XAware data services

On the one hand the data integration market is consolidating; on the other, more and more vendors are appearing to focus on particular aspects of the market. One such is XAware, which is focused on the emerging “data services” market.

I guess I had better start by explaining what I mean by data services. It is essentially an analogue of web services in that you need to provision web services (or, more generally, applications) with data, especially within an SOA environment. IBM calls it “information as a service” but it is the same thing. Other major vendors such as BEA and SAP (via Business Objects) also refer to data services. However, these are big, expensive solutions while XAware (the name of both the company and the product) is open source except for the company’s Enterprise Runtime Management product, which provides SNMP support, deployment version control and scheduling, and monitoring capabilities for large organisations. Commercial licenses are also available for systems integrators.

The other elements provided by XAware include the XAware Designer, the XAware engine and connectors and adapters (which support standards such as ACORD, IFX and so on). While the Engine and connectors speak for themselves the Designer, which is an Eclipse plug-in, provides a visual drag-and-drop environment of the usual sort, along with a variety of wizards. On a more general level the way that XAware works is by using “XML views” that span your data sources and which can read, write or transfer data between these sources within the context of a distributed transaction. The aim, of course, is to hide the complexity of accessing these data sources; hence the use of views.

Although you may not previously heard of XAware the company has in fact been in business since the turn of the century so it is by no means a new start-up. Moreover, it has some big name customers such as ING, AXA and The New York Stock Exchange, as well as partnerships with Atos Origin, Unisys, IBM and others. However, it only adopted an open source model in late 2007 so it is still relatively early days in that sense.

The challenge that XAware faces is an old one: do you want a single platform that does lots of things (in this case, data services, ETL, data migrations and consolidations, and so on) or do you want specialised tools for specialised purposes? There are always arguments for and against, and companies that will choose one or the other. Going open source in this sort of environment makes sense since it enables companies to try out XAware’s solution easily and inexpensively. And, if you like what you see it makes it much easier to opt for XAware’s specialised capabilities rather than having to fall back onto a generalised platform.