Customer Reviews

Linux on the Mainframe

5 star:		(0)
4 star:		(2)
3 star:		(0)
2 star:		(1)
1 star:		(1)

 

 

"Server consolidation" is the latest buzzword for downsized IT staffs. Many believe this means reducing the number of Windows servers running on Intel hardware. "Linux on the Mainframe," (LOTM) written by experts from IBM, offer an alternative: virtualization on the IBM zSeries and S/390 mainframes. Virtualization is the process of running dozens or hundreds of operating system "images," each of which thinks it is running on dedicated hardware. LOTM explains the improvements in reliability, availability, and serviceability from implementing this sort of system.

LOTM doesn't assume the reader has mainframe experience. (It doesn't assume that much Linux experience, either.) As a result, the book provides the background to understand differences between PC and mainframe architectures and designs. LOTM frequently refers to two sample deployments (StoreCompany and ISPCompany) to show how mainframe technology can be deployed vertically (to mainly support individual applications) or horizontally (to clone similar systems). LOTM gets its point across using numerous diagrams and tables, each clearly showing the concept behind running multiple Linux servers and network devices on a single mainframe hardware platform. LOTM also offers a glossary to help mainframe newbies learn that "storage" in mainframe lingo equals "memory" or "RAM" on the PC. Those that find the first two-thirds of the book too light on detail will enjoy the last one-third, where technicalities like the PSW and mainframe registers are explained.

LOTM is a book which explains IBM's image of computing as an "on-demand utility," like electricity. This seems to be the only way to protect and maintain systems whose operators cannot defend themselves. Although LOTM at times reads like an IBM commercial, I bought into the book's thesis. It is impressive to consider an entire ISP architecture (servers, routers, etc.) running within a single mainframe. I recommend anyone trying to understand the latest server deployment trends read LOTM. While you may not plan to deploy this arrangement, it will definitely affect IT in the years ahead. Expect Microsoft and Sun to continue to compete in this space as well.

I just started reading this book and it is already clear to me that it is nothing but IBM marketing hype. All authors are hardcore IBM'ers and it shows in spades. Page after page of "oh how wonderful IBM is". Give me a break.
I intend to read some more but given the choice I would much rather have my $$ back and would like to buy an objective tome.
Save your money. I wish I had.
Roger Gordon

I recently placed this book on my Safari on-line book shelf. I have been anticipating a decent book on using linux on a mainframe for several years. Being a practitioner in the art of adding value to legacy (circa 1980) mainframe applications which IS departments are frightened of moving to more modern environments, it has struck me that there is a huge market for such a book. Unfortunately, this book doesn't address that market.

In the late `90s and early 2000s, I spent a few years in a Midwestern bank using older software on older mainframe hardware. A common enough combination I understand. This bank had many customers who wanted their data in easier to read formats and to be able to access their data on the web. Given the simple nature of the requests, I thought at the time that it would be easy to provide. There was only one fly in the ointment. The data in question was sequestered in a combination of outmoded IBM file formats and managed by batch based processes. Now a modern computer science graduate could be forgiven for assuming that you could simply log on the the mainframe and copy the relevant files, unfortunately, that isn't how it works on a mainframe. In fact, on a mainframe of this vintage, there isn't even a disk operating system a modern computer science grad would recognize. The silly solutions I came up with invariably involved FTPing files off the mainframe which the admins had created by performing a "data set extract". I was always unimpressed by the options for getting data off the mainframe or with integrating with its processes.

When this book came out I thought that someone had finally put together a description of several solutions to the sticky problems presented by such legacy hardware and software. Hopefully a description of how to access data sets with some native linux libraries provided by IBM, or even a way to access IBM COBOL programs in a way that made porting them easier. Instead, what this book does is to provide someone in management with assurances that linux on the mainframe won't cause thermonuclear warfare or add to global warming.

"Linux on the Mainframe" says it is intended for business people and IT architects. Realistically, it serves neither audience well. The financial aspects of zLinux are barely touched on and the technical treatment is far too high-level to interest an "IT architect."

What the book does offer is a good deal of historical background and a broad overview of the way Linux is architected on the IBM mainframe. The book might be useful to computer science students or application programmers who may be generally familiar with Linux but have limited technical knowledge of mainframes.

The book is liberally sprinkled with diagrams and tables and has a very complete index. That makes it useful as a reference manual. The language is simple, but dull.