Data Access in the SOA Context

Preferred SOA Approach

Click graphic for a full-sized image.

If you don’t plan ahead for data access in your SOA initiatives, it’s unlikely that your services will be able to meet the necessary flexibility, performance, and scalability requirements.

In addition, if you don’t decouple the data access logic from the business logic, the reuse and agility benefits you receive from SOA quickly erode.

Abstracting the data access out into a data tier gives you one point of change vs. having to update every service every time the physical data source is changed.

Poor choices in data access middleware that underlie your services can limit both performance and scalability. It is this middleware that enables services to access heterogeneous data sources in the most efficient, flexible way possible. For example, using Hibernate to abstract data will suffer performance penalties if the underlying JDBC driver is not fast and scalable. Or a slow ADO.NET provider used as the foundation for a data layer leveraging the Microsoft Data Access Application Block (DAAB) will inevitably slow down the services.

Data access pitfalls in the SOA context include:

  • Performance and scalability issues
  • Database platform and version differences
  • Differences in how the various data sources handle standard data access operations such as create, read, search, update, and delete
  • Network performance issues
  • Varying data source security priorities
  • Differences in how SQL statements are processed
  • Exception handling and reporting

Data access has long been a limiting factor in scalability, performance, and interoperability, but SOA magnifies this problem because organizations are more likely to reuse the underlying code.

What to Look for in Data Access Middleware for SOA

Choosing best-of-breed data access components is an important enabler of SOA and the first step to resolving these data-related issues. Even if you’re just getting started with SOA, you should use the very best data access middleware to avoid problems down the road.

Furthermore, choosing the right data access middleware will improve the overall performance, scalability, and flexibility of the system. To that end, your middleware should:

“To meet our business objectives and satisfy the evolving needs of customers we wanted to partner with an established player in the .NET community. With its robust product offering, proven .NET expertise and superior support organization, DataDirect Technologies was an excellent choice."
Pervasive Software

  • Include capabilities for boosting query performance
    such as connection pooling
  • Support tunable data access performance such as adjusting network packet size
  • Be thread safe and offer client load balancing
    and failover to alternate servers for maximum
    scalability and availability
  • Support different types and versions of databases
  • Address the subtle variations in SQL versions supported by different databases
  • Support multiple computing platforms, chipsets, and operating systems
  • Obviate the need for database client libraries and software to eliminate the overhead and maintenance issues of off-the-shelf database drivers
  • Support the full range of standards, including JDBC, ODBC, ADO.NET and over time, SDO
  • Support secure communications that covers both network security and database security

The DataDirect Connect family of database drivers has several differentiating characteristics for SOA enablement that put them in a different category from other alternatives, including those drivers that the database vendors provide. In particular, the wire protocol design of DataDirect’s drivers eliminates the need for database client software and libraries, simplifying installation and administration as well as improving performance.

Customers leverage DataDirect’s data access products to create an expanded data tier that is a critical enabler of SOA. DataDirect drivers deliver on the key value propositions of SOA with performance, scalability, interoperability, and the ability to deal with multiple versions of a single database. DataDirect Connect resolves database version issues, platform and operating system differences, and also provides for SQL leveling, which addresses the differences among SQL versions.

To learn more about data access in the SOA context, read the white paper from ZapThink, “Service-Oriented Data Access: Building Interoperable, Robust & Reusable Data Services.”

If you would like to speak to an expert about how DataDirect products can meet your SOA needs, please call 1-800-876-3101 or submit a Request for Information form and an account executive will contact you.