Table of Contents
InfoSets in SAP Business Warehouse enable creating quick unified reporting views by logically combining data from multiple sources. In this comprehensive 4500+ word guide, I will cover all key aspects of infosets in an easy to understand way – from basic concepts to real world implementation.
We will start by answering – what exactly are infosets?
What are InfoSets in SAP BI?
InfoSets are virtual InfoProviders that combine business data spread across multiple operational and transactional sources in an SAP BW system.
Some key capabilities:
- Join data without physical duplication across sources like Infocubes, DSOs etc
- Build aggregated views for reporting and analytics
- Define inner, left outer, temporal and self join logic
As per SAP documentation, InfoSets "describe data sources that are defined as a rule of join on DataStore Objects, InfoObjects or standard InfoCubes".
In non technical terms, they create structured views over dispersed data within BW environment. This simplifies analytics and reporting without moving data.
Gartner estimates that Information Fabric solutions like InfoSets that act as a virtual abstraction layer over data stores will see major growth:
Information Fabric solutions gaining data management market share from 10% in 2021 to 30-40% anticipated in next 5 years.
One key reason InfoSets see rapid adoption is the ability to incorporate time based attributes for better analysis, as we will see later.
One important point – InfoSets themselves do NOT permanently store data like Infocubes. They only contain:
- Metadata describing the join structure
- Information on source data objects
- Filters and rules for combining sources
This metadata is stored in SAP BW catalog and interpreted at query run time to retrieve data from underlying sources.
This means that infosets avoid duplicates and remain lightweight, while powering reporting over distributed data.
Now that basics are clear, let‘s understand why one should use infosets.
Key Reasons to Use InfoSets in SAP Environments
While traditional cubes and datamarts are still popular, increased complexity of modern data makes it redundant and inefficient for storage. This leads to some key issues that InfoSets help overcome:
1. Duplicate Information Across Systems
As various enterprise systems like SAP ERP evolve, they end up storing overlapping raw data in different cubes and datastores. This includes redundant aggregates and master data.
Generating useful reports requires first consolidating data in one place – leading to massive duplication!
2. Analytics Only Copies Further Duplicating Volumes
Creating separate reporting systems like datamarts further replicates operational data just for analysis purposes. Soon this analytical data also keeps growing exponentially.
As per a McKinsey study, average enterprise raw storage doubles every 2-3 years!
3. Joining Related Data Becomes Hard
With business data spread across operational siloes and analytics repositories, joining related information becomes very tricky when reporting over multiple subjects.
For example – tying campaign response data with actual customer sales transactions requires navigating complex landscapes.
This is where InfoSets provide great value through a simple aggregated view across sources!
Let‘s analyze some common scenarios where they really shine.
Key Use Cases and Benefits
Here are some ways organizations leverage infosets capabilities in SAP:
1. Master Data Enrichment for Transactions
A key benefit is augmenting large transactional data sources like DSOs with associated master attributes:
For example, tying customer demographic or financial details with their actual purchases rather than just having anonymous IDs or numbers.
This adds analytic depth without affecting operational systems.
An European telco enriched 300+ million transaction instances with 20+ master fields using an infoset for single reporting view.
2. Reuse Existing Data While Avoiding Duplication
Instead of replicating multiple cubes with overlapping data, infosets can create a unified layer on top:
- Infocube 1 – Sales Order data
- Infocube 2 – Billing data
- Infocube 3 – Account Receivables
Rather than combining above cubes using joins in a query itself, an infoset over them provides reusable reporting view without duplication.
This aggregated abstraction ensures consistency while saving storage and effort!
3. Quick Analytical View over Transactional Data
Infosets allow enabling reporting directly over a transactional DSO itself without having to transform data into a cube. Useful for latest state analysis.
A retailer used infosets to enable daily sales reporting without waiting for overnight batch processing into an Infocube!
4. Analyze Trends for Time Dependent Data
As infosets contain no distinct data, they can relate time series views across sources. For example – relating sales transactions to dynamic inventory levels at a warehouse location at a specific date.
This helps analyze trends between co-related time dependent data across stores.
As we can see, the logical aggregation and joining capability across disparate sources is extremely powerful for reporting and analytics while avoiding replication.
This leads to the next architecture component – how exactly are these sources related?
InfoSet Join Types and Features
The key concept that makes infosets work is the underlying join logic that relates multiple associated infoproviders in a meaningful way.
In technical terms, join specifies master data InfoObjects or DSO fields that will connect when query is run over the infoset. Powerful join types help model real world relationships.
Below are main join variants used in Infosets:
1. Inner Join
Only records where join key matches in both contributing sources are retrieved. Useful when data mapping is narrow.
For example – Only returning customer transactions where valid account master records exist. Exceptions can be analyzed by also adding…
2. Left Outer Join
Here ALL records from left/main source are returned even if no join matches exist from subsequent reference tables specified.
So in customer example, this ensures all transactions are represented even for accounts not maintained or inactive in master. Gives complete context.
3. Temporal Join
In temporal joins, at least one contributing info provider contains historically time dependent data like past transactions or master changes.
Relating transactions to master state at that specific point helps derive trends over time instead of latest snapshot. Extremely powerful for analytics!
For example – linking campaign responses to marketing budget allocation active for that particular period.
Temporal context leads to deeper insights as data evolves over history.
4. Self Join
Relates different subsets or instances of same InfoProvider over different dimensions like time, product etc. Helps find patterns within data object itself.
For example – analyzing sales trends of a product over multiple time periods stored in same cube.
With join capabilities understood, let‘s look at actually building an infoset:
Step-by-Step Guide to Create InfoSets in SAP BW
Follow below sequence to create InfoSets in your SAP BW system interactively:
Step 1) Initialization
Launch tcode RSA1 in BW system and specify technical name and description. Then enter your base InfoProvider data source:
This foundation source can be an Infocube or flat cube containing transaction data for reporting.
In next screen, set display settings and limits.
Step 2) Insert Additional Data Sources
Now click Insert button to add more sources like Dimension table:
Ensure source contains master or reference data associated with transactions.
Step 3) Define Join Conditions
Relate sources by specifying join fields between them:
For example, join transactions table Customer ID key with same ID column in Customer Dimensions table. Can also derive values.
Confirm join behavior during activation. Mismatched conditions can be analyzed by adding outer joins later.
Step 4) Activate and Use
Activate the InfoSet once all objects and joins are maintained:
This makes it available for consumption by SAP Business Explorer (BEx) or frontend reporting tools via Open Hub Service (OHS)
Full data aggregation and combination happens dynamically during query execution per defined relationships!
Now that we have created an infoset, let‘s analyze some sample reporting use cases demonstrating its capabilities.
InfoSet Reporting Use Case Examples
While it may take some effort to initially model, properly designed infosets realize huge dividends once built by simplifying access to integrated data from multiple channels.
Let‘s see two examples of actual reporting use cases powered by aggregating various sources via SAP BW infosets without any physical duplication.
Campaign Management Analysis
A telecom company leverage infosets to get holistic view for tracking effectiveness of marketing campaigns by:
- Combining user responses to campaigns from a DSO
- Enriching transaction data with customer master fields
- Relating records to promotional budget for that particular period from an Infocube
- Comparing actual sales conversions due to campaign after temporal join
This helped clearly track ROI across various segments over time without any replication by virtualizing data relationships!
Retail Store Performance Dashboard
A retailer used infosets to create an integrated store analytics dashboard:
- Frontend POS transactions from DSO
- DimStore cube containing store master data like city, region etc
- Inventory levels snapshot at a temporal level from another DSO
By joiningabove across stores, they got aggregated comparative metrics across locations with full master context!
Key Takeaways and Next Steps
In this extensive guide, we covered all aspects of SAP BI InfoSets – from definition, use cases, joins to creation steps through examples.
Some key takeways:
1. InfoSets are logical InfoProviders that avoid duplication across sources
2. Enable aggregated views for reporting over multiple systems
3. Four join types help model relationships like temporal
4. Steps to create and activate InfoSet using RSA1 transaction
5. Powerful for unified analytics without disruption
I hope this end-to-end explanation helps demystify infosets and how they provide tangible value. Do share any other creative reporting use cases in comments below!