For many IT shops, the new year will bring a new look at some of the mainframe legacy applications that have chugged along through many years. Their steadfast years of service will be weighed against the value of shiny new Web applications, with organizations' business leaders asking why they ''can't turn off that old mainframe.''
Today, choosing which applications to transform is still something of a balancing act.
''You have to go where the savings are,'' said Dale Vecchio, analyst, Gartner. ''Look at [legacy applications] where you are consuming a large part of your budget.''
Vecchio said the conversion from a legacy platform can be expensive, so the work to convert the application must be ''paid for'' with substantial savings, not marginal gains. Implementers should recall that the apps that take up a big portion of budget have the most impact, Vecchio told an audience at last month's Gartner AADI Summit 2011 in Las Vegas, Nev.
''T-I-M-E'' method applied to legacy study
When you take a close look at the application and its role in a company's present application portfolio, you find there are various paths available. Analysis may lead you to decide to leave the application as it is, to invest in an update, to move it to another platform or to just turn it off.
Ultimately, the alternative paths for legacy transformation can be encapsulated in the handy ''T-I-M-E'' mnemonic device that Vecchio describes – with ''T-I-M-E'' standing
As with other long-term programs, overall application modernization programs must show success with their first projects. So, said Vecchio, while the legacy application you first choose to address should have significant budget impact, it should at the same time not be the most complex application in your portfolio.
You should remember that the most monolithic, most spaghetti code filled legacy application may be hardest to successfully change. The transformation might fail, putting an entire application replacement program under peril.
Stakeholders' roles in business transformation
Business-side people must be lobbied for support, in order to ensure successful application transformation programs, said James Duggan, analyst, Gartner, who joined Vecchio in the AADI Summit 2011 session on ''Application Overhaul.'' Besides urging, the business-side people may need some education, since they have become increasingly accustomed to Web applications and related open source software.
''We have to get the stakeholders involved. Part of the reason they are interested in new stuff is they think everything now is free,'' said Duggan. ''You have to build a process for the stakeholders.''
Gartner's Duggan and Vecchio suggest development leaders consider pace layering approaches to portfolio overhaul. These methods admit that different application types should change at different rates, or paces.
''If I use pace layering,'' said Duggan, ''I can say, 'Let's plan on replacing [an existing application platform] with something else,''' adding that pace and layering need not merely apply to ''the old stuff.''
The pace for mainframe application change is "tens of years,'' said Duggan. ''For client-server, it is about ten years.''
''For Web apps the average life is two or three years,'' he said, referring to the popular new platform type. Continuous planning, as well as continuous portfolio refreshing is in order, the Gartner analysts suggested.
What are your thoughts? Email us and let us know.
This was first published in January 2012