IBM Announces the "Gameframe"

The world's press is abuzz with IBM's announcement of the "gameframe." IBM will work closely with Hoplon Infotainment to marry the Cell processor with the IBM mainframe, resulting in a single system that can realistically simulate virtual worlds for huge numbers of online gamers.

The New York Times, Associated Press, International Herald Tribune, Forbes, and numerous other media outlets have the story. One typical example is here.

What's your reaction?

by Timothy Sipples April 26, 2007 in Innovation
Permalink | Comments (8) | TrackBack (0)

Privacy Protection (Or: How to Avoid Going Out of Business)

There's a security crisis in our industry. Quite simply, businesses are failing to protect their customers' privacy, and corporate information is leaking everywhere it shouldn't be. IT architects need to think seriously about how to avoid these problems. If they don't, it's going to get a lot worse: at least as bad as the 2005 CardSystems case, for example.

I humbly suggest that businesses should follow an architectural pattern of "data recentralization." I hope people don't think this notion is a radical one. Think about the privacy problem statistically for a moment. It's much easier to protect information -- to apply reasonable and appropriate business controls -- if it isn't scattered everywhere. It costs a lot less, too, to have fewer, central data stores with authentication, authorization, audit trails, and data protection (backup, disaster recovery, etc.) There will always be tension between the need for convenient, immediate access to corporate information and strong governance. However, I don't agree that these objectives are necessarily mutually exclusive. In fact, data recentralization helps promote a "single version of the truth," something that provides tremendous business value.

So why are some so-called experts totally confused about the technology solutions to this problem? Take this recent example in the trade press, which proclaimed that, "Encrypting data on a mainframe is difficult...."

Where on earth did that myth start?  Fortunately, an anonymous commenter immediately stomped on that false rumor, calling that particular assertion "dangerous" even. The commenter added, "The world would be a lot safer if our financial transactions relied exclusively on IBM mainframes, and many businesses are rushing to do exactly that."

by Timothy Sipples April 24, 2007 in People
Permalink | Comments (0) | TrackBack (0)

What's the Point of Microsoft Host Integration Server? Still Wondering....

Charles Fitzgerald (sort of) and James Governor both replied to my original question: "Is there a good use for Microsoft Host Integration Server 2006?" Thank you both. I appreciate the dialog.

I mostly agree with James, although I already noted his reservations in my original comments about the financials. I think we both agree that, while there are software costs, two things are true: Microsoft will offer a "special deal" if pressed, and the software costs are largely irrelevant to the huge other costs organizations would face with this sort of architecture, not to mention the QoS problems.

I found Charles's reply both fascinating and non-responsive. I don't have space to pick it all apart, and any omission on my part in responding to one of Charles's points in no way suggests I agree. So let's cherry pick together....

I am most amused at the idea of mainframe guys arguing cost.

I'm an architect. Any competent enterprise architect must be familiar with mainframe technology and its business implications. Yesterday I recommended a database that installs on handheld devices. Does that make me a handheld guy?

I am even more amused by a leader of the world's largest software company, founded and managed by the world's richest individual, arguing cost.

One fundamental reason for [Microsoft] HIS is because it is always preferable to run an instruction off the mainframe as opposed to on it, given a choice....

How does a product that connects to the mainframe, demanding it perform certain units of work, move instructions off the mainframe?  MS HIS may not have a purpose as it is, but it really doesn't have a purpose without a mainframe doing something.  And if it's "always preferable," why are so many businesses doing exactly the opposite? For example, take a look at Gartner report G00138213 (March, 2006) which studied database spending and deployment trends. Here's the OS ranking for customers' RDBMS deployment target plans according to the survey: Linux (including mainframe Linux), z/OS, UNIX, Windows, Other.

Whatever their rationale, customers seem to prefer HIS. The last time I checked about 60% of mainframe installations have an instance of HIS or precursor running.

I have no way of verifying that claim. But I'm suggesting, in 2007, 59.9% of them may not need it -- or at least they have architecturally superior, lower cost options.

You made an assertion that IBM "continues to jack prices." Did you even check Gartner, Enterprise Systems Journal, or several other available expert references before you wrote that? Please share with us what Microsoft's price trends have been in the past few years. If you could point me to the price reductions for Microsoft Windows and Microsoft Office, I'd appreciate it.

However, to echo James's central point, I'd be willing to applaud even a modest software price increase if customers received value from their investments. Software costs are only part of the financial picture. What really upsets me is Microsoft's abusive licensing terms, and you didn't respond to that complaint. Why does Microsoft dictate that customers cannot use MS HIS client components to connect directly to the mainframe, even if they pay full list price for Microsoft software? That's obscene. Enterprises should not tolerate license terms that force a particular architectural pattern. I'd genuinely appreciate your efforts to eliminate that license restriction in a price-neutral or better way.

To put this in perspective, we’ve just been through the biggest computing buildout ever in the last decade with the Web, and the mainframe is nowhere to be seen.

If you're referring to public Internet-facing HTTP stacks or static Web content serving, indeed, mainframes are not typically found in those roles. (No reason why they cannot, and some are.) If you're talking about everything else that happens in back of the stacks (credit card authorizations, banking transactions, airline reservations and ticketing, package tracking, insurance claims processing, tax filing, etc., etc.) then your assertion is ridiculous on its face.

I'm still wondering what the answer to my original question is.

by Timothy Sipples April 12, 2007 in Economics
Permalink | Comments (0) | TrackBack (0)

What's the point of Microsoft Host Integration Server? A response.

I asked Charles Fitzgerald, Microsoft general manager of platform strategy, what he made of Timothy's trenchant blog post yesterday about Microsoft Host Integration Server (which, now suspiciously, seems to have a less inflammatory headline). He  was kind enough to respond, and i am sure the response will  - ahem-exercise the mainframe blog community. I have taken the liberty of posting it in full from own blog, MonkChips.

I am most amused at the idea of mainframe guys arguing cost. In the interest of transparency, Timothy might also do the math on what the customer would save if they ripped the mainframe out altogether. Not only do they not need to buy HIS, but they can save a lot more on the recurring costs of the mainframe.

One fundamental reason for HIS is because it is always preferable to run an instruction off the mainframe as opposed to on it, given a choice. The reality is the mainframe is the most expensive place in the world to run code and getting more expensive in relative terms. Welcome to the land that Moore’s Law forgot. Last time I checked (and I haven’t been paying attention for a while), the mainframe was three orders of magnitude more expensive per MIP than x86 and was falling further and further behind. You’re crazy to put any new workload on the mainframe. This is why IBM is always peddling the fad of the moment on the mainframe to see if they can hoodwink people to maintaining or even increasing the workloads on their mainframe. Run Linux and Java on the mainframe, they say, never mind the fact that these cross platform approaches should drive people to the lowest cost hardware, not the highest cost. This week they’ve probably got a sales pitch for how you can Second Life on the mainframe. Anything to keep the mainframe annuity flowing in.

IBM is welcome to disprove this by actually publishing industry standard benchmarks for the mainframe, but for some reason they have not been willing to permit any apples to apples comparisons for years.

Whatever their rationale, customers seem to prefer HIS. The last time I checked about 60% of mainframe installations have an instance of HIS or precursor running.

For all the ways IBM has pitched the “mainframe renaissance” in recent years, the reality is the mainframe base continues to shrink and IBM continues to jack prices for the customers who remain. Perhaps a mainframe blog is just what it will take to turn this around in the face of terrible economics, limited investment, a dying ecosystem and a graying talent pool that isn’t being replenished.

To put this in perspective, we’ve just been through the biggest computing buildout ever in the last decade with the Web, and the mainframe is nowhere to be seen. Even when IBM has tried to pay customers to run portions of their web sites on a mainframe, they have failed. There is a lot of inertia to the installed base, and IBM has been very active in prolonging it as much as possible, but it is a dead end.

Hats off to Charles. He admits he hasn't examined the economics lately, but is confident enough to make some assertions about platform cost. IBM has been working hard to show price slashing credentials through its mainframe charter, but evidently for some reason Charles didn't get the memo (that might be because if Google is anything to go by IBM appears to have pretty much forgotten about the initiative from a marketing perspective). There are some new faces in the space (although why the "community page" doesnt link to mainframe blogs is beyond me). Anyway - I just thought some of you might care to respond. And keep it clean please folks...

by James Governor April 11, 2007 in Economics
Permalink | Comments (10) | TrackBack (0)

Microsoft Mainframe Architecture

A customer prompted me to review Microsoft Host Integration Server 2006's architecture. They're considering MS HIS as a replacement for their Novell NetWare for SAA branch office gateways. My question was, "Do you even need a gateway?" "No" appears to be the correct answer.

There's a basic fallacy at work here. If you pretend that there's nothing new in mainframe networking for at least the past decade, then MS HIS is one of several possible options. Otherwise, does anybody know what the point of this product is? I fear that a lot of customers are wasting their money.

Let's start with the money part.  For each branch you need an X86 server running Microsoft Windows.  Then you need a Microsoft Host Integration Server 2006 server license, at $2,499 per CPU.  Add a Microsoft SQL Server 2000 license to that.  Let's suppose you can get away with SQL Server 2000 Workgroup Edition and that you spend about $1,500 for one CPU plus 10 user licenses. (Let's assume your branch office has 10 users.) That's now $4,000 per CPU in software, excluding anything else you might buy for operational management. Multiply $4,000 by 500 branches and you've got a cool $2 million, although maybe Microsoft will cut you a special deal. Except you've got to add the annual support fees, and did I mention the hardware, installation, and your own burden for managing 500 remote servers? Talk about expensive!

What do you get for all that? Well, if when your branch server crashes, every client in your branch office is dead and cannot connect. (Then should you buy two for every branch?) If you ever decide you want to get rid of the Microsoft HIS servers because you're frustrated with the diminished service quality, sorry: "Customers cannot use the client pieces of HIS to connect directly to the host." (Even if you pay for the software!) You also get to write to proprietary APIs which nobody else except Microsoft has, like FMI (a.k.a. 3270 EIS), so you'll spend time and money trying to migrate in the future.

Or you can just...connect every client directly to the mainframe! With TCP/IP! You can buy SNA-related APIs on the client to support your existing code while maintaining TCP/IP connectivity across the wire, but you only buy that if you need it. (IBM products such as the CICS Universal Client, Personal Communications, and the Communications Server for Linux on z Remote API Client are examples, and there are others, including non-IBM products.) That's leaving aside capabilities such as Web services on the mainframe, HTTP on the mainframe, Java on the mainframe, etc.  Fun historical fact: Stanford University installed the first HTTP server anywhere outside of Switzerland, and it was on a mainframe.

So that's what this banking customer is going to do, connect directly to their mainframe. They just avoided a multi-million dollar mistake.

Is there a good use for Microsoft Host Integration Server 2006? What is it?

Minor spelling error corrected April 12, 2007.

by Timothy Sipples April 10, 2007 in Economics
Permalink | Comments (4) | 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.