I read with interest your article on legacy integration using web services (Operation recycle), but was a little taken aback by some comments.
Len Erlikh of Relativity Technologies, for example, claims that organisations cannot expose their legacy applications as web services in a non-intrusive way.
Adegis is a specialist web services legacy integration software vendor, set up in September 2001 to address the specific requirements described in your article. It is clearly a falsehood to state that organisations can't expose legacy applications as a web service in a non-intrusive way. In fact, Adegis and several other companies including Mitem, Jacada, Seagull, WRQ, and Attachmate have been enabling customers to do this for some time.
The Adegis Bridge accesses the legacy systems via a number of non-invasive techniques. Not a single line of code on the legacy system needs to be changed, nor does any code need examining – techniques which make the process slower and more complex.
In fact, Adegis challenges its customers to allow us to prove this approach by giving us five days to present a legacy transaction as a web service.
You are right to point out that legacy systems do not necessarily require the kind of radical re-engineering advocated by ‘legacy renewal' companies such as Relativity. There are a host of different approaches to integrating legacy applications using web services standards, and our goal was to illustrate some of these approaches. Relativity argues that web services architectures are comprised of software components that perform common functions such as error reporting and transaction logging. It claims the way to eliminate processing redundancy and to ensure consistency in system behaviour is to standardise these components in a system-wide reusable utility library – hence, the company's argument for a more invasive approach.
I was encouraged to read the article on companies' disaster recovery strategies since 11 September (Disaster Zone).
Our experience shows that in the aftermath of those events, companies have placed a renewed focus on disaster recovery but in doing so, have neglected the 24 x 7 availability of all business and IT systems.
According to Synstar's most recent Pressure Point Index (PPI), which surveyed 600 IT managers across Europe, 66% of UK IT managers see IT continuity as a major issue.
Yet on average, 10.5 hours of downtime are experienced within their companies every month. At many companies, IT downtime becomes critical in less than 30 minutes. According to the results or our survey, it is mission critical within eight hours at 76%. Only 6% said their organisation could operate for 24 hours without IT.
Of course, contingency planning for a major disaster is essential for any business looking to be around in the long term. But the estimated £28 billion lost by UK businesses each year through downtime could be reduced considerably if organisations were to look at the wider picture of business availability.
Director of Business Continuity
I read Rob Buckley's article on measuring the potential return on investment (ROI) from integration projects (The cost of integration).
My company has been wrestling with the ROI of integration for quite some time. Such an ROI is built on a few key parameters that are, as of now, and you correctly stated, hard to measure. The reason is that most companies do not measure the costs effectively, as they tend to be broken up across all the processes, systems, and divisions being integrated. It is a common adage in IT that "the one who pays is usually not the one to see the pay-off". In integration projects, measuring this pay-off is difficult, if the enterprise does not know what [a lack of integration] cost before.
Having said that, there are a few areas, where there can be huge differences between one system and another in both cost of entry and total cost of ownership.
In the Information Age Business Briefing on security (On patrol), details of how to contact Integralis did not appear. More details on Integralis can be found at www.integralis.com.