This article is part of an Essential Guide, our editor-selected collection of our best articles, videos and other content on this topic. Explore more in this guide:
1. - Approaches and tools for modernizing old apps: Read more in this section
- Top considerations before upgrading old apps
- Using a services approach to modernizing legacy apps
- Mobile computing's impact on application modernization
- Approaches and tools for app updates and architecture modernization
- Application modernization needs on the upswing
- How to tackle application portfolio assessment
- Challenges and trends modernizing apps and migrating to the cloud
Explore other sections in this guide:
- 2. - Utilizing the cloud for app modernization
- 3. - Getting mainframes and databases up to speed
- 4. - App modernization and cloud: Must-know terms
There's a big drive in my organization to get efficiencies from applications. Which approaches and tools make sense for application updates and architecture modernization today? Is cloud now the best strategy for keeping our system current?
Look out into the business and start counting all the spreadsheets that you, the architect, are using to augment the weaknesses in your current state applications architecture. Often, only one-third to a half of applications' functionality is being used to meet business needs.
What's behind this drastic waste of resources? Probably the business has built out its systems with desktop applications, some of which have become fairly elaborate. So, the business is working with a 15- to 20-year-old strategy with 50% utilization, at best, and has no strategy for doing it differently.
Some people think a new modernization strategy only requires moving applications to the cloud. They're thinking: "I'm going to fix significant underlying architectural problems by smoothing over something with a bunch of new front ends." Unfortunately, they're unlikely to make progress in this area. If it's a move to the cloud that's wanted, analyzing and decoupling and re-architecting current state systems are a must. If a business is going to make those investments, it had better have a business-driven approach to doing it.
Getting into cloud and virtualization without looking at this approach from a business-driven, holistic perspective is a fool's venture. If I was the CIO, and an IT guy came to me without a business strategy for driving application updates and architecture to the cloud, I wouldn't give him a plug nickel.
Often, only one-third to a half of applications' functionality is being used to meet business needs.
Now, it's back to those spreadsheets you're using to cover up problems. What's a different approach, if it's not cloud? What can be done to deal with the underlying architectural weaknesses, fragmentation, redundancy, inconsistency, lack of automation and other things going on in the current applications and data architecture? Without looking at that, business and IT are basically living in a very, very limited technology bubble.
In business transformation, many current systems are going to continue to run as incremental improvements -- prioritized by business needs -- to the overall architecture. By making sure that there's a business driver for each step, IT won't be fighting a battle of pleading with the business for money. It should be more of a flipside, where the business is pushing money on IT to solve business problems.
On the application modernization tools side, there have been improvements but not enough, far from enough. I alpha tested what I consider the first commercially available software reengineering tool in 1980. It was a COBOL code restructuring tool. There's still some of that technology floating around today. It's pretty rudimentary. It worked on a program at a time. During the 1990s, there was a brief continued evolution of some of those tools and technologies, but they never got over the innovation hump, largely because there was no big drive for modernization. Until recently, most tools for application updates were used for the purposes of maintenance, really.
Now, there's resurgence in business investments in transformation, coming mostly in three categories:
- Migration tools -- Some are very sophisticated. Look at Philip Newcomb and The Software Revolution Inc. You'll find a fairly sophisticated language platform transformation technology; it's very good stuff but very technologically oriented.
- Screen scraper technology -- We've taken that as far as we can go, but it's still widely used.
- Holistic business analysis and transformation tool suites -- This range of tools can be used to holistically review entire systems. For example, I'm working on a project now using Hatha Systems. This type of business-driven modernization tool set looks at entire systems and can parse and load the entire system. There's a series of off-shoot technologies included, such as cross-reference reporting and intelligence on the system or systems. There's additional information for business rule extraction. Right now, these types of tools do have some scalability issues.
Evolution isn't on the fast track here. I would have been stunned if someone told me in the '80s that 20-plus years later there would only be incremental improvements made in modernization tools. We need demand from businesses and software professionals for vendor investment migration and holistic tools. The tool vendors have the capacity and capability to build out these technologies to the next generation of modernization. That's not happening much now, as most vendors are sticking with tool-assisted modernization with manual support or utilities that do some of this work.
Businesses that reach an understanding about the benefits of application updates and architecture transformation will gain the competitive edge, and their success stories can drive the tool evolution.
About the author:
William M. Ulrich is president of TSG INC., a management consulting firm that specializes in business and IT planning and transformation strategies.
Editor's note: This expert advice is taken from an interview by Jan Stafford, executive editor of SearchSOA.com.