Mainframe Mindset

Whether you're moving from mainframe to NT and Windows 2000 or vice versa, this book can help you translate your new world.

• By Greg Neilson
• June 01, 2000

Windows 2000 and Mainframe Integration serves two audiences: those with a mainframe background who need to understand the world of Windows 2000 (all of those Y2K programmers need to go somewhere) and, conversely, those with a PC background who need to work with and understand IBM mainframes. Despite some of the predictions in the early '90s, mainframes certainly aren't going away; and increasingly we in IT need to integrate the two transparently for the advantages that each platform provide. Some of the latest technologies added by Microsoft to SNA Server (which is covered in the book) allow distributed objects from both platforms to invoke each other as needed. To be able to design and build these types of applications requires a good deal of understanding of the architecture of both platforms and how they can be connected.

Zack's book takes various aspects of operating system concepts--architecture, memory management, security, networking, and the like--then discusses how IBM's mainframe OS, OS/390, works, and compares and contrasts this with the way Win2K works. Those of us from one background trying to understand the other environment often find the concepts very similar. There's just a staggering array of new words and acronyms that describe the same thing-or worse, the same acronym is used for totally different things. For example, SMS in the Microsoft world, of course, stands for Systems Management Server; in the IBM mainframe world it stands for Storage Management Subsystem.

Having made a career move from mainframes to the PC world a few years ago, I see many IT professionals out there who can make use of this book to get moving quickly. There are clear and concise explanations of mainframe concepts here that had taken me years in working with the platform to really understand. The same is true of the explanations for the Win2K arena as well; this is clearly the work of someone who has been working with both platforms for a long time and understands them well.

The last chapter discusses the $64 million question of scalability; unfortunately, it's still too early to evaluate exactly how useful the Data Center version of Win2K is going to be. For mainframe folks who are used to almost continuous availability and enormous degrees of scalability, this really will be the test for whether Win2K takes a hold as an enterprise-class system or whether it's just another wannabe. The high availability and NT clustering provided now still has a long way to go until it reaches the power of a mainframe parallel sysplex, which can support up to 32 mainframe nodes. The author also cautions the reader from making too much of Microsoft's "Scalability Day," which, while an interesting demonstration, didn't prove a great deal.

If there were any area I thought needed more detail, it would be coverage of the Unix-System Services add-on to OS/390, which hardly rates much of a mention in the book; yet this important product is the basis for many Web technologies, including Domino, to run on the mainframe platform. However, I guess it was probably a battle to get the book down to the size it was with as much information as it has.

If you're making the transition from Win2K to IBM mainframes or vice versa or simply need to understand how to work with the other platform, this book would be a great start. You should be able to use this to make a running jump into your New World, although, of course, you'll need more detailed information later as you delve deeper into these technologies.