Arcati Mainframe Yearbook 2008 user survey

The Arcati Mainframe Yearbook 2008 is available for download free from www.arcati.com/newyearbook08. It’s full of interesting stuff, but particularly interesting are the results of the user survey.

Only users from sites that actually had a mainframe were included in the survey results – although nine additional people who didn’t have a mainframe completed the survey! Respondents all completed a survey on the Arcati site during December 2007. 42% were from Europe and 36% from North America, with 22% from the rest of the world.
48% of the respondents worked in companies with upwards of 10,000 employees worldwide, while 21% had 1001 to 5000 staff. 11% of respondents had 0-200 staff and similarly 11% had 5001-10000. In terms of MIPS 41% of respondents (down from last year’s 47%) had fewer than 1000 MIPS installed, 30% fell into the mid-sized category between 1000 and 10,000 MIPS, and 23% were at the high end.

Looking at MIPS growth produced some interesting results. Larger, more mature businesses (above 10,000 MIPS) were mostly experiencing healthy growth – with nearly 30% in the 26-50% range. However, over 10% of sites in the 500-10,000 MIPS range were experiencing a decline, while at the same time 42% were experiencing 10-25% growth. Sites below 500 MIPS had a similarly mixed picture. Just over 28% of respondents said that their MIPS growth was static or in decline, whereas just over 28% reported growth above the 10% mark. The mainframe market does appear to be quite fragmented with competitive pressures at the lower end of the mainframe market.

Opinion is clearly divided over the role of the mainframe in new applications. Some companies regard the mainframe as a separate legacy environment while others are leveraging the strengths of large systems by using them to deploy new workloads and technologies. 48% of organizations said that their mainframes participated partly or fully in Web services. 41% say they run Java-based applications on the mainframe. Also, 68% of organizations were Web-enabling their CICS sub-systems, with 62% Web-enabling DB2.

There are a huge number of other gems of information in the survey. Comfortingly, it concludes that the "mainframe offers an increasingly cost-effective, secure, and powerful platform for companies that have the necessary background and expertise in place to support it".

Download the whole Yearbook for yourself.

The Arcati Mainframe Yearbook 2008

The 2008 edition of the celebrated Arcati Mainframe Yearbook is now available to view or download (free of charge) from www.arcati.com/newyearbook08.

It includes a mainframe strategy section with a selection of articles by industry gurus and vendors on topics such as: managing z/OS FTP, energy efficiency, using refurbished mainframes, integrating WebSphere Service Registry and Repository with WebSphere Enterprise Service Bus, cross-platform Java application performance management, using ping with CICS, and modernizing mainframe systems.

For many people the highlight each year is the mainframe user survey. This illustrates just what’s been happening at users’ sites. It’s a good way for mainframers to compare what they are planning to do with what other sites have done. I will be looking at some of the survey highlights next week.

The full vendor directory section gives contact details and other useful information in a summary format for all major vendors, consultants, and service providers in the z/OS environment.

There’s also a media guide for IBM mainframers. This lists information resources, publications, and user groups for the z/OS environment.

Next comes the glossary of terminology, which contains a large number of definitions of mainly mainframe-related terms.

The final section contains technical information – z9 model tables; mainframe hardware timeline 1952-2008; and mainframe operating system development.

In order for the Yearbook to be free, it is supported by sponsors and advertisers. This year’s sponsors were Software Diversified Services (SDS), Epoka Group A/S, Macro 4, and Iona.

To see this year’s Arcati Mainframe Yearbook, click on http://www.arcati.com/newyearbook08.

Now that the Yearbook is complete... if anyone needs a good technical writer etc, please contact me at trevor@itech-ed.com.

Data integration and mainframes

SHARE surveyed its members back in July and August 2007 and has published the results in a report entitled, The New Mainframe: Data Integration and Service-Oriented Architecture, Big Iron Style. Its 431 respondents led SHARE to conclude that, "mainframes are evolving into a leading role both as a source of mission-critical data, as well as key services". They went on to say that this "immense, scalable, real-time transaction platform for the highly-integrated environments" is needed by today’s businesses. Coming from a mainframe organization these are not startling conclusions – they just really confirm what we already knew.

Obviously, SOA (Service-Oriented Architecture) is on everyone’s mind, and the survey found that nearly one in four respondents’ organizations are currently working on SOA projects. They also found that one in three are planning or considering SOA, of which at least half will employ mainframes in a central role. The survey also found that although most SOA efforts don’t reach all the enterprise, many companies are preparing SOA to meet real-time requirements. The figures showed that about 40 percent of companies are deploying or considering event-driven architecture.

What was, perhaps, surprising was that at least half of the surveyed sites still use hand-coded scripts to move data from their mainframes to other platforms or databases. It also found that while mainframes store and manage much of their enterprise data, most of it remains inaccessible. In fact, the majority of mainframe sites share only a small portion of their data across enterprise systems.

What this clearly illustrates is the need most enterprises have for something like SOA that will make the data they store – and remember figures between 70 and 90 percent are quoted for the amount of a company’s data that is stored on a mainframe – available to non-mainframe users. Perhaps it also illustrates that SOA is not as easy to achieve as a million Powerpoint presentations suggest!

One problem that many sites face is easily identifying programs – or even parts of programs – that can be turned into Web services. Another problem is that most of the mainframe programs are written in COBOL while the Web-based programs are mainly Java. There is a big difference between the two languages and it is quite hard to convert COBOL programmers to Java programmers. A third difficulty many sites face is that their experienced COBOL programmers are getting older – with many approaching or passing retirement age.

So what technologies are enabling mainframe SOA? The survey figures were 35% application server on another platform, accessing the mainframe; 31% application server residing on the mainframe; 31% XML infrastructure devices; and 27%, composite applications on another platform, but accessing mainframe. The other options scored below 15%.

The report makes interesting reading. But it looks like mainframes and data integration with the rest of the enterprise still has some way to go.

Next week: the new Arcati Mainframe Yearbook 2008.

IBM futures – Elseworlds

If you’ve ever read DC comics – their characters are Superman, Batman, Wonder Woman, etc – you know that sometimes the authors come up with a great idea for a story featuring one of their heroes that doesn’t quite fit in with the regular continuity. So the story is published as an Elseworlds story, and we the readers can enjoy it, and regular storylines in the DC universe are not affected by it. Marvel comics run something similar under the "What If …?" tag.

So, this blog is an Elseworlds story. It looks at what could happen to IBM in the next few years – but it won’t impact on normal continuity!

So here’s the thing… what if IBM was at war with Fundamental Software (the FLEX-ES people), Hercules, and the others because it (IBM that is) was planning to release in September 2008 an Intel-based version of zVM.

You would be looking at the ultimate virtualization server – with 40 years of VM experience more than its nearest rival. Customers would have a product that might well do more work and use less power than anything else on the market. It would be a green solution. One little box (or rack) would do the work of lots of individual servers – which is the usual rational for virtualization – but this way customers get the advantage of IBM’s years of experience and IBM gets to sell more software – lots more software. And, as I’ve said before, once customers are seduced into the IBM way with small machines, they might want to get a big one – a mainframe.

zVM runs multiple operating systems concurrently, it continues even when over-utilized (they claim up to 123%), monitoring and management are well understood, it links well with Linux, it’s everything a company moving towards virtualization could want. It’s big drawback at the moment, as far as many potential customers are concerned, is that it comes on a mainframe.

Companies adopting Intel-based zVM would reduce their power consumption, their cooling costs, the amount of space the data centre takes up, and their manpower. IBM would have a runaway best seller on its hands.

On the downside, they would lose a few smaller mainframe customers to the cheaper new option, but what they’d gain is a presence in almost every medium-sized data centre, something they’ve not had since the 80s.

Like I said, it’s just an Elseworld idea.

Getting back to the real world – all you high-end IBM users with IMS installed will be pleased to hear that the second Virtual IMS Connection user group meeting takes place on Tuesday 5th February at 4:30 GMT (10:30 CST). Anyone using IMS can sign up for free membership at www.virtualims.com. The meeting is completely free, and contains a presentation by NEON’s Bill Keene on IMS disaster recovery preparation.

Blog tag – I’m it!

I just got an e-mail from Craig Mullins (http://www.db2portal.com/2008/01/blog-tagged.html) saying that I have been blog tagged. I had to check on his blog site to see what it all meant, and apparently it’s the latest game. Each blogger reveals eight things about themselves (originally five) and tags eight (originally five) other bloggers – who in turn do the same. It’s like a chain letter without the final sentence of doom to anyone who breaks the chain. It seems like a bit of fun – so I thought I’d join in.

Craig was blog tagged by Willie Favero (http://blogs.ittoolbox.com/database/db2zos/archives/ive-been-blogtagged-i-guess-im-it-now-21817), who says he was tagged by Lewis Cunningham. Willie also says that the first blogger he can find mentioning the game is Jeff Pulver (http://pulverblog.pulver.com/archives/006087.html).

So here are some facts about me:

1. I live in the UK near Bath and Bristol and not too far from Silbury Hill and the ancient stones of Avebury. My family moved here 21 years ago, and my wife and I have stayed put ever since.
2. I’ve been a school governor for nearly 15 years, and chairman of a local charitable trust for eight years.
3. I have two grown-up daughters who are both at university at the moment – other parents will recognize just how much this costs!!
4. I play the guitar (badly) and make feeble attempts at playing the saxophone. But I enjoy it.
5. I have written and published three books – two on VM performance and one on MVS automation.
6. I set up my own company (iTech-Ed Ltd – www.itech-ed.com) in February 2004 and it is still going.
7. I have enjoyed holidaying in Kenya (safari) and China – you can see these photos on Facebook, where I also started the Eddolls group for people across the world with the same surname (obviously!!).
8. I taught myself to juggle and am now trying to learn how to ride a unicycle.

So, who to tag next? There are just so many people who have already been tagged. So here’s my starting point:
Aseem Anand http://aseemanand.blogspot.com/
Marc Wambeke http://mainframe-watch-belgium.blogspot.com/
Ian Steyn http://www.pcs305.com/
(I’ll add more later…)

They are now "it"!

Bringing home the bacn!

I’ve known about spam for years. On many occasions I have ranted about the fact that sometimes 90% of my inbox is spam. I have opted not to use the executive-must-have toys such as a Blackberry because I know the "important" e-mail that has just arrived will be offering me – well you know the sort of thing these kinds of e-mail offer.

What I didn’t realise, until a recent interesting conversation down the pub last week, is that someone has divided e-mail into two categories – spam (the stuff you never want) and "ham" (the stuff you do want). That all seemed to make sense. It’s useful if everything has a name. Although I must admit I’ve never heard anyone come into the office and say, "nearly half my e-mail is ham today".

What got me checking the date to see whether April Fools’ Day had come early was when the conversation turned to a third type of e-mail called "bacn". I looked it up on Wikipedia later – and there it is fully defined etc. So definitely not a hoax. I hadn’t been drinking too much.

Now readers of this blog will fall into two groups, those who already know about bacn and those who, like me, had never heard of it. So, for those of us in group 2, bacn refers to those e-mails that you want, but you don’t really want at this minute. For example, you subscribed to the Trainspotting newsletter one afternoon when you weren’t to busy at work and now, when a new edition arrives, and you’re completely inundated with work, you leave it in your inbox until later – a later that may never arrive before the next newsletter turns up. So bacn is basically solicited e-mail that goes unread.

I’m not sure how you’d classify the e-mail from Facebook telling you a "friend" has just written on your wall. Wouldn’t it be easier just to put the message in the e-mail and save me logging in? But then perhaps I’d just e-mail my friend and miss out the whole Facebook experience.

Anyway, back to bacn. The term bacn, apparently, was originally coined at PodCamp Pittsburgh 2. Since then, according to Wikipedia, it has become popular amongst the blogging community – but not so popular that I was aware of it until last week. Perhaps I should get out more – or perhaps I get out too much and I should be at home reading everyone else’s blog!

So if the conversation turns to e-mail, you can confidently refer to unwanted e-mail as spam, personal and wanted e-mail as ham, and whatever’s left as bacn. You’ll either sound like and expert or an idiot!! Let me know which.

Going back to Wikipedia, I’ve always been a big fan. I use it as a TV episode guide, to research obscure topics that crop up in conversation, and everything else. I was surprised that Google was setting up a new equivalent called Knol. I’m sure competition is a good thing, but it’ll mean checking two sources rather than one. A bit like a man with two watches never quite knowing what’s the right time. So, nil point to Google for that idea.

And finally... With spam, ham, and bacn, what kind of e-mail would be the equivalent of a pork scratching? I’d like to be the first blogger to use the term "pork scratching" to refer to an unread RSS feed you have set up!

Costs, age, and hen’s teeth

The Arcati Mainframe Yearbook 2008 user survey has recently closed. At the very end of the survey was a box that gave respondents and opportunity to add any comments they wanted. This has made very interesting reading. And, without giving away any of the survey results before publication, I would like to draw your attention to some of the comments.

Now bear in mind that everyone who completed the survey works with IBM mainframes, so the comments are from died-in-the-wool, Big Blue, true believers – not Microsofties or secret Unix enthusiasts. And, of course, not everyone left a comment, and there were comments on other topics as well.

Cost was a definite concern for many respondents and two highlighted IBM support – or the lack of it. I don’t know whether other people have been dissatisfied with IBM support recently, but I’d be interested to hear. Both these respondents were reluctantly moving off the mainframe – which must make sad reading for Sam Palmisano and IBM shareholders.

Other respondents commented about the age of mainframe staff, with one giving details of the forthcoming 60th birthdays of the individuals who look after the mainframe. The word “grey” was used a lot. Age itself is not so much the issue as the fact that CICS, IMS, DB2, and z/OS itself need to be supported, and the experts who can do that are approaching retirement age. How can an organization plan ahead when the very software that the business runs on may not have experienced people to run it in the medium-term future?

Another respondent likened experienced mainframe support staff to hen’s teeth – ie pretty rare.

These respondents indicated that the thinking at their sites was to migrate off the mainframe while they still had people who understood the mainframe, and run the business on hardware and software that was less effective than mainframes, but which had plenty of support staff and experts that could be employed as necessary for the foreseeable future.

Although these are not necessarily new concerns, it must be a worry to IBM that they could be losing business while at the same time having the best hardware and software in town simply because of an ageing and shrinking population of mainframe experts. Let’s hope they already have a recovery plan and its coming to fruition in 2008.

And, as I said last week, I am currently one mainframe “expert” that is available for consultancy, writing work, etc. You can contact me at trevor@itech.com.