FAQ
Q: Isn’t it more then just translating between data structures, what about where the rubber meets the road and column names are different for the same data?
Q: Doesn’t the universal view or context still need to be dealt with as individual tables and columns?
Q: What source of data can be integrated?
Q: Can unification include non-formatted data sources like XML, text documents and emails?
Q: How long does it take to set up an SOA system using contexts?
Q: Ok, if it implements fast does it take a long time to learn?
Q: What daily tasks are involved to take advantage of all of this flexibility and business agility?
Q: Are we forced to rely on a big set of proprietary APIs for our low-level implementation and maintenance?
Q: Do we have to use a specific programming language?
Q: Does the unified approach to interacting with data consume a lot of network overhead?
Q: Should we implement an enterprise-wide governance policy?
Q: Doesn’t putting more servers in the path of data slow access to that data?
A: This can increase network performance and move data 300% faster. It is all in the use of machine horse power to better prepare data for transport across the network.
In unification there is synergy. For example: Data triggers which limit or delay the transportation of the unused portion of data requests Cashing data along the way.
Q: Isn’t it more then just translating between data structures, what about where the rubber meets the road and column names are different for the same data?
A: The translation layer automatically translates between data types. Data requests are made via their context which is a universal view of the data.
For example: You could define this view using any kind of data structure, you could even make up your own kind of data types. The system relates the semantics of data typing to a brand and implementation of a specific database system. Known translation are automatically applied while the data is in route.
Q: Doesn’t the universal view or context still need to be dealt with as individual tables and columns?
A: In some ways yes, naturalized data can conform contextually, but, a context can ignore the boundaries of tables and schema and even of database implementations.
For example: A context can treat some columns from table A and some other columns from table B as a single virtual table.
Q: What source of data can be integrated?
A: Any database management system with a Java driver is compatible. This includes the new more application specific database engines.
Q: Can unification include non-formatted data sources like XML, text documents and emails?
A: The current release is focused on universal database access. We intend to add support for non-structured data as that part of the system continues development.
Q: How long does it take to set up an SOA system using contexts?
A: It is surprisingly fast, a complex database implementation with 50 tables averaging 30 columns each can be set up in as little as 6 minutes.
Q: Ok, if it implements fast does it take a long time to learn?
A: Training on high-level implementation takes between 10 and 20 minutes.
Training on lower level aspects takes between 50 minutes to 2 hours.
Q: What daily tasks are involved to take advantage of all of this flexibility and business agility?
A: There is no requirement to maintain at the code level. System adjustments are made via a high level GUI (the same one that does the original high level implementation).
Q: Are we forced to rely on a big set of proprietary APIs for our low-level implementation and maintenance?
A: No, there are only two classes required for all data interactions. In fact, you can use databases by simply referencing the context and the server will automatically generate the SQL. Although you can still use traditional SQL against any of the unified databases.
Q: Do we have to use a specific programming language?
A: To use our client-side API, yes, the API (and the system) are in Java. You can use any language using a message formatting technique.
Q: Does the unified approach to interacting with data consume a lot of network overhead?
A: No, in fact the unified method uses far less bulk then modern methods, and only a fraction of the data transportation overhead.
Q: Should we implement an enterprise-wide governance policy?
A: All of the experts agree that governance is the key to a successful SOA. But there is a catch 22 in this concept. The experience and knowledge to implement effective control over the technology is needed before you start.
The reality is that it is a process that does evolve. One way to start is to have a working map of the data flow in the enterprise.
Implementing a smart enterprise service bus will address data integration and cause the by-product of a complete data map.
Your data can then be used on an enterprise scale and it becomes a valid foundation for SOA. Governance can then be established before the next step of dividing application components across the network.
Build a solid foundation for both the business management (governance) and technology.