Which do you want—centrally managed, secured, and governed data or flexibility for the many different users of the data? With SAP Data Warehouse Cloud, you can have both.
Businesses today face a dilemma when it comes to implementing a data warehousing solution. The enterprise as a whole, needs to ensure data governance, quality, completeness and security of data. But smaller groups within the organization—departments, working teams, lines of business—need maximum agility. They have their own datasets, which only partly overlap with the company’s central repository. This often includes their own specialized marketing data, as well as data from social channels and other sources. And they have their own analytical needs, which requires developing their own reports, their own models, etc.
A large enterprise typically has in place a traditional enterprise data warehouse (EDW), which will provide enterprise-grade services such as model consolidation, data lifecycle management, and data governance. In such an environment, IT cannot give access to the central data infrastructure. This leaves functional groups no choice but to start building their own data marts and data warehouses—over which IT has limited in no say where governance is concerned, meaning they can no longer assure data quality, completeness, integrity, security, etc.
Of course, these groups within the company also have the cloud option. They can
subscribe to cloud services to get that kind of agility. But these services are typically not enterprise-ready. They work best for smaller scenarios. They allow the group to do whatever they want, with maximum agility, but they have lost the connection to the central governed data model.
In either scenario, you run the risk of having multiple data sets in the organization. Data is not updated. Divergence increases. Multiple different answers to the same questions. It’s chaos.
Introducing Spaces
SAP Data Warehouse Cloud turns that model upside down with its new Spaces concept. Spaces makes it possible to let many more users access the data warehouse and in different ways, without taking the kinds of risks described above. Using spaces, IT can give each group their own access and their own space in the same cloud environment. Each group gets the memory and computer power it needs. Each space is isolated, but still tied to the central repository, models, security, etc.
The Spaces concept unites the conflicting needs of the business for agile service with the needs of IT to provide centrally governed service. With their own space, the team is able to make their own building blocks. They can acquire data, build an analytics model—their own, distinct model—and build their own reports. Say you have a central analytics model, provided by IT or a central analytics group. IT can expose this model for read, allowing a user from one of the other spaces to access the model without modifying it.
That user can then begin building their own model and join it with the model provided by IT—creating a new joined model. Now in the data warehouse space we have a data map between a centrally governed analytics model and a separate departmental model running its own data warehouse or data mart. This is maximum agility while maintaining central data governance.
Moving to the Cloud
Suppose you have an on-premise data warehouse with excellent data and a robust data model. This was the case for one of our largest customers. They have a HANA-based data warehouse with more than six terabytes of data. They have made a
significant investment in generating cleansed, consolidated data that’s always up to date. It’s a great example of everything an enterprise data warehouse should be.
Within this company there were many functional groups and individuals who wanted access that data. But resources were limited, and they could not provide access to everyone who needed it.
For this customer, SAP Data Warehouse Cloud is the perfect solution. Say one of the marketing groups, for example, needs access, but not to the full 6 TB of the data warehouse. They are looking at a much smaller slice of the data—about 20 gigabytes. Now they can implement a cloud data warehouse, subscribing to the existing central warehouse, just for that smaller slice of data they need to work with. Basically, the system replicates the 20 GB of data, and publishes the model to their space in the cloud. And because we are talking about replicating data from SAP HANA, it is a very quick and efficient process.
Now the marketing team can do whatever they want to with that data. They can build whatever analytical models they want to on it. They can add whatever social media data they need. And their data from the central repository will continue to be updated.
The Spaces Advantage
So far this all sounds good, but some might wonder whether you really need the Spaces concept to address this situation. Why so much concern about a single data mart? But, of course, this is not the only functional group within the company—not by a long shot. Soon another group from marketing needs to do something similar, only with their with 10GB slice of the data warehouse; then a group from sales needs its own slice; then a product team wants in; then another product team speaks up; then more groups from marketing and sales say they need something, too. Each needs a somewhat different subset of what’s in the central repository. Each has their own data, too, and their own reporting and analytics needs. And to make matters more complex, some need to publish and subscribe to each other’s data.
Now the
power of the Spaces concept becomes clear. It gives you visibility to manage this complex set of needs. The users in each of the various groups are getting all the flexibility they need, but the integrity of the central repository and model is still preserved. If each of those team’s environments was a separate, one-off environment, chaos would very likely ensue.
With the Spaces concept we can allow each team full access to a centrally governed data warehouse, which is normally not an option. The team has full ownership, but still with central governance, central data quality, security. The age-old dilemma between agile and centrally governed data is solved at last.
