CIOZone.com - Professional Network for CIOs and IT Professionals

1. 03-15-2010 07:31

I DECIDED TO BLOG THIS ARTICLE. 
I HAVE POSTED MY THOUGHTS BELOW -- BUT IT IS EASIER TO READ THEM AT HTTP://BLOG.CLABBYANALYTICS.COM. 
Regards, 
Joe Clabby, President, Clabby Analtics 
 
A Counter Opinion to Your Mainframe Up, Down, In, Out Article 
By Joe Clabby, President, Clabby Analytics 
I read with great interest yet another “the mainframe is dead” article coming from the technical press — this one having been written by FORMER IBM employee Michael Neubarth. The report is entitled “The Mainframe: Up, Down, In or Out” and can be found at http://www.ciozone.com/index.php/Server-Technology-Zone/The-Mainframe-Up-Down-In-or-Outu.html.  
My overall impressions of this article are these: Mr. Neubarth is an excellent writer — and he clearly does a lot of research and talks to a lot of vendors as well as research analysts. And several of his criticisms of the IBM mainframe are valid. But I was very disturbed by what I consider to be omissions in this article. And I was particularly disturbed that he would recommend to his readers to take the writings of certain analysts “with a grain of salt”. (Perhaps in his next article he can include a list of specific analyst papers that his readers should not read — after all, why would he want his readers to view opinions that are exactly counter to his?).  
In response to this article, I’ve written the following open letter to Mr. Neubarth. 
____________________________________________________________ ________________________ 
Dear Mr. Neubarth: 
Please accept the following critique of your article. 
Let’s start with your advice on listening to or reading what certain analysts have to say. In a section that you entitle “Pundits in the Pocket”, you observe that “an examination of news reports and articles on the mainframe shows particular analysts appearing over and over, providing commentary favorable to IBM\'s position”. You go on to say “because of these close associations, the analysts\' views should be taken with grain of salt”. You also suggest that “the whitepapers and surveys that are sponsored or conducted by vendors should be viewed with a skeptical eye”. Just FYI, I’m one of the analysts whose case studies and reports are often found on mainframe sites — and I don’t take kindly to your advice to your readers to view my reports with a skeptical eye. In fact, after reading my counter points (below), I think your readers might want to view your reports with a skeptical eye.  
Before I start my critique of your article, let’s get a few things straight here: 
• I am an advocate of mainframe computing — I think mainframes are the most sophisticated and elegant commercial computer design on the planet — and I have no problem telling people why I believe this (and better yet — I can fully substantiate my opinion versus any other systems architecture in the industry). I even operate a Website to promote mainframe ideas (www.gomainframes.com). It should, therefore, come as no surprise that, given my positive impression of the mainframe and my unabashed willingness to research and write about mainframes, that IBM publishes a number of my reports and case studies. (Incidentally, I’m also a reporter for Mainframe Executive magazine — just FYI — in case you want to give your readers a heads-up to take my reports in that magazine with a grain of salt, too). 
• Several of my opinions run very counter to those contained in your article. Perhaps this is why you recommend that my research should be viewed with a skeptical eye (you wouldn’t want your readers to take my counter-opinions very seriously, would you?). 
• I am in no one’s “pocket” — just like CIOzone is in no one’s pocket. 
Now, for the fun of it, allow me to present a counter argument to your “Pundits in the Pocket” section — just so that we understand each other a little better: 
• Your article takes a decidedly negative slant on the mainframe’s future; 
• Your article has a great big, flashing Microsoft banner advertisement right next to it on all eight pages (and we both know that Microsoft has a mainframe migration program in place); 
• Your article quotes an HP executive (the one in charge of their mainframe migration program) several times — but quotes no IBM executive. 
Using your Pundits-in-the-Pocket argument, would it be fair to say that there is a monetary reward (advertising revenue) from a particular advertiser paid to your organization to underwrite the publishing of this article? And, given that Hewlett-Packard is mentioned several times in a very favorable light — does your organization intend to sell reprints of this article to Hewlett-Packard? 
Also, should I have named the above section “Press in the Pocket”? 
Omissions 
Okay — now let’s get down to some stuff that I felt was missing from your article. 
• You start by pointing out how IBM mainframe sales have dropped 27% from a year ago (this is just before your “death watch” section). By comparison, I notice that you neglected to mention that the HP Itanium-based machines in HP’s Business Critical Systems division have been performing extremely poorly over the last four quarters (-22%, -33%, -30%, -29%). That’s an average drop of 28.5% per quarter — a worse financial performance than the mainframe. (I’m curious, are you planning to write a follow-on article about the death of Itanium systems? If so, I recently wrote a piece about how Intel’s own Xeon multi-cores are going to have Itanium-bases systems for lunch — and I’d like to talk to you about it…). Also, are you planning to write a follow-on article entitled: “Itanium-based Servers: Up, Down, In or Out”? 
• You quote an HP executive as saying that there is no way that IBM is pleased with such poor mainframe results. I always love it when vendors with vested interests are allowed to make such comments about their competitors. Too bad that no IBM executive was given the opportunity to speak for him/herself anywhere in your article… 
• You are to be congratulated for your box war section. You let the Hewlett-Packard executive tell his side of the story again — about how companies are moving from mainframes to Hewlett-Packard servers. And then you respond with evidence that IBM has been very successful moving HP/Sun customers to IBM mainframes (for the most part, I liked your balance in this section). The part I think you missed is that the migrations from IBM mainframes were likely made to HP Itanium-based servers — you know, the ones that I described earlier that are declining at a greater rate than IBM mainframes — and the ones that have been obviated by Intel’s new Xeon multi-core x86 servers. If I were an HP customer who migrated from a mainframe to an Itanium-based server, I’d be more than a little concerned about the wisdom of that decision right now.  
• In your “Moving Off” section, you talk about the U.S. House of Representative’s mainframe migration —and about Union Pacific’s pending move. I think there’s more here than meets the eye. The energy savings reported by the House of Representatives was based on a comparison to a much older, less energy efficient mainframe model. Had they upgraded they may have found that the energy they used would have been significantly less than the new server farm they just installed. And on Union Pacific, they are planning to spend $150 million to $200 million to migrate away from a mainframe — and its going to take them three more years to deploy their new environment? I hope that Berkshire Hathaway (Union Pacific’s primary investor) fully understands the gamble that Union Pacific has undertaken. (Don’t get me wrong — the new x86 multi-cores are excellent machines — but Union Pacific has to also redesign its business processes, business logic, networking infrastructure, business resiliency plan, security plan, and more to make all of this work. I’m not convinced they can get where they need to from a service-level perspective given where the x86 ecosystem is today). By the way, they declined an interview by me — same thing with you? (I noticed that you didn’t talk directly with either the House or Union Pacific for your article). 
• Just FYI, I write a lot of case studies — which means I go to mainframe shops, talk to mainframe managers, and ask them about the workloads they are running. And what I’m finding is that a lot of new Linux and Java workloads are running on the mainframe — and IBM is repositioning the mainframe as a new, consolidated workload Linux server (it is a major part of their mainframe marketing effort and key to IBM future growth strategy for the mainframe). Why was this not mentioned in your article? (And note, you’ve discouraged your readers from visiting IBM Web site to read articles published by analysts like me — so how are they going to learn about this growth strategy if you fail to tell them?)  
• You raise mainframe skill sets as an issue. I’ve just written a report that claims that this skill set issue is an urban myth/legend — something that makes logical sense but has no basis in fact. I think this alleged generational skills set chasm is a ruse being fed by mainframe competitors trying to create fear, uncertainty, and doubt (FUD). But, oh, that’s right — my opinions are to be “taken with a grain of salt” — so never mind… (Just FYI, when I did the research for this report I found no panic at the mainframe executive level — they’re growing new talent from within, outsourcing, and hiring newly graduated students [bear in mind that IBM’s System z Mainframe Initiative people are reporting that 50,000 students are taking mainframe courses — and the number of schools teaching mainframe courses has almost tripled over the past 5 years]). If you can substantiate that this skills chasm really exists, I’m all ears. If not, why is this FUD in your article? 
Just for the record — I did enjoy the “golden goose” section. I learned some new things (which is why it is important to read material from a bunch of sources — even if you don’t agree with them).  
Last, but not least, when you recommend that your readers view with skepticism research done by certain analysts, I believe you do your readers a disservice. I hope they’re smart enough to find a report that offers opinions (such as yours) and counter opinions (such as mine), and then draw their own conclusions.  
Hope this helps. 
Sincerely, 
Joe Clabby 
President, Clabby Analytics

Registered

2. 05-04-2010 18:54

Dear Mr.Clabby 
 
Thank you for your comments...It provides so needed balance to the article above. My company is serving one big client that has been trying to move off the mainframe for decades. They managed to create nice screens, but it was a huge volume of batch processes that anchored them with zServer forever, I hope:). So my advice to the client is: just relax, be proud the you are a big guy that needs a considerable computer power to process data, enjoy high security, availability, and scalability.  
 
Again, thank you for your comments. 
 
Sincerely, 
 
Allen Gindler, 
President, Cyberrex Associates

Registered

3. 05-04-2010 18:54

Dear Mr.Clabby 
 
Thank you for your comments...It provides so needed balance to the article above. My company is serving one big client that has been trying to move off the mainframe for decades. They managed to create nice screens, but it was a huge volume of batch processes that anchored them with zServer forever, I hope:). So my advice to the client is: just relax, be proud the you are a big guy that needs a considerable computer power to process data, enjoy high security, availability, and scalability.  
 
Again, thank you for your comments. 
 
Sincerely, 
 
Allen Gindler, 
President, Cyberrex Associates

Registered