Published: February 02 2012, 06:11 PM | no comments
by Marcel den Hartog

Everybody in IT knows that most of the IT budget is spent on "keeping the lights on". This is seen as "not good" because, according to some, IT should spend more money on innovation, adding business value etc. Spending 60-80% on keeping the lights on is equivalent to "wasting money". And somehow, nobody ever seems to question statements like this. Well, I do...

Please look back 24 months. We hardly caught our breath after crisis # 1 when crisis # 2 hit us. Frozen budgets for up to 3 years in some cases, but budget was invested in innovation because of projects using Cloud and Virtualization. However, because budgets remained flat, the money had to come from savings elsewhere, so we cut some staff, we delayed the acquisition of new hardware, we didn't not upgrade some software and all was well. After almost 15 years of continuous growth, we suddenly felt that we actually COULD do the same with less. We thought.....

But did we actually do a proper risk-assessment when we made those decisions? What we did not realize is that we are more dependent on IT than ever, and that it would take some time for things to take effect. And a well managed system doesn't turn into a "sub-optimal" managed system overnight......  Some 8 months into crisis #2, we see the effects of NOT spending enough money on "maintaining the current state".  Maybe its anecdotal evidence, but I see more systems failures than I have seen in a long time. I personally experienced the unavailability of Internet Banking systems, Airline booking systems, travel websites and others and I read about many more in the trade magazines and online publications or newsletters.

Let me first say that I too know that we sometimes spent money on the wrong things in the past. Many projects to make IT better exceeded budget, many hobby projects were too expensive and things could have done smarter. But simply making a sweeping statement that spending 50% of the IT budget on maintaining and optimizing the IT Infrastructure (which happens to be the backbone of the company) is the wrong way to spend the IT budget is naive and misguided. To make the comparison, a production company doesn't spend 60-70% of the production budget on innovation; they spend most of the money on keeping machines running that make the goods that generate revenue.

Am I comparing apples and pears?  I do not think so. All I DO know is that in the past few years, selling software that simply helped to optimize the existing IT Infrastructure was sometimes seen as a sub-optimal investment. Here is a surprise, it's not. How much does it cost a financial institution when their Internet Banking goes down for several days and how does this compare with the savings that created the unstable IT Infrastructure that caused this? Spending money to make the existing IT Infrastructure more reliable is well worth it.

Please leave a comment, I am curious to hear your opinion.

Share this post:  Email

Share

Published: January 12 2012, 11:57 AM | no comments
by Marcel den Hartog

2012. A year where nobody seems to know what to expect.  There is uncertainty about almost everything.  Since we KNOW it will be an uncertain year, at least we can prepare ourselves for it, which, in itself, is a comforting thought.

According to the Chinese Zodiac, this is also the Year of the Dragon, which officially begins on January 23, 2012 and ends on February 9, 2013.   This is very appropriate to Mainframers.  Why?  The Dragon is a creature of myth and legend, and representing royalty and power.   Similarly, the mainframe is legendary within our industry, but when you want to talk about the state of the mainframe in 2012, opinions by many people are largely based on myths. 

This subject has been covered in many articles and blogs.  It's the historical facts about the evolution of the mainframe that we don't see enough about.  Mainframe myths are no more accurate than stories about dragons. For those wanting to educate people, tell them to watch the video series: Big Iron - The Mainframe Story.  And once you've educated them, they will also see why the mainframe has an important role to play in the IT strategies of many companies this year.

So I guess it's safe to say that the 2012 could be the year of the legendary mainframe. Why? I have created a small list of solid, economical, environmental and business reasons why this is not just a statement.

• 1. Green IT.
  A recent survey from Arcati shows that 56% of the respondents (down from 72% last year) don't seem to bother with "Green IT". I personally think people are confused by the difference between "Save the planet" and "Save Money". Once they understand they can save money by moving workload to the mainframe, and as a consequence, reduce their energy bills, suddenly green can be translated in "green dollars". And in these times, that's more than welcome.
  
• 2. Cloud.
  Yes, workloads have moved to the Cloud already, and more will follow. 2012 will be the year where we will not only see the real cost of doing this; we will also realize that our existing data and applications will not go away as quickly as promised by many Cloud evangelists. Many Cloud Services will be driven by data that resides on the mainframe. So unless you are going to duplicate all of it, you better make sure the mainframe can support the additional and often unexpected workload that is the result of these new Cloud Services.
  
• 3. Aging Workforce.
  Even though some seem to disagree, customer surveys and my own anecdotal evidence (people at GSE conferences telling me they will retire in the next 12 months) tell me we are experiencing a mainframe brain drain. We MUST start using the right tools in 2012 to capture the knowledge of these people so it can be easily re-used by the next generation mainframers.
  
• 4. New Mainframe Management Solutions.
  New generation of mainframers need new tools. Not only because we can't simply expect them to work with an arcane 3270 interface (no matter how powerful), but also because the mainframe today is NOT the mainframe from 25 years ago and today's complexity simply demands a more powerful and easy to use interface that helps analyze massive amounts of data in seconds.
  
• 5. Big Data.
  Yes, this IS new, even though many mainframers think it's not. Big Data means actually DOING something with all the data we have. And no matter what some people might say, the mainframe is a very cost effective platform to run things like this without having to copy massive amounts of data to other platforms.
  
• 6. Pragmatism.
  If there is one thing the history of IT should have taught us, it's pragmatism. "Old" things don't just disappear; "new" things are not always better and more efficient. Doing things only because "you can" will no longer be acceptable by our educated Business Managers. Bias needs to go. It was never healthy to start with, and it limits our efforts to run IT as efficiently as possible. Biased people tend to fanatically hold on to what they know and this is not the right attitude in times where budget constraints are the norm, rather than the exception. "Fit-for-purpose" will evolve to be the best way to run IT. Some things will fit better in the Cloud, some on the mainframe and some on anything in between. Good. Be pragmatic about it.

Share this post:  Email

Share

Published: November 16 2011, 10:19 AM | 1 Comment(s)
by Marcel den Hartog

The World of IT evolves. And like things in the real world, we have big revolutions (often viewed as hype) and in between those, we have evolutions. The evolutions are necessary to keep things going and to clean up some of the mess that is left by the revolutions. With this in the back of our minds, what does this make CA Technologies recent “Next Generation Mainframe Management” announcement?

To put it simply, some components have been gradually improved, some have been enhanced to make revolutions possible while others will cause a small revolution by themselves. The mainframe, by nature, is an environment where the word “revolution” does not fit. We like things to be stable, reliable and predictable. But even the mainframe sometimes needs a revolution to make sure it’s still viable in the decades that lie ahead of us.

The revolution in this case is definitely IBM’s zEnterprise and the zBX. The latter allows us to run and manage distributed workloads that depend heavily on the mainframe almost as though they are actually housed of the mainframe. So making sure that this complex environment and new technologies like Cloud are supported is key, and you will see some revolutionary changes in our strategy and products.But we are faced with many other “challenges. Challenges like: lowering the TCO of the mainframe, knowledge transfer to the next generation of mainframers, support for new technologies etc. All these challenges require changes that could be classified as evolutionary but once you look closer, you will realize that even those changes will have revolutionary elements. 

• CA Mainframe Software Manager, a new and easier ways to install, maintain, deploy and configure software using a trusted technology like SMPE but with an interface that allows the less experienced to perform these complex tasks with ease
• The CA Mainframe Chorus platform, that drives DB2-, Storage- and Security management solutions which have been around for 20+ years, offers a huge leap in user-friendliness, productivity, flexibility and knowledge sharing
• Solutions that have been enhanced to allow them to manage Linux on System z as an integrated part of the existing zOS environment and last but not least;
• Mainframe solutions that integrate closely with their distributed counterparts to help manage and monitor complex multi-platform applications

Share this post:  Email

Share

Published: November 15 2011, 12:43 PM | no comments
by Scott Fagen

So, I’m a couple of days into CA World ’11 and my feet hurt from being upright for double-digit hours every day, but it is heartening to see that our solid core of mainframe customers are here in force.  In fact, there are more mainframers here than the two previous CA Worlds!  The Exhibition Center is buzzing, especially around our new products, but there is also excitement around the new capabilities of the mainstays, as well. 

It’s hard for me to believe that I’ve only been with CA for a little over four years now and that the work we’ve been doing has made such a positive impact on the mainframe world.  Last week, I received an email from one of my colleagues, Greg Shriver, who was part of the infancy of what became “Mainframe 2.0.”  In the email was a set of PowerPoint slides that contained digital photographs of what we were trying to do – change the way the mainframe is managed forever.  His words echoed my thoughts, “Wow, pretty amazing that this was less than 4 years ago…”

CA World ’11 is actually marking a transition for the mainframe business at CA.  When I joined in September of 2007, the company had just formed (or, better, reformed) the mainframe business, coalescing products that had been under separate management chains for many years.  While such an organization isn’t necessarily a “bad thing,” it was not reflective of what was going on in the market.  The IBM System z10 was just around the corner and customers were beginning to recommit to (or at least admit to the fact that they weren’t abandoning ) the platform. 

For us, Mainframe 2.0 was more than just a set of things to do, it was the metaphor for changing how we went about doing our business.  I will freely admit that much of what we did borrowed liberally from my IBM past, but many of the ideas were improved upon by my CA team of architects, developers, testers and even management.  Mainframe 2.0 became the rallying point for our setting high standards around the management of software.   It helped customers manage cost through identifying ways of getting greater value from the software the customer already has and expanding the use of specialty engines within our software.  Mainframe 2.0 is also the springboard for the advent of the next CA mainframe strategy.

Here at CA World ’11, we are kicking off the “Next Generation Mainframe Management” strategy, which builds upon the foundation laid by Mainframe 2.0.  None of the initiatives within Mainframe 2.0 are forgotten, they are now considered “business as usual.”  With “Next Generation Mainframe Management” we address the next set of challenges ahead:  the transitioning of the workforce from the experienced gurus who’ve been at this since the mainframe was the only game in town to a new cast of “ambidextrous” IT staffers who work “across the silos” instead of vertically within them.  Our new products and product releases are targeted at not only making this new workforce able to function this way, but to do so effectively. 

And, I’m not just talking about the silos within the mainframe discipline, we have been working on solving many of these issues across the datacenter, with products like Cross Enterprise APM (Wily and Sysview), Workload Automation (ESP, CA 7, and Autosys), we’ve only begun to break ground on some of the exciting new technologies that we are working on in the lab, like bringing our AppLogic technology to create a compute grid that not only encompasses Linux on System z, but allows for easy connectivity to z/OS transaction and database management.  With zEnterprse, we’ll be able to extend that grid to application serving on the blades in the zBX.

Yes, we’ve been working very hard the last four years, laying the foundation for some very exciting advances in the near future.  It’s exciting to see that the customers here at CA World really “get” what we are doing and how supportive they are of our efforts.  Of course, the reward for hard work is, as always, more hard work.

Share this post:  Email

Share

Published: November 09 2011, 05:11 PM | no comments
by Marcel den Hartog

Every single day during the last 20 years in my career I have been confronted with the (sometimes) bitter arguments between people who manage mainframes and those who control the distributed environment. Often, the discussions were emotional, sometimes even bitter and almost always they had no positive effect on the reputation of the IT department. Call me an optimist, but I sincerely hoped that every time a new "wave" in IT occurred, it would be the reason for both sides to start working together more.

Let me explain; in companies that owned (still own) a mainframe, distributed systems were embraced as the platform that would solve all the so-called problems the mainframe had; high cost, inflexibility and complex management. History has taught us that these promises have not worked out as expected. Of course, we now do more with IT than ever before, we run different workloads, have more users, etc, but ask any businessman about the state IT in his/her company today and he/she will tell you it is too expensive, inflexible, not agile, too much money is spent on managing and not enough on innovation. Hence the enthusiasm for Cloud. Again, we believe/hope that this new technology will solve all the "problems" IT has today.

And what happens in the meantime in the data centre? There is more competition than ever. Instead of joining forces to make sure that we do not automatically move everything to a Cloud based infrastructure (which would be just as wrong as trying to move everything off the mainframe when distributed systems were introduced) IT folks argue. Instead of focusing on a fit-for-purpose data centre where things are implemented on the platform best suited to run a particular workload, we disagree on even the simplest things. We overpromise savings so workload can be "stolen" away, and then hope that by the time people realize these savings cannot be achieved, the project can no longer be stopped.

Even the threat that it is easier to move distributed workloads to a Cloud infrastructure than it is to move mainframe workload to Cloud doesn't seem to motivate people to start working together in order to make the data centre more cost effective, efficient and integrated. Management tools that manage workload across platforms are seen as tools from the dark side. X-platform Application Performance Management solutions don't get used to bring mainframe and distributed closer together and the question "who owns Linux on the mainframe" seems to be more important than asking what it can do for the business.

It often it takes a "common enemy" to make people realize that they need to join forces. Calling Cloud an "enemy" is not correct, I realize that. Cloud IS happening for all the right reasons and has great advantages for certain types of applications and workloads. But only if we team up can we help prevent our companies from making the mistake many of us made years ago - and that is using a new platform for "everything", just because they can.

Oh, and yes, please forward this article to your distributed colleagues....

Share this post:  Email

Share