Professional Documents
Culture Documents
Agenda
3. Data Flow Design 4. Administration & Monitoring 5. Maintaining Data Security 6. Performance Management
5.1. 5.2. SAP NetWeaver BI Accelerator Other Performance Techniques
Provide each organizational unit or better each role with the needed reliable, consolidated, integrated, up-to-date, and historical information
e.g., at business unit level provide a: local/subsidary view regional view global view at headquarter level across business units provide regional view global view
EDW
Ad Hoc
BI App
Planning Layout
Enterprise Report
MS Excel
Web Analyzer
Report Designer
Analyzer
Web Services
BI Layer Query Designer VirtualProvider Planning Modeler Analytic Engine Master Data Meta Data Repository
Data Sources
3rd-Party BI Data
BI Accelerator
Analytic Engine
Calculation Caching Security Planning Services
Near-Line Storage
Source Systems
SAP 2007 / Page 6
Monitoring / Administration
Performance Management
Performance Management
Provides all means to accelerate query performance, in particular the SAP NetWeaver BI Accelerator.
Agenda
3. Data Flow Design 4. Administration & Monitoring 5. Maintaining Data Security 6. Performance Management
5.1. 5.2. SAP NetWeaver BI Accelerator Other Performance Techniques
Overview
Usability Features
Favorites Personalization Advanced Search Complete data flow at a glance
Agenda
3. Data Flow Design 4. Administration & Monitoring 5. Maintaining Data Security 6. Performance Management
5.1. 5.2. SAP NetWeaver BI Accelerator Other Performance Techniques
Information Access
Any Source
Sources
Support of virtually all sources
DataSource
DB Connect UD Connect BI Service API File Interface Web Service
BAPI
SAP Source
File
XML
Legacy Applications
Source sytems categories: SAP vs. non SAP File vs. database Relational vs. Multidimensional DB ABAP vs. Java XML vs. Text/Binary Pull vs. Push Realtime vs. Batch
General Information Descriptions Reconciliation flag (not functional) Opening Balance (inventory) Error handling (duprecs)
InfoProvider
Transformation
InfoPackage
DataSource / PSA
InfoProvider
Transformation
Restrictions: Not optimized for mass data transfer No packaging of data Full Mode Only
SAP 2007 / Page 18
DataSource / PSA
Agenda
3. Data Flow Design 4. Administration & Monitoring 5. Maintaining Data Security 6. Performance Management
5.1. 5.2. SAP NetWeaver BI Accelerator Other Performance Techniques
DTP
DTP
InfoProvider
Transformation
Transformation
Target
Intuitive UI Unit conversion Unified transfer + update rules into all-in-one capability Integration of Open Hub Service
Semantic Groups *
* optional
SAP 2007 / Page 21
Transformation Definition
Transformation Graphical UI
Source fields
Target fields
Note: Key figures, characteristics and date fields are shown on the same level (transformation group)
SAP 2007 / Page 23
Transformation Rules
InfoProvider
Transformation
Transformation
DataSource / PSA
InfoProvider
Transformations InfoSource 1
InfoSource
Transformation directly links from a source InfoProvider (or DataSource) to a target InfoProvider An InfoSource is usually not needed New InfoSource architecture is used (flat InfoObject-based structure) Scenarios for (flexible) InfoSource A flexible InfoSource is necessary in order to use currency or unit conversion from the source DataSource Define InfoSource as an intermediate structure You can use a flexible InfoSource as a uniform source for several targets; the InfoSource can the be target from different sources (see next slide) Note: for direct InfoSources (for master data updates), there is no difference between old and new InfoSource, i.e. you can define a transformation as well as transfer rules Pre-requisite: InfoObject is defined as InfoProvider
Transformations InfoSource 2
InfoSource
Scenario: InfoSource as a uniform source for several targets and as target from different sources SAP NetWeaver Business Intelligence InfoProvider InfoProvider 2 Transformation InfoSource Transformation DataSource 1 DataSource 2 DataSource DataSource n
InfoProvider 1
InfoProvider m
Transformation Groups 1
Transformation Groups
Summarize key figures with the same characteristics assignments All key figures of one transformation are updated based on the same characteristic values If other characteristic updates are necessary for particular key figures, a new transformation is created
Transformation Groups 2
Transformation Groups
Use / Example Scenario: overview on bonus-relevant sales of all employees An employee generates a certain sales volume, which is the basis for his/her bonus The manager of the employee will be assigned 10% of the employees bonus as managers bonus relevant two transformation groups are generated (e.g. employee and manager)
Source Employee Johnson Sales Volume 1000 Manager Giles
Transformation Group 1
Employee Employee Sales Volume Bonus-relevant Sales
Transformation Group 1
Manager Employee Sales Volume*0,1 Bonus-relevant Sales
Agenda
3. Data Flow Design 4. Administration & Monitoring 5. Maintaining Data Security 6. Performance Management
5.1. 5.2. SAP NetWeaver BI Accelerator Other Performance Techniques
Primary Usage
Structure
Activation Queue
EDW layer
ODS Layer
Others
Active Data
Change Log
On request level
Staging layer esp. for large sets of data with (generally) unique key for external applications and analysis processes (APD)
via staging (DTP) via APIs, Staging into subsequent targets possible
No
Details
Option Generation of SID Values Improves query performance Queries are also possible if SID values are not generated Option Unique data records Only available if Generation of SID Values is set Activation process is optimized (only inserts, no sorting, no before image) Note: error if key already exists For (non-reporting) scenarios, write-optimized DataStores are recommended instead of standard DataStore objects with unique flag Performance Improvement Rollback Instead of rolling back in serial and in one transaction rollback now is in parallel and for each data package there is a single task
SAP 2007 / Page 32
Structure
Activation queue Used to store data to be updated in DataStore Object which has not been activated After activation the data can be deleted Technical key: Request SID, Package ID, Record number Active Data Table Same structure as the DataStore Object definition
Change Log Change history for delta mechanism from the DataStore Object into other InfoProvider Key fields: Request GUID, Package ID, Record number
Scenarios
Fast EDW inbound layer (no activation needed) For large sets of data records on detailed level (e.g. document level) wide structure is possible (16 key fields, 749 data fields) Load of new records: Every record has a new key E.g. for POS data Load & Drop: Full Upload into DataStore Object Update subsequent InfoProvider Drop DataStore Object data Continue with full uploads No update, only inserts
Details
Definition Only active data table (key: request ID, Packet No., Record No.) No change log and no activation queue Technical key is unique
Partitioned on request ID No SID generation Nevertheless: Reporting is possible (but not optimized for performance) Fully integrated in data flow: usable as data source and data target
Export into InfoProviders via request delta Can be included into MultiProvider or InfoSet
Uniqueness of Data Checkbox Do not check Uniqueness of data Performance improvement during data load Does not create/maintain unique index on semantic key
Details
Definition Only active data table Can be used as data target within APD cannot be used for transformation (upload) scenarios no loading process within BI but export into next InfoProvider is possible Reporting is possible Scenarios Used for direct input of (external) transactional data E.g. BI table for user interaction An API is available with a set of function modules (some are RFC enabled)
Agenda
3. Data Flow Design 4. Administration & Monitoring 5. Maintaining Data Security 6. Performance Management
5.1. 5.2. SAP NetWeaver BI Accelerator Other Performance Techniques
InfoCube
InfoCube
Star Schema optimized for multi-dimensional reporting Master Master Data Data Dimension Dimension Dimension Dimension
Dimension ID
InfoCube
MultiProvider
MultiProvider
Logical definition without data persistency Abstraction level for queries Can integrate the following InfoProviders via union operator InfoCube InfoObject DataStore Object VirtualProvider InfoSet Aggregation Level
InfoSet
InfoSet
Logical definition without data persistency Can integrate InfoCubes, DataStore Objects and InfoObjects join and outer join operator
InfoSet
w/o data persistency
InfoCube
DataStore Object
Agenda
3. Data Flow Design 4. Administration & Monitoring 5. Maintaining Data Security 6. Performance Management
5.1. 5.2. SAP NetWeaver BI Accelerator Other Performance Techniques
TR
Process Chain
DTP
DTP
TR DataStore Object 1
TR
DTP
DataStore Object 2
DTP
DTP
IP
IP
Source System 1
SAP 2007 / Page 43
Source System 2
With filter it is possible to load a set of data to the data target instead of the complete volume of data. Different data selections can be made via different data transfer processes for the same or for different data targets.
Prerequisite
DataSource is enabled for direct access (table ROOSOURCE) Or SAP Basis Plug-In Release 2005.1 SP8, or SAP Plug-In Release 2004.1 4.6C Support Package 13 Or if you want to have the correction before the above support packages are released, please apply note 923783.
IP DataSource (PSA) DataSource (PSA) Source System There is no error handling available for an InfoPackage. In case of invalid records, data needs to be reloaded from the source system.
DTP
Error Stack
Error DTP
Invalid records can be corrected in the error stack and updated into the data target
SAP 2007 / Page 47
Error Handling
Possibility to choose in the scheduler to... abort process when errors occur process the correct records but do not allow reporting on them process the correct records and allow reporting on them Number of wrong records which lead to a wrong request Invalid records can be written into an error stack Keys should be defined for error stack to enable the error handling of DataStore object Temporary data storage can be switched on/off for each substep of the loading process Invalid records can be updated into data targets after their correction.
Error Handling
Error Handling Once errors occur, the whole Data Package is terminated. The request is not released for reporting. Valid records are updated. After manual release of the request, data is valid for reporting. Valid records are updated and available for reporting
Error Stack 1
Error Stack Stores erroneous records Automatic checks: Existence of master data, conversion exit (restricted, e.g. Alpha) Customer-defined checks in transformation routines (see appendix for more information) Keeps the right sequence of records for consistent DataStore handling Key of error stack defines which data should be detained from the update after the erroneous data record After correction, Error-DTP updates data from error stack to data target Note: Once the request in the source object is deleted, the related data records in error stack are automatically deleted
Error Stack 2
Error Stack Key of Error Stack = Semantic Groups Subset of the key of the target object Max. 16 fields Defining which data should be detained from the update after the erroneous data record (for DataStore Object) Semantic groups bundle records with the same semantic group key into the same request see transformation chapter for more details (for DataStore Object and InfoProvider)
Temporary Data Storage Help for tracing the erroneous records and transformations Data records from different steps within the data transfer process can be stored temporarily Stores complete set of data (erroneous as well as valid records) Scenario: If the debugging mode is switched on Trace the erroneous records Trace Transformation
Settings for Temporary Data Storage Level of detail Tracing the erroneous records Tracing transformation by package Tracing transformation by record Deletion of temporary storage With request status green If request is deleted After X days
Switch on/off the temporary data storage for data loading steps
DTP Monitor Integrated in InfoProvider management screen Integrated in DTP maintenance Additional information: duration of each step Temporary storage access if activated Error Stack is displayed in DTP Monitor
Error Stack
Agenda
3. Data Flow Design 4. Administration & Monitoring 5. Maintaining Data Security 6. Performance Management
5.1. 5.2. SAP NetWeaver BI Accelerator Other Performance Techniques
Start
Activate Data in DataStore Object Load into InfoCube
Agenda
3. Data Flow Design 4. Administration & Monitoring 5. Maintaining Data Security 6. Performance Management
5.1. 5.2. SAP NetWeaver BI Accelerator Other Performance Techniques
Easy administration for complex Enterprise Data Warehouses using the BI Administration Cockpit
including context-specific
Drill-down to details Processing options Exceptions (optional)
in the areas of
Enterprise Data Warehousing Enterprise Query, Reporting and Analysis Business Planning and Analytical Services
to make administration easier and faster and thus to lower the TCO
Graphical display
Business Package BI Administration 1.0 from the Portal Content Portfolio. BI Administration Cockpit can run in a central or in a local portal. Technical Content for SAP NetWeaver BI (software component BI_CONT, release 7.0.2), The Technical Content is entirely based on SAP BW 3.x functionality not requiring BI_JAVA. SAP NetWeaver 7.0 BI technology (software component SAP_BW)
BI Platform
BI Suite
ed nd e m om ec R
a load transaction ST03 M Persistent Data Storage and Remote Access to BI Statistics Information Query Runtime Data Load Data Load
Statistics
at d
ry o
BI Statistics
Status
Statistics
Warehousing
SAP 2007 / Page 70
Detailed Runtime Statistics Data collection for various BI BI Data Objects in Data Warehousing, BI Enterprise Reporting and Planning
Platform Suite
Bu
I lti
SAP NetWeaver BI
Main enhancements
New Technical Content for new and enhanced BI Statistics New Query Runtime Statistics Process Chain and DTP Statistics BI Object Request and Process Status Technical Content for direct access and analysis on persistent data Per default, queries from the Technical Content filter on reading from persistent InfoProviders only Reading from Virtual Providers can be enabled on query level by customers Technical Content on detailed and aggregated level For Query Runtime Statistics New maintenance for statistics data collection Enabling statistics and selection of detail level for statistics
Agenda
3. Data Flow Design 4. Administration & Monitoring 5. Maintaining Data Security 6. Performance Management
5.1. 5.2. SAP NetWeaver BI Accelerator Other Performance Techniques
BI Accelerator
Analytic Engine
Calculation Caching Security Planning Services
Near-Line Storage
Source Systems
SAP 2007 / Page 74
Monitoring / Administration
Online Database
Near-Line Storage
Classic Archive
Data Warehouse
BIA Engine
NLS
InfoProvider InfoCubes DataStore-Objects Multidimensional Model High Performance Capabilities High Volume Capabilities Optimized TCO
NLS Engine
Offline Archive
BIA
NLS
Indexing Staging
Archiving
RDBMS
BI
InfoCube
RDBMS
Reporting Aspects
Business Explorer Suite (BEx) Transparent Access
No Access
BI
Adjoint InfoProvider
InfoProvider
BIA Engine
NearlineProvider
RDBMS
Dataflow Aspects
timeslices + dimensions ADK, ADK/NLS, NLS new process type in ProcessChains flexible for structural changes Archive and delete inEngine NLS one LUW write protection for removed areas in Data Store objects DAP
BI
Data Mart
EDW
DTP
DTP
PSA
Sources
SAP 2007 / Page 80
Scheduling via Process Chain Archive Type Offline, ADK only (like BW 3.x) Near-Line only Offline and Near-Line (NLS indexing Offline Archive) Selection Schema Time-Slice Archiving
relative archiving periods, delta oriented, DSO and compressed InfoCube, range protection for incoming data
Pure Request-based
for uncompressed InfoCubes
Online DB
SAP 2007 / Page 82
Archive
Flexible Selections
no support for periodic processing
Agenda
3. Data Flow Design 4. Administration & Monitoring 5. Maintaining Data Security 6. Performance Management
5.1. 5.2. SAP NetWeaver BI Accelerator Other Performance Techniques
Authorizations Levels
On characteristic level
Authorization
Authorization
Authorization
Authorization Check ok
Query results will be shown if query selection is a proper subset of the authorization
Query Selection
Authorizations
Scenario: A group of users is authorized only to specific sales organizations (e.g. Berlin and Birmingham)
A group of users is authorized only to specific sales organizations (e.g. Berlin and Birmingham)
(Berlin) (Birmingham)
Possible Values
EQ: single value BT: range of values CP: contains (simple) patterns ending with * or + (e.g. XY*)
SAP 2007 / Page 88
If you want to grant authorizations on navigational attributes, mark them in the attribute tab strip as authorization relevant.
Authorizing Hierarchies 1
On the same level like the value authorization, you can also grant authorizations on hierarchy levels. Assume youll have a sales organization as depicted.
Authorizing Hierarchies 2
Now you grant access for the complete Americas and France.
You can also use variables for flexibly and dynamically determining hierarchy nodes.
Special Authorizations
Special authorizations
* (asterisk): denotes a set of arbitrary characters + (plus): denotes exactly one character (e.g. 01.++.2005 until 10.++.2005 : allows access only the first 10 days of each month in 2005 - only available for time validity (0TCAVALID)) : (colon): allows only aggregated access to data (e.g. allows information on all sales areas only on aggregated level not on particular countries)
Key figure authorizations For key figure authorizations, you can include 0TCAKYFNM as characteristic into the authorization. Note: hierarchy authorizations are not allowed on this characteristic. Note: Once you define 0TCAKYFNM authorization-relevant, key figures are checked for every InfoProvider.
Check of Authorizations
Selection of query will be checked against the union of the authorizations Example: One authorization grants access to cost center 1000 for year 2004, a second one grants access to the same cost center for year 2005 Access to a query selection with cost center 1000 and years 2004 and 2005 will be granted
Year 2004 Year 2005 Cost Center
CC 1000 Year
Note: In the former concept of authorization objects, the query selection had to be in the intersection of the two authorization object if the authorization should be checked (i.e. the mentioned query was not authorized)
Comparison Analysis Authorizations <= SAP NetWeaver 2004 vs. SAP NetWeaver 7.0
Most important differences <=SAP NetWeaver 2004 Technical Foundation Maintenance Number of objects Navigational Attributes Hierarchy Authorizations Authorization Objects Not Changeable Afterwards 10 objects Only on global basis SAP NetWeaver 7.0 Analysis Authorization Changeable Number of InfoObjects not limited Individually
Via GUID and Equivalent to value 0TCTAUTHH authorizations Composition of Intersection of business Union (as expected) authorizations objects Per InfoObject AND Authorization Relevance Only InfoObject setting InfoCube
Migration
Migration Support
ABAP program RSEC_MIGRATION (use transaction SA38) No complete, automatic migration, but support About 80% automatic migration expected The more complex the existing authorization concept, the more manual migration work might be necessary Customer-exit variables for 0TCTAUTHH cannot be migrated; the respective hierarchy nodes must be assigned manually Intensive tests are highly recommended Singular event, not for scheduling During migration to the new authorization concept, the existing concept wont be changed
Agenda
3. Data Flow Design 4. Administration & Monitoring 5. Maintaining Data Security 6. Performance Management
5.1. 5.2. SAP NetWeaver BI Accelerator Other Performance Techniques
Increasing data volume Increasing number of information workers Additional Administration effort
Information at the speed of thought Quick and easy scalability Reduce cost of operation significantly
SAP NetWeaver BI Accelerator for high performance BI A new transparent approach to boost BI query performance
Performance speedup factor between 10 and 100 Without changing the BI user experience (transparent to users) Pre-requisite: BI in SAP NetWeaver 7.0
Queries
Queries
BI Accelerator
X
Database
Ready for high data volumes Queries that routinely involve access to many millions of records and may involve up to billions of records Examples: retail, utilities, telephone companies Challenging response time SLAs Example: service level agreements for call center operators demand short response times for good closure rates Unpredictable types of queries Far more different data sets and aggregations than traditional optimization and caching strategies can handle Excellent response times for any drill-down, slice & dice, Examples: on-demand reporting for different user groups, ad hoc analyses Minimizing costs of operation Maintenance of aggregates can be significantly reduced Reduced roll-up and change run times
Agenda
3. Data Flow Design 4. Administration & Monitoring 5. Maintaining Data Security 6. Performance Management
5.1. 5.2. SAP NetWeaver BI Accelerator Other Performance Techniques
Aggregates
Aggregates
Pre-aggregated (sub-)InfoCubes Alternative to SAP NetWeaver BI Accelerator
Analytic Engine
Month Revenue July 30 August 30
Database / Selection
Month Revenue July 30 August 30
Aggregate
Query Cache
Query Cache
Stores query results in cross-transactional application buffer Re-use of similar query results also for other users Can be actively used for performance improvement information broadcasting pre-load the cache via
Query Cache
Modeling options
MultiProvider (semantic) partitioning Line-item dimensions
Database features
Indexing Database Statistics
Compression
Compression
Move data from F to E fact table Compression usually reduces the number of records by combining records with the same key that has been loaded in separate requests When dealing with non-cumulative key figures, it is highly recommended to regularly compress (also when using SAP NetWeaver BI Accelerator) Double fact table As InfoPackages are added, F F Table fact table partitions are created Up loa Request Information d Typically small F - Table Optimised for Loading REQUEST No. Time Material Sales and Deleting E Table Fact Table Compression Optimised for Queries E - Table Typically large
SAP 2007 / Page 105
Weitergabe und Vervielfltigung dieser Publikation oder von Teilen daraus sind, zu welchem Zweck und in welcher Form auch immer, ohne die ausdrckliche schriftliche Genehmigung durch SAP AG nicht gestattet. In dieser Publikation enthaltene Informationen knnen ohne vorherige Ankndigung gendert werden. Einige von der SAP AG und deren Vertriebspartnern vertriebene Softwareprodukte knnen Softwarekomponenten umfassen, die Eigentum anderer Softwarehersteller sind. SAP, R/3, mySAP, mySAP.com, xApps, xApp, SAP NetWeaver, Duet, Business ByDesign, ByDesign, PartnerEdge und andere in diesem Dokument erwhnte SAP-Produkte und Services sowie die dazugehrigen Logos sind Marken oder eingetragene Marken der SAP AG in Deutschland und in mehreren anderen Lndern weltweit. Alle anderen in diesem Dokument erwhnten Namen von Produkten und Services sowie die damit verbundenen Firmenlogos sind Marken der jeweiligen Unternehmen. Die Angaben im Text sind unverbindlich und dienen lediglich zu Informationszwecken. Produkte knnen lnderspezifische Unterschiede aufweisen. Die in diesem Dokument enthaltenen Informationen sind Eigentum von SAP. Dieses Dokument ist eine Vorabversion und unterliegt nicht Ihrer Lizenzvereinbarung oder einer anderen Vereinbarung mit SAP. Dieses Dokument enthlt nur vorgesehene Strategien, Entwicklungen und Funktionen des SAP-Produkts und ist fr SAP nicht bindend, einen bestimmten Geschftsweg, eine Produktstrategie bzw. -entwicklung einzuschlagen. SAP bernimmt keine Verantwortung fr Fehler oder Auslassungen in diesen Materialien. SAP garantiert nicht die Richtigkeit oder Vollstndigkeit der Informationen, Texte, Grafiken, Links oder anderer in diesen Materialien enthaltenen Elemente. Diese Publikation wird ohne jegliche Gewhr, weder ausdrcklich noch stillschweigend, bereitgestellt. Dies gilt u. a., aber nicht ausschlielich, hinsichtlich der Gewhrleistung der Marktgngigkeit und der Eignung fr einen bestimmten Zweck sowie fr die Gewhrleistung der Nichtverletzung geltenden Rechts. SAP bernimmt keine Haftung fr Schden jeglicher Art, einschlielich und ohne Einschrnkung fr direkte, spezielle, indirekte oder Folgeschden im Zusammenhang mit der Verwendung dieser Unterlagen. Diese Einschrnkung gilt nicht bei Vorsatz oder grober Fahrlssigkeit. Die gesetzliche Haftung bei Personenschden oder die Produkthaftung bleibt unberhrt. Die Informationen, auf die Sie mglicherweise ber die in diesem Material enthaltenen Hotlinks zugreifen, unterliegen nicht dem Einfluss von SAP, und SAP untersttzt nicht die Nutzung von Internetseiten Dritter durch Sie und gibt keinerlei Gewhrleistungen oder Zusagen ber Internetseiten Dritter ab. Alle Rechte vorbehalten.