Dropping Soon: IBM’s Big Application Modernization Redbook

February 4, 2014 Alex Woodie

Last year in Rochester, before the Polar Vortex swooped down and turned vast stretches of the Upper Midwest into arctic tundra, IBM convened a meeting of the minds like no other. Big Blue flew in dozens of IBM i modernization experts from all over the world to work on a forthcoming Redbook, with the idea of putting down on paper a complete guide for how to remake existing IBM i apps into sleek and powerful programs that will run businesses for decades to come.

That Redbook is almost complete. According to sources in the IBM i community, we’ll soon get to see what the modernization experts came up with during those warmer Minnesota days. “I have been looking forward to this set of material for a long time,” IBM i chief architect Steve Will wrote in his blog last week, “and I am excited that it’s almost ‘ready for prime-time’ because it’s something customers need.”

While hanging Web and mobile interfaces off RPG programs is currently all the rage, there seems to be a growing recognition that a more thorough modernization from the bottom up is the proper prescription to prepare for long term success of IBM i customers, IBM i ISVs, and IBM itself. As Will noted in his blog, there are many shops that have not touched their database in decades.

“Despite the relational database capabilities that have been a part of this operating system for more than 25 years, they are still running their businesses using older flat-file or hierarchical designs,” Will writes in his blog. “Now that they are growing, and trying to interoperate with more modern technologies such as Web services, they are hitting limitations that will hamper their ability to handle that growth.”

That Will is speaking about this topic is good news for folks like Marinus van Sandwyk, the managing director of TEMBO Tech Labs, who for the past couple of years has been trying to gain some traction in the North American market with his DB2/400 (DB2 for i) database modernization tool. van Sandwyk is adamant that, by moving databases from the legacy DDS architecture to a modern SQL-based architecture, businesses can retain the value in their IBM i applications, speed application performance, and minimize the pain of future upgrades and modifications. TEMBO is now represented in the Americas by inFORM Decisions.

It’s also good news for other proponents of SQL on IBM i, such as Fresche Legacy, which last year acquired Databorough and its flagship X-Analysis tool for re-engineering IBM i applications and databases. You can also count the folks at Resolution Software, makers of the Xcase for i database modernization tool, among the interested parties.

All three of these vendors support the notion that IBM i applications are best served when using the DDL-SQL database engine, and that the days of relying on the DDS engine are over. And IBM is obviously a big supporter of SQL, too. It has published many Redbooks advising IBM i shops and ISVs to adopt SQL as a standard.

There is nearly unanimous agreement on where the IBM i platform needs to be, and what IBM i shops ought to be doing: using the DDL-SQL database engine (SQE), and not the old DDS engine. That engine served its purpose when applications were written monolithically, but it is proving to be a major hindrance in the migration to the modern world of segmenting applications according to the precepts of the model view controller (MVC) framework.

The problem, of course, is how to get there. (The devil is always in the details.) Is a big bang migration best, or should it be undertaken in piece meal over time? Should existing RPG applications be entirely recoded and re-engineered to use SQE and the SQL precepts–including turning much RPG business logic into database routines? Or are customers best served by employing an abstraction layer to trick existing RPG apps into thinking they’re talking to the old DDS engine when in fact they’re talking SQL? What’s the best way for managing a big database modernization project, and how can you sell it to the boss? What role do new technologies, like RPG Open Access, play in all this?

There are a lot of questions that come up when one undertakes a big IT project, like a database migration, and they all need to be studied and understood thoroughly. While the GUIs and mobile apps may get more attention from the employees and other users, the CEOs and business owners understand that there is a lot more at play when the guts of an IBM i application are opened up and tinkered with.

There are potentially big rewards for IBM i shops that successfully modernize their applications’ database around SQL: simpler maintenance of code, ease of finding technical talent, a clearer modernization path, and faster performance (primarily for random data access). But there are also risks, which is why the upcoming Redbook is so highly anticipated.

While it sounds bad, doing nothing is also an option. But there’s already been a whole lot of “doing nothing” done up to this point in the IBM i community. IBM itself has been a big enabler of the do nothing route, as it seeks to avoid rocking the boat and upsetting customers.

“In my recent customer visits, I have been reminded once again what a ‘double-edged sword’ we have created as we’ve allowed customers to remain on older technology without forcing them to modernize,” Will writes in his blog.

Let’s hope that in the upcoming Redbook, that IBM gives us clear bearings on which direction IBM i shops ought to go–and gives them the proper incentives to reach their destination. Although the Rebook was originally expected to be available in January, it is now being predicted for a March arrival.