SAP BW Archiving Object Architecture

activates
BW Repository

generates

Archiving Object
Archive
DataManager

reads Administration
InfoCube Write (SARA)
deletes Schedules
Delete
ODS
Read Datamart
Object Extractor

ADK

File system, CMS, HSM

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 32
Technical Architecture: BW and ADK Integration

File system External Storage Systems


BW database

InfoCubes Archiving Object


• Optical
Data Objects
Storage system

Archive
Files
• HSM
System
ODS Archiving Object
object
Data Objects

• Storage-
system
Archive
Files

SAP Functionality - ADK 3rd party products


© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 33
SAP BW data archiving technology : Write process

Time Slot archiving is a powerful tool for selection of time


dependent transactional archive data.
ODS Object
Current Date: Mar 2003
Sep 2002 Archive file

Oct 2002 Sep 2002


Nov 2002 Oct 2002
Dec 2002 Time slot archiving: Dec 2002
Archive complete Years 2002 & 2003
Jan 2003
Only complete fiscal years
Feb 2003 Exclude Nov 2002 & Mar 2002
Mar 2003 Protect archive areas

Pros: Complex time selection options


Cons: Limited to time selections only

New Data Dec 2002 Mar 2003


loads
© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 34
SAP BW data archiving technology : Write process

Field Data selection archiving is a very powerful tool for free


selection of transactional archive data based on any criteria:
ODS Object
Current Date: Mar 2003
comp 10 Oct 2002 Archive file

comp 10 Nov 2002 comp 10 Oct 2002


comp 20 Nov 2002 comp 10 Nov 2002
comp 10 Dec 2002 Data selection archiving: comp 10 Dec 2002
comp 20 Jan 2003 Archive Years 2002 & 2003
Company 10 only
comp 20 Feb 2003
comp 20 Mar 2003

Pros: Flexible data selection


Cons:Time selections more complex
No protected archiving areas

New Data Comp 10 Dec 2002 Comp 20 Mar 2003


loads
© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 35
SAP BW data archiving technology: Extraction/Reloading

BW Extractors and Export


DataSources are ‘Archive enabled’
InfoCube
ODS object InfoPackage is extended by an
option 'archive selection'
Selection options for available
Update Rules archive sessions and files
Only full extraction supported
InfoPackage Archive files are scanned with
selection criteria of the request

Export DataSource
Reload to original DataTarget is
possible but not recommended
Reload Recommendation:
InfoCube Archive Files
Extract to a copy of original
DataTarget instead
ODS object Use MultiProvider to combine
remaining data with reloaded data

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 36
Why would we consider Near-line storage?

Reasons for Direct Query Access to Archive Data

Business-driven reasons
Introducing new characteristics with a historical background
Strategic Analysis of data over long periods
Just because it’s our data! ☺

Legal reasons
Regulatory and industry specific requirements
Data is immediate accessible from a legal and technical perspective
Example: GDPdU demands a ten-year period for retention for tax
relevant data and…
the data’s immediate legibility and machine evaluation within the entire
period for retention (§§ 146 u. 147 AO)
Example: Food & Drug Administration (FDA)

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 37
Near-line Storage Solution for SAP BW (Key points)

Separating frequently used (kept in database) and infrequently used data


(stored in Near-line Storage)

Supporting both InfoCubes and ODS Objects

Transparent access to "non-archived" and "archived" data for queries

Hierarchical Storage Management (depending on the provider)


First Level: BW Database
Second Level + further Levels: Near-line Storage

Intelligent Data Access


Data Selection Analysis and Feedback
High level index in BW DB
Low level index in Near-line Storage

Openness
StorHouse® / FileTek, CBW® / PBS Software,
DiskXtender® for BW / Legato, …

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 38
Overview of SAP BW 3.X “Project” Solution

Query Query
Proprietary Interfaces
to select
MultiProvider 3rd Party Providers
No Intelligence


Virtual InfoCube
Basis Cube or
ODS Object Copy Near-line Storage
Adapter

Near-line Storage Partner Solution

BW Database Optical
Robotic Jukeboxes
Tape
Libraries
High speed disk NAS or
Low-Cost Disk

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 39
SAP BW 3.X “Project” Solution – Sample screen shot

3rd Party
Options

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 40
SAP BW 4.x Open Solution (Planned)
Data Flow
Query Control Flow

InfoCube/ODS Object with Near-line Services

Union
High level index
Split and Dispatch

Archive/
BW DB Interface Restore Near-line Storage Adapter

BW Data Base Near-line Storage Partner Solution

Optical
Robotic Jukeboxes
Tape
Libraries
NAS or
High speed disk Low level index Low-Cost Disk
Data Manager
© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 41
SAP BW Near-line Storage technology

A data view of BW and Near-line Storage solution integration


ODS Object
Relocate complete Near-line DB table
Sep 2002 Year 2002
Oct 2002 Sep 2002
Nov 2002 Oct 2002
Dec 2002 Nov 2002
Jan 2003 FTP Dec 2002
Feb 2003
ODS Object
Mar 2003

Multiprovider - Provides consistent view of data


Dec 2002

New Data Mar 2003


loads
BW Queries
© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 42
Agenda

Information Lifecycle Management

Information Lifecycle Management

Data Aging Strategies in a


Heterogeneous Application Environment
SAP ERP Components
SAP Business Information
Warehouse

Summary

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 43
What we plan to deliver in the future …

Operational Data Store


Non-SAP
Source

Persistent Inform-
Staging ation
Area Arch- Access
Data itected
SAP Warehouse Data
Source Marts

New!
Archiving & Near-Line Storage

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 44
What about your Data Volume growth?

700.00
Expected size
without Archiving
600.00

Allocated DB size
500.00

400.00

Allocated DB content
300.00

'Without' Initial
Archiving Archiving With regular archiving
200.00

100.00
DB growth: Reduction: DB growth: ~7 GB/month
~15 GB/month ~60GB
0.00
01.09.2002

01.11.2002

01.01.2003

01.04.2003

01.07.2003

01.09.2003
01.10.2002

01.12.2002

01.02.2003

01.03.2003

01.08.2003
01.06.2002

01.07.2002

01.08.2002
01.03.2002

01.04.2002

01.05.2002

01.05.2003

01.06.2003
© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 45
The right time to start ....

1 The “healthy“ System


Data Archiving shouldn’t be the
final step to prevent a system

2
going into cardiac arrest!

Early Planning
Proactively maintaining and
sustaining performance in
the system

3 Interdisciplinary Process
Data Archiving makes a high level of
coordination necessary between IT (technical)
and Application (functional) groups
© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 46
Further Information

Public Web:
www.sap.com Services Rapid ROI and Low TCO
SAP Customer Services Network: www.sap.com/services/
Consulting Contact
Roy Wood, VP SAP NetWeaver Consulting Practice (R.wood@sap.com)

SAP BW 3.x Near-line Pilots contacts


Americas – Tony Andreacchio (Anthony.andreacchio@SAP.com)
EMEA – Rainer Uhle (Rainer.uhle@sap.com)
Related SAP Education Training Opportunities
http://www.sap.com/usa/education/
BW360, SAP BW Performance and Administration

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 47
Questions?

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 48
Feedback

Please complete your session evaluation


and drop it in the box on your way out.

© SAP AG 2003, ASUG BITI Dallas 2003 Ron Silberstein / Mike Eacrett/ 49

Potrebbero piacerti anche