So you want to roll out one of SAP’s Rapid Deployment Solutions for Analytics but you’re worried about performance, scalability, management and maintenance (who isn’t?). As a part of the scalability question, you worry about adding on additional ad-hoc analyses that all business and financial analyst KNOW will come as soon as you show the boss answers to her FIRST set of questions! After a successful go-live, can we integrate SAP Rapid Deployment Solutions with other SAP Analytical solutions? Yes! Fortunately, SAP Rapid Deployment Solutions can be easily integrated with existing data marts and analyses. You might choose to start quickly with a rapid-deployment solution (implementations in 8-12 weeks!) and then further build upon that foundation with other SAP Analytical solutions.
One way to do this is to build your analytic platform on Sybase IQ - and we have some very compelling reasons to do so:
1. Sybase IQ offers extremely fast complex ad hoc query response times
2. It partitions well to support high volumes of concurrent users
3. It offers high compression and data partitioning with excellent user security
4. It easily integrates with SAP BI and Mobility solutions.
OK, so how does it all work?
First of all, the fundamental architecture for most BI apps is pretty straightforward regardless of your technology decisions. It’s pretty much: Input ->Processing->Output
How do you overlay this with a consistent and maintainable Sybase / SAP technical infrastructure? As you probably read in my last blog (Keep Calm and Carry On - July 19, 2012), there are only four things to worry about:
1. Data Lifecycle Management
2. Business Rules Management
3. Visualization and Presentation
4. Information Delivery
You might chose to address Data Lifecycle Management with tools from one vendor, Rules Management and Presentation with tools from another vendor, and delivery (say, Mobile) with tools from third vendor. But from a budgetary perspective, you have to consider the cost of building, integrating, maintaining and supporting an architecture not specifically designed for integration. Why not connect-the-dots with a set of tools from a single vendor (let’s say SAP)?
Your primary focus is on creating value added analyses and reports in a timely manner. In designing this system, always work from the end user back to the source system. That is:
- First focus on content – Are the metrics that I want to render relevant to my business users? SAP offers numerous rapid-deployment solutions to address this need for pre-defined content
- SPEED – Do the dashboards paint fast? Do the reports run fast? Do any required batch engines run within your batch window? This is a relative concept that is defined by …you guessed it…the BOSS.
- Next, produce intuitive visualizations – Your users want one-stop-shopping. “I know the problem, what is the answer?” At a glance, are you providing the answer?
- Does your data fully support your analysis? Can your users explain to their boss (who wants it NOW, but also wants it RIGHT) where the data came from?
All of this brings us to Sybase IQ. SAP Sybase IQ is a very fast, very powerful base for cutting edge analytics. It is a column based, scalable database management system designed for high performance – just what the doctor ordered for analytics! How would I integrate IQ into my standard BI architecture?
Well, first let’s assume that you have the right content (maybe an insightful rapid-deployment solution) with some stunning visualizations (using BI 4.0 tools). You’ve grabbed the boss’s eye. Now it’s down to speed and the “four pillars” – and here’s where Sybase IQ shines. Let’s look at it by layer:
- Source Systems - Sybase as the database underlying ECC – pretty straightforward
- Transport – depending on your need for speed, you could use either SAP Business Objects Data Services for ETL – or you could use database level replication (Sybase Replication Services). Either way, this is simply configuration of your transport tool. Since IQ is fully ANSI compliant and most transports rely on ANSI SQL, this is a pretty straightforward process.
- Next you want to stage your data. You may choose to create a few tables, a departmental DW or even an Enterprise DW. Either way, Sybase IQ can easily serve as the DBMS to house and stage your analytic data
- Now to the RULES and the Data Marts – SAP / BusinessObjects / Sybase covers this is a couple of ways:
- The first place to build and maintain business rules is through your ETL jobs (Business Objects Data Services). By defining aggregation, transformation and calculation rules in the ETL tool, it is fairly easy to move data to reporting tables.
- The second place is in the semantic layer, the Universes – essentially these are logical reporting data marts. These are “rules engines” broadly defined. As an ANSI SQL compliant source, Sybase can provide the data foundation for your Universes using direct connections between the DBMS and the Universe through the Universe Builder and IDT.
- If you choose to build your semantic layer in SAP BW, still no problem for IQ. As a SQL compliant DBMS, IQ can easily feed BW.
- Finally, you connect Sybase IQ to your visualization tools. Again, this is a simple matter of ANSI Compliance. Further, since IQ uses standards based(ODBC, JDBC, OLE-DB) interfaces that are easily consumed by most visualization tools such as Dashboards, SAP BusinessObjects Web-Intelligence, Explorer and Crystal, this is not a problem.
So you see, from a BI perspective. Sybase IQ gives you not only what any other DB engine provides but much more (remember from above? user scale, speed, data volume). It is a very fast, highly integrative standards based DB engine. And the integration for SAP Rapid Deployment Solutions for Analytics and BI apps is a simple matter regardless of where you place it in the BI tool value chain.
If you'd like to learn more about SAP Rapid Deployment Solutions, look here: http://www.sap.com/solutions/analytics/applications/solutions-overview.epx
Follow SAP Rapid Deployment Solutions on Twitter: @SAPRDS