RIM Needed a Mainframe

Has anybody else noticed that IT service delivery is bad and getting worse? Hardly a day has gone by when there hasn't been a big security or availability failure.

For example, Rearch in Motion (RIM), the maker of the Blackberry, certainly needs — needed — a mainframe. The company is already in trouble, rapidly losing marketshare, and now they can't even keep their messaging service running. If businesspeople cannot rely on RIM to keep their Blackberries in service, then they'll stop buying Blackberries. RIM might go out of business as a result of this disaster. How much does bankruptcy cost?

R.I.P., R.I.M.

by Timothy Sipples October 19, 2011 in Business Continuity, Current Affairs
Permalink | Comments (4) | TrackBack (0)

My Personal Mainframe: Late 2011 Edition

Back in 2006 and again in 2010 I illustrated how just about anyone, with relatively little money, can buy and run an IBM mainframe. In my case, that's to run my mission-critical home computing workloads. My only real barrier is that, like most data centers that are overstuffed with other servers, I don't have physical room in my home for even another laptop computer.

That aside, I keep an eye on the marketplace. The trends are entirely helpful in making it easier and more affordable to acquire your first mainframe, whether for your home or for other, more demanding situations. So I'd like to update all Mainframe Blog readers who share my fascination and interest in jumbo shrimp small mainframes. And, if you have some stories to share about your small mainframe in your home (or your small office), please post some comments.

First, I'd like to set some unofficial rules:

  1. My personal mainframe must run licensed z/OS. I like all mainframe operating systems, and I may run some others as well, but z/OS is my base requirement.
  2. My software licensing should not unduly restrict who can use the system (or how), nor should the licensing be artificially time-limited (like some of Oracle's licensing). IBM's Monthly License Charge (MLC) software licensing is perfectly fine, even excellent, and I'm quite confident it'll be reasonably stable (or even declining in price) given IBM's track record.
  3. On the other hand, I'm OK with software licensing that restricts what products I can run. All commercial software licensing does that.
  4. I want to be able to write and run some programs in one or a couple popular programming languages. I want at least basic transaction management, basic database capabilities, and Web user interface support. As in 2010 that means I'm getting z/OS which includes VSAM, plus the C/C++ compiler and a few other optional z/OS features, plus WebSphere Application Server for z/OS.

In other words, I'm considering full commercial software licensing. If I want to set up an account for a remote user connecting via the Internet to my machine, that's no problem with full commercial licensing. Some of IBM's special pricing, such as zNALC and Solution Edition, can meet my personal requirements. Of course, your personal mainframe requirements may be very different. Read on for information on development-only options, as a notable example.

With those rules out of the way, let's consider the changes in the marketplace since 2010 and how they affect my choices and finances. As mentioned above, the news is again all good:

  • IBM dropped the starting price for its smallest new z114 mainframe configuration under $75,000 (U.S. price). IBM's new model introduction and price reduction means there are more used models at lower prices that will meet my personal mainframe requirements. I would be quite satisfied running a z890 (model 110), z9 BC (model A01), z10 BC (model A01), or of course a z114 (model A01). In my 2010 review I said that I wouldn't buy a used z800 unless I happened to get an incredible deal. That's still true, but with all the other, newer used models available that would have to be priced well below $75,000, it's hard to imagine any used z800 would make the cut. And all of those other, newer models run all currently available z/OS software, unlike the z800 (which can't run DB2 10).
  • IBM discontinued the DS6000 line of FICON-compatible storage products, but they're available in the used market. For this 2011 update I also found this interesting option which provides about 6.5TB of total storage. That unit would also meet my personal mainframe requirements. If I grow, I have several storage options, including NFS-compatible Network Attached Storage (NAS). IBM shipped tapeless z/OS loading and installation support, and so I would skip tape for my personal mainframe.
  • IBM has made it easy (and even more affordable) to order my preferred software licensing with hardware with their zEnterprise Solution Edition for WebSphere offering. Alternatively, if I opt for a used machine, I could just order the software licenses separately, still applying for zNALC qualification. WebSphere Application Server V8 for z/OS is now available, and that version offers even better performance on every machine, including my 26 MIPS configuration, so I'm delighted with that. That's more performance per dollar. And JZOS has been enhanced to improve VSAM access, plus there are various z/OS improvements in C/C++ and UNIX System Services. All of those improvements are fantastic for my personal mainframe.

Of course, you don't have to buy a physical machine to enjoy mainframe computing. You can access cloud-based z/OS remotely, across the Internet, from several different companies. If you're interested in development-only licensing, then IBM's new Rational Developer for zEnterprise Unit Test Feature is a great choice. Information Technology Company is a one-stop shop that makes it easy to satisfy your mainframe development needs.

Now, if I can just clean out my home data center....

by Timothy Sipples October 7, 2011
Permalink | Comments (2) | TrackBack (0)

In Memoriam: Steve Jobs

Photo © 2011 Apple Incorporated

Steve Jobs passed away earlier today (U.S. time) at his home in California at the age of 56. There's so much to say about Jobs and his impact, but I would like to highlight his significant impact on mainframe computing, specifically.

I "met" Steve Jobs once, in 1988. I might have shook his hand — Steve's famous reality distortion field might have warped my memory, though. Jobs was then the CEO of NeXT, a company he started after he left Apple. (Several years later Apple bought out NeXT, and Jobs returned to lead Apple.) I was lucky to be invited to the unveiling of NeXT's first computer, the NeXT "cube," at NeXT's event in Boston aimed primarily at the area's huge academic community.

Even by then, Jobs had perfected his technology solution stage show. I was seriously impressed with the thinking behind the NeXT cube. Jobs and his NeXT colleagues saw several unsolved marketplace problems and went about trying to craft a unique computing solution to address those unmet needs. As it turned out, NeXT was Apple's incubator, and way back in 1988 Steve Jobs foreshadowed what would come — what he would cause to come to the marketplace, actually. The $6,500 NeXT cube (university discounted price, 1989 dollars) was too expensive for most individual students, but Jobs explained that each student could carry his/her whole computing experience in a (large) pocket, sharing a NeXT-equipped lab. The NeXT machine could read and write huge (for the time) optical disks which could hold the operating system and an entire undergraduate education's worth of information. Arguably that was a cloud computing — or even mainframe service bureau — concept, reborn and adjusted for the technology limitations of the time. But that idea of computing presaged the iPhone and iPad, too. NeXT also featured Display Postscript, a reflection of Steve's tireless focus on making sure his products always looked their best. That focus was later manifest in his work at Pixar and in such technologies as the iPhone 4's "retina display." But in 1988, Display Postscript, even on a grayscale display, was seriously radical stuff. And there was UNIX, which served as the NeXT's operating system. But it wasn't an ugly, hard-to-use UNIX. It was the progenitor to Apple's Mac OS X and iOS, and NeXT engineers had worked hard to make sure the software was polished and approachable. They mostly succeeded at that time, and of course they've been wildly successful after Apple adopted most of those NeXT-incubated technologies.

NeXT was born almost exactly at the same time the Web was born, at CERN. History records that a NeXT machine was the very first Web server in the world. (One of Stanford's IBM mainframes was the second — and Stanford's mainframe also served the first Web application with dynamic content.) Maybe the Web would have been born without NeXT, maybe later, maybe differently, or maybe not at all. However, the birth and explosive growth of the Web has resulted in renewed interest and growth in mainframe computing generally (i.e. "host computing," "cloud computing," etc.), and IBM mainframe computing in particular. I think NeXT was the point in his career when Steve Jobs first understood the balance between personal computing and shared, networked computing. Apple's products continued to be intensely personal, but they also led (and continue to lead) the market in terms of community-building and service delivery. For example, the iPod is a great media player, but the iTunes Music Store was the key breakthrough that made the iPod a must-have. And the iTunes Music Store sure resembles a mainframe in many ways, although it's the combination that's magic.

I've also heard a lot of reports from banks, in particular, wondering why their carefully prepared five-year mainframe transaction volume forecasts are not accurate. Answer: smartphones, and particularly the iPhone, are generating more transactions (and associated batch processing) volumes. Thank you Steve Jobs for that.

Finally, I'm quite sure Steve Jobs spurred many IBM employees and managers to get off their butts, to stop living off the past, and to innovate. IBM's near-death experience in the late 1980s and early 1990s almost coincided with Apple's. Those two companies are more alike than most people realize. I think Steve Jobs realized it, though. Apple and IBM collaborated on a number of technologies, some with success and some not. Their business models are quite similar, even though they don't compete at all. Both companies are continually trying to figure out what unmet needs (and desires) the marketplace has and then, using unique technologies and packaging, to bring real solutions to market to address those needs. They don't try to be the cheapest, but they always try to provide maximum value-for-money. So let's always keep innovating.

Thanks, Steve.

by Timothy Sipples October 5, 2011 in People
Permalink | Comments (1) | TrackBack (0)

The postings on this site are our own and don’t necessarily represent the positions, strategies or opinions of our employers.
© Copyright 2005 the respective authors of the Mainframe Weblog.