Mainframe Reports that Customs Clearing Agency Needs a Mainframe

I'm expecting an important letter from China, and it's being shipped via one of the world's biggest express shipping companies. Unfortunately it's now two days late (and counting). I have been following the letter's progress online, and here's an actual record from the system:

Shanghai, China29/06/20110:33Clearing agency computer system breakdown;
temporarily unable to transmit shipment info

That reminds me of another true story. IBM mainframes offer a standard "call home" feature, and most customers use it. The system itself can contact IBM and report any urgent issues that might impact service. One day a mainframe automatically contacted IBM and (paraphrasing) reported that "It's getting warm in here. I'll keep running, but if it gets much hotter then I'll reduce CPU speed and give priority to the most important services and users. Before I do that, it'd be nice if a human investigated." An IBM technician immediately telephoned the customer and asked that they check their data center. Sure enough, a major server vendor's (not IBM's) blade server had caught fire and was burning out of control. (If there was a fire suppression system it didn't activate.) The mainframe literally saved the data center.

by Timothy Sipples June 30, 2011 in Business Continuity
Permalink | Comments (3) | TrackBack (0)

Herculean High-end Hacking

"Hercules, Hercules, Hercules, ..." -- Mama Klump

A gentle reader inquired about my Hercules configuration. He is having trouble getting IP connectivity between Linux running on Hercules and the rest of his network. I fear I won't be of much help: I only ever get Herc Linux to talk to the host system these days. (But that is more than my friend has at the moment.)

IBM is in a comfortable niche w/r/t their high end hardware. For "real developers", there is zPDT. It ain't free. In fact, it is well beyond what most FOSS volunteers can afford. So we have free (as in beer) ways of emulating a mainframe, like Hercules. The guys who shouldered the monmuental task of emnuating the mainframe with open-source did an amazing thing. And there are some things Hercules cannot do. So there is a place for zPDT. Maybe IBM is just trying to keep certain business partners comfortable. I hear that many developers are not. (are not "comfortable", that is)

Important note: You cannot legally run most IBM operating systems on Hercules. You can run them on several other, non-free alternatives.

For Love or Money?

The very thought of this class separation, with "real developers" being those who can afford a pricey hardware dongle versus those who write code voluntarily is, frankly, insulting. Hey, I get paid too, and I hope to continue. (There is my food addiction for which I need a daily fix. Pushers like Kroger and Safeway don't sell their Coke for free.) But I also do a lot in my spare time ... things that I want to do ... things that maybe zBoss doesn't want to pay time for. (Things where neither of us really see a biz case in the near enough term.)

Someone is going to chime in and tell me I can get a guest system at some IBM-funded site. Cool. I know that. My friend knows that too. He already has such service. So do I. But there are limits on that use. Neither of us are any more inclined to do 100% of our mainframe development using guest "z" access than we are to do 100% of our PC developement using cloudware. Would you do differently? The full range of options is essential, even for the non-free developments. (Ghasp! I use personal systems to help with the work I do for my boss. So it cuts both ways.)

Whining or Wise? (In other words, do I have a point?)

Yes, there is a point to this rambling. Whether through guest IDs or Hercules or zPDT or (careful! get this cleared with your boss first!) use of the company's mainframe, we need more open source. People like my friend (and me!) need more and better means of doing what we do. The first unfettered sharing of software happened in the mainframe world. So ... let's do it again! But we need more hardware, more "available to the hacker" hardware. (Maybe if they would license z/VM on Hercules? I'm just sayin.)

IBM should be shown how much they stand to gain from better hardware pricing. Supporting their circle of biz partners, sure. But don't shut out the volunteers. The computing public (such as the fine readers of this journal) must let them know. The high end hackers need more support.

People ... Linux lives. Linux on the mainframe is a major factor, growing daily. Combine it with z/VM and you have a virtualization platform extreme value, largely unknown (though documented) and largely untapped. But aside from the virtual, "Linux on z" is robust, reliable, scalable, serviceable, and a hot performer.

There will be a VM (and Linux) Workshop again this year. It's been 13 years since the last such workshop. Remember Y2K? The last VM Workshop was before that. For reasons that I won't bore you with here, in those days VM/ESA was being trounced like a two-digit year. Now, of course, the sun is shining and z/VM is popular again. It hosts Linux so exceedingly well. Therefore, the workshop is the "VM and Linux Workshop".

At this workshop, there may be a few zPDTs. (I'm not counting on it.) There may be some Hercules. There will be demonstration of Linux on z and especially of Linux on z/VM, and many applications on that. And there will be lots of free software, CDs and DVDs given out like donuts to cops. (They'll serve coffee too.) Y'all come.

-- R;

Oh ... and I did send my Herc config to my friend.



by sirsanta June 28, 2011
Permalink | Comments (2) | TrackBack (0)

Travelodge Needs a Mainframe (Updated)

The hotel chain faces fines up to £500,000.

UPDATE #1: The Arizona Department of Public Safety needs more safety: a mainframe.

UPDATE #2: Starbucks Singapore still needs a mainframe. It's Monday in Singapore, and Starbucks Singapore cards aren't working at Starbucks Singapore shops again. That's just ridiculously, embarrassingly bad. It's the payment equivalent of going to Starbucks and discovering they've run out of coffee. Memo to Starbucks Singapore: Call NCS, First Data, and/or IBM, ask them for a mainframe-hosted payment card solution, and fix this problem already. Sincerely, your caffeine-addicted (and now dwindling) customer base.

Meanwhile, over at The Coffee Bean shops in Singapore, there's a buy one get one free promotion if you use your Visa payWave card. Visa has a mainframe and uses it. Last I heard they've had a total of a very few seconds of outage (planned and unplanned combined) over the past decade plus — if you merely swiped (or waved) twice, you wouldn't have noticed.

UPDATE #3: To give you a better idea how serious this problem is, the Starbucks Singapore shop near my office has a corporate-issued (and professionally made) "Our cards aren't working" sign posted atop the counter. In other words, Starbucks cards are so unreliable the corporate office had to issue signs to its shops, similar in quality and appearance to its menu boards.

My dear Starbucks friends: do you make your coffee with portable electric tea kettles? Pick the right tool for the job. That would be a mainframe for payments. Mainframes work, and you can buy or rent one. In the meantime, if you need some help you can find me over at The Coffee Bean.

by Timothy Sipples June 24, 2011 in Business Continuity, Security
Permalink | Comments (3) | TrackBack (0)

The Register: "Reconsider the Mainframe"

The Register just published a new paper: "Reconsider the Mainframe." The authors encourage IT organizations to understand the modern mainframe's leading capabilities, and they also summarize the key organizational issues preventing proper consideration of the mainframe's role. Many of those same issues hinder quality IT service delivery generally.

There are a lot of false myths and superstitions about mainframes. I think this paper helps dispel at least some of them.

by Timothy Sipples June 22, 2011 in Blogs, Economics
Permalink | Comments (2) | TrackBack (0)

Windows Server 2008 and Itanium: No, Obviously Not Better

I had a really good laugh watching this video:

That video was uploaded to YouTube on November 29, 2009 — less than two years ago. But just barely four months later, on April 2, 2010, Microsoft announced that Windows Server 2008 R2 would be the last Windows operating system for Itanium systems.

No, Windows 2008 and Itanium are not a "better alternative to the mainframe." They aren't even alternatives at all.

The Itanium turmoil in recent months is a good reminder of one of the hallmarks of mainframe computing: durability. As IBM celebrates its 100th birthday this week, I think the company's most astonishing accomplishment is the continuing fulfillment of a promise made in 1964, with the announcement of the System/360. It was a simple but powerful promise, one that requires extraordinary focus and continuous investment to keep: never again would people have to re-write applications unless there was a business reason to do so.

Here we are, over 47 years later, and IBM has kept that promise. Some people say the mainframe is "old." No, it's Itanium that's old, because Itanium couldn't even accomplish what the System/360 did starting in the mid-1960s: preserve your investment in your valuable code, written in any language — even 24-bit assembler.

Now, that's not to say you should keep all that old code. You might improve it, extend it, renovate it, adapt it, enable it, or otherwise do something with it. But those decisions should be based insofar as possible solely on business reasons, not on the whims of vendors. Breaking code is really, really expensive. It's durability that's extremely valuable — and thoroughly modern.

So if you've got a business process that you want to automate, and if the process (or at least its steps) might be around for a while, it's a very good idea to develop on the mainframe where you can keep running as long as you like. But it's not only that. That old code isn't trapped in some time capsule. You can fold, spindle, and mutilate it as much or as little as you want. You can run that old code right alongside code written 5 minutes ago, with old and new cooperating in myriad ways. It really is a magnificent accomplishment, one that hasn't been replicated.

Happy 100th birthday, IBM. And thanks for the durability.

by Timothy Sipples June 17, 2011 in Application Development, History
Permalink | Comments (1) | TrackBack (0)

More Itanium Meltdown News: HP Sues Oracle

In the wake of Oracle's decision to cease all development of new Oracle software releases for Itanium-based servers, Hewlett-Packard has filed a lawsuit against Oracle. HP accuses Oracle of anti-competitive behavior. Oracle issued a statement which ends with this sentence: "Intel’s plans to end-of-life Itanium will be revealed in court now that HP has filed this utterly malicious and meritless lawsuit against Oracle."

I'm not a lawyer, but it's hard for me to imagine HP prevailing in a lawsuit. Even if Oracle is anti-competitive, what's the remedy? Would a court order Oracle to continue writing software for a particular platform? Which products, for how long, and would anyone want to buy products Oracle didn't want to create but which a court forced them to ship? In all the cases involving anti-competitive behavior that I recall, any legal remedies involved stopping a particular activity, disclosing information, and/or a financial settlement. None of those remedies are going to solve customer problems in this situation. Also, Oracle will no doubt point out that they continue to develop software for IBM platforms (AIX and Linux on System z), for Linux running on HP servers, and for Microsoft Windows running on HP servers. Furthermore, Oracle would argue that they were only following Red Hat and Microsoft in abandoning HP's Itanium servers. In other words, the market has spoken, and no vendor should be required to support a fading UNIX platform longer than they wish — and, besides, Oracle has done more to sustain Itanium than anyone. Moreover, history is littered with fading UNIX platforms for which vendors no longer develop new software: Silicon Graphics, NCR, SCO, NeXT, Data General, Wang, etc. HP has discontinued development and/or support for many operating systems, too. Are any Tru64 or MPE/iX customers happy with HP?

There were a few HP customers that were lobbying Oracle to reverse or at least modify its March announcement. The odds of that happening were always low, but the lawsuit (and Oracle's response) have dropped the odds to zero. Oracle isn't coming back to Itanium.

I suspect HP knows they cannot prevail in court, but sometimes it still makes business sense to file. As I mentioned previously, most Oracle-HP customers, now that they are forced to choose sides, will choose Oracle. Software trumps hardware, quite simply. However, some percentage of customers find alternatives to Oracle. HP probably hopes that this lawsuit will encourage more of those customers to side with HP. We'll see.

For prior analysis of the Itanium Meltdown, including recommended solutions and strategies for Oracle-HP customers, please see here, here, and here.

by Timothy Sipples June 15, 2011 in Current Affairs, Systems Technology
Permalink | Comments (9) | TrackBack (0)

A Century for IBM, Half a Century for Bob

IBM celebrates its official centennial this week on Thursday, June 16. By sheer coincidence — or with impressive planning! — Robert Neidig celebrates 50 years of service to IBM on the same day. Bob now serves as the Program Director for the System z Project Office, focusing on mainframe sales in Asia. His first day at IBM was June 16, 1961. I have absolutely no memory of that day because I was several years away from being born, and that was also about four years before first shipments of the IBM System/360 machines. In other words, that was a long time ago.

The Mainframe Blog congratulates Bob on his 50 years of IBM service and his contributions to the IT industry. We also wish him the best in his next 50 years. And if you've seen Bob or worked with him, you know that's the least we can expect.

by Timothy Sipples June 15, 2011 in People
Permalink | Comments (0) | TrackBack (0)

About "(Blank) Needs a Mainframe"

By now Mainframe Blog readers have seen several "...needs a mainframe" posts. We try to set some trends here, and that's the whole point about these posts.

The central premise (if you'll pardon the pun) of mainframe computing is about quality. Sure, you can add, subtract, multiply, divide, and branch on a PC or an iPod. Lots of computers are Turing-complete, including mainframes. But if you have a business or government to run, and if at least some of your business processes are important, then, quite simply, you need a mainframe — and you need to use it. Otherwise, it's going to be much harder to deliver the security, reliability, and other qualities real people increasingly demand.

The information technology industry solved these quality problems a long time ago, and the solutions to those problems involved relying on the highest quality infrastructure (i.e. mainframes) combined with centrally focused, highly disciplined operations and change management (i.e. mainframe-related development and operations), end-to-end. We know that formula works. Yet there are way too many businesses, governments, and their IT organizations that have lost the plot, implementing obscenely complex, Rube Goldberg-esque application architectures to fulfill even the most common and critical business functions. Such architectures are costly, fragile, and vulnerable.

Unfortunately, as we've seen over just the past few weeks, quality is deteriorating. Major businesses are crashing and burning, hard, with security and availability crises causing major disruptions. Public "cloud computing" isn't going very far unless quality improves dramatically and quickly. Only the fit will survive: the organizations that have or adopt mainframes and actually use them for their critical business processes, end-to-end. It's really that simple: "Fit for Quality."

One technology company that distinguishes itself on quality is the world's largest technology company: Apple. Here's a 30 second video example from 1995:

Apple is a remarkable company. Apple has mastered the "it just works" segment of the consumer technology market. As technology (and life) gets ever more complicated, and as the value of time increases, more and more people value technology like Apple's. The same is true in the world's data centers. Businesses and governments want solutions that deliver secure, reliable service. Those qualities are becoming more important every day. And I think IBM ought to press home its advantages and repeat this simple phrase:

Get a mainframe... and use it!

by Timothy Sipples June 14, 2011 in Business Continuity, Security
Permalink | Comments (1) | TrackBack (0)

U.S. Airways (Now United Airlines) Needs a Mainframe

U.S. Airways, which is merging with United Airlines, had to stop flying on Friday when key applications, notably their boarding system, became unavailable. The airline claims that a power outage near their (only?) data center in Phoenix caused the outage, which in turn caused chaos at U.S. Airways counters and boarding gates despite clear skies and good weather.

If U.S. Airways had a pair of IBM mainframes — one in their primary data center, one in a second data center — if they configured them in a remote cluster (using an appropriate flavor of Geographically Dispersed Parallel Sysplex), and if they actually used those mainframes to support their most critical business processes, end-to-end, then it's extremely unlikely they would have had this problem — and certainly not for hours. That particular infrastructure formula should be familiar. Was U.S. Airways following that formula? If not, why not?

UPDATE #1: The International Monetary Fund needs a mainframe.

UPDATE #2: The United States Senate needs a mainframe.

by Timothy Sipples June 12, 2011 in Business Continuity, Security
Permalink | Comments (3) | TrackBack (0)

Birthday Party for the New Internet

Some of you already know this, or parts of it ... bear with me.

IPv6 is "the New Internet", Internet Protocol version 6. It replaces IPv4 which we have all come to know and love over the past decades. Next Wednesday will be a birthday party of sorts ... maybe more like a debut. z/VM and z/OS are ready (the stack). Are your apps ready?

The new information superhighway has been under construction for years. It is now begining to carry noticable traffic.

(For the pedantic and curious, yes, there was an IPv5, but that story is told elsewhere.)

World IPv6 Day

To celebrate, and to run a bit of a test drive, organizers will hold "World IPv6 Day" next Wednesday, 2011 June 8th. That day, many major internet sites (Google, Yahoo, etc) will activate full and preferred access to their wares via IPv6. IPv4 will still be available during the event, but users with IPv4-only will come in second. (You probably won't notice any delays. Thee point is to find out if and when and where you would.)

If you're on z/OS or z/VM or zLinux, your stack is ready. Your apps might not be ready (and most of us are not running IE on MVS), but the operating system is ready. (Talk to your app writers.) Linux, as shipped, is probably 99% ready, maybe better.

Old and New ... at the Same Time

When the Eisenhower Interstate System was under construction, the old "US Highways" routes were still fully available. Eventually, interstate was completed and today it is preferred, though many of the "US" routes are still in use. Similarly, IPv4 and IPv6 coexist at the moment. Few sites are likely to close their IPv4 access for a long time yet. (Most maybe never, for varying values of "never".) Let's all reminisce about Route 66.

Some sites ( is one) don't have IPv6 listed yet. They might have IPv6 connectivity, but they're not telling anyone. Other sites (Google, Facebook) have special DNS names for their IPv6 connections. You have to take pains to use the IPv6 connection. On IPv6 day, the new route will be right there with the old IPv4 route. You won't have to do anything special to get it.

The Big Snore ... an Analogy

This is infrastructure. Most people don't get excited about it. Most won't care. Most don't care. The majority probably should not care. In your car, if you want to go from Biloxi to Baton Rouge, you just go. Sure, the interstate is faster (now), but if you're not driving you really don't care about these details. US 90 is the scenic route. I-10 is the usual pick. So by analogy, pre-interstate compares to IPv4. Interstate compares to IPv6.

If you're an executive in charge of IT readiness, you need to wake up. If you're a programmer or an engineer, you need to pay attention (and keep your boss informed).

When you're on the internet, you use the Domain Name System to get around rather than typing numeric addresses ... most of the time. On World IPv6 Day, one thing that will happen is that the DNS will direct you to IPv6 addresses first, if possible. This virtual TomTom will direct you to IPv6 or IPv4. You'll just go. Again, most users won't see any difference. And it's just one day ... for now.

As an example, if you hit, the DNS will give you either an IPv4 address or an IPv6 address. Your web browser and other such internet software will figure out for you which kind of address it wants for which information superhighway it prefers, old or new. Web browsers like Firefox have been doing this silenly for months. (Everyone please buy the Mozilla developers a beer at SCIDS.)

Again, executives need to think about this before there is a crisis. When the infrastructure is humming along, it's easy to get lulled to sleep. Don't! And you (we!) technicians are responsible too. We need to get the facts, assess the risks, and deliver the plan to management. Don't panic. There is no emergency. But there is change.

Like Phone Numbers ... more Digital than Analogy

Internet addresses are like phone numbers. These days, most of us use our contacts (on cell phones, at least) instead of punching in the 10-digit string. Using the Internet DNS is like having an automatic contacts list.

10 digits ... remember when it was less? like 7 ... or even 5? (Maybe fewer in some early exchanges, but I'm not that much of a telephone system historian.) Originally, you only needed to dial 10 digits if you included the area code. But we ran out of room, added more area codes, and now have "overlay" area codes. Area Code 212 used to refer to all of NYC, but now is only used for Manhattan telephone exchanges, and the borough has two other area codes.

What has happened is that the Internet has run out of addresses. Like telephone numbers, the addresses are doled out in blocks. New York City did not blow past 10 million phones (all possible combinations of 7 digits) when they had to break out of the 212 area code. But they had come close to 1000 exchanges ... close enough for the engineers to take action. And so it is in this case. The original internet (IPv4) can address 4 billion computers. If there are 4 billion computers in actual working order, they are not online at the same time. The problem is that the chunks have been handed out.

IPv4 uses 32 bits. That's roughly equivalent to 9 digits (base 10) and then some. IPv6 uses 128 bits. It's something like 39 decimal digits. That's a lot of numbers. No where near a googol, but big. We should be able to slice out blocks and chunks for several years.

More to Come

Sir Santa hopes to pitch IPv6 for z/VM and zLinux at the upcoming VM Workshop in Columbus.  (This is assuming that he gets on the agenda, which hasn't happened yet.) If you're interested in a geek fest, consider yourself invited.

-- R;

by sirsanta June 4, 2011
Permalink | Comments (2) | 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.