Published: June 19 2009, 06:12 AM | no comments
by Reg Harbeck

If the title of this blog entry got your attention, I hope it also gets the attention of everyone that has been saying "death to the mainframe" for the past three decades.

After all, given the number of people who have been trying to portray the mainframe's invisibility as some sort of demise over the years, it seems like that's actually a good indicator of something that works - so well that nobody notices it. Just like that old saying, "housework is something nobody notices unless you don't do it."

So, now, Linux, that friendly operating system with the penguin mascot (named "Tux" as it turns out), has arrived on the invisible platform. Actually, it's been there for nearly a decade - but, as I mentioned in my blog entry from March 12, 2009, its progress has also been somewhat invisible.

As Wikipedia reminds us, IBM first announced mainframe Linux in 2000.

And, ever since then, everyone's been waiting and watching and trying to figure out what has happened to this penguin on its journey into the world of mainframe.

Well, someone has finally found the answer - and it's good news!

According to a Press Release that CA issued this past Wednesday, the folks at TheInfoPro have done a survey of large mainframe shops that have or are getting mainframe Linux, in order to find out what people are doing (see http://ca.com/mainframe/linuxresearch). The answer? They're growing it, and moving both new and existing (especially distributed) applications to it, in order to take advantage of the significant strengths (virtualization, security, scalability...) and cost savings available on the mainframe.

Since CA has a large and growing stable of products for managing mainframe Linux (see ca.com/mainframe/linux), this is good news for CA and the rest of the mainframe world, as it affirms the choice to join this penguin on its sojourn.

So, what about the death of the penguin? Just as with the mainframe, nothing could be further from the truth. It's just been too busy taking root under the surface (if I may mix metaphors), and taking on many of the same production qualities that we've come to take for granted on the mainframe.

Something tells me we'll be hearing plenty more about this before long.

What do you think? Are you or your organization using or considering Linux on the mainframe? What do you have in mind for it?

Published: June 01 2009, 07:04 AM | no comments
by Reg Harbeck

I just checked Wikipedia.org, and the disambiguation page for "Overdrive" took me by surprise with the large number of entries, all of which seem to hearken back to the original, automotive concept of a high gear for fuel savings when you're already cruising at full speed (see http://en.wikipedia.org/wiki/Overdrive_(mechanics)).

Interestingly, one definition I didn't find would refer to the effect that IBM's zAAP and zIIP specialty engines have on certain mainframe software, and by extension the whole mainframe.

Well, I'm going to use it for a CA software product that, as we're announcing today, is going into overdrive by taking advantage of the zIIP.

As you may recall from my blog entry of January 8, 2008, when IBM introduced their specialty engines, CA immediately announced that we'd be taking advantage of them. And we've certainly kept that promise, with a growing list of products that allow your mainframe, already running at full speed, to give you software economy by running some of our software on specialty engines that don't entail additional software costs. The growing list of such products even includes new ones such as our CA CMDB Connector for z/OS.

And today, we've fulfilled that promise with another important GA announcement: Datacom r12 has just gone GA with zIIP exploitation!

That makes a total of 14 CA mainframe products that now bring value to the zIIP processor (see http://ca.com/mainframe/zIIP).

So, what do I mean by overdrive in this context? Well, suppose your mainframe is running around 100% busy (as mainframes often do), and you add in a zIIP. Suddenly, you're able to get more work done without increasing your billable capacity - a lot like shifting into overdrive at high speed to go further on less fuel.

One of the cool things specific to CA Datacom (and CA IDMS - our other outstanding mainframe database), however, is that by becoming zIIP-enabled, it is bringing a higher level of fulfillment to the original promise of IBM's zIIP: make mainframe database processing more cost-effective. In fact, the case of our databases, you're not limited to new workloads: any kind of work you're already running on CA Datacom (or CA IDMS) can be a candidate for running on the zIIP.

You know, all this talk of overdrive and business value reminds me of something. It's a song by Bachman-Turner Overdrive: Takin' Care of Business. ‘Cause that sure is what we're doing!

Published: May 27 2009, 07:13 AM | 3 Comment(s)
by Reg Harbeck

Hello from CA's Prague Mainframe Center of Excellence! This week, I've had the pleasure of meeting with some members of our new generation of mainframers in person.

As you'll know from my blog of May 5, 2009 (see http://community.ca.com/blogs/execio/archive/2009/05/05/gdg-generations-doing-good.aspx) or our New Mainframers video (see http://www.ca.com/files/videos/mainframe-recruit_205295.asx), CA has been building a new generation of mainframers around the world, including here in Prague.

This is my first visit to Prague, and it has been a great experience to meet the outstanding people here. We even got together for a group picture:

 During my visit, in discussing what it means to be a new generation mainframer, we came up with a new term, or perhaps job title: SysPrague.

Of course, it's pronounced similarly to the more established title, "Sysprog" which is short for "Systems Programmer". But, being here where a new generation of mainframers is spreading their collective wings, I thought: maybe it would be good to come up with a definition for this term.

So here's my suggested definition (you're welcome to offer improvements to it if you wish). A SysPrague is a new generation mainframer that knows all about the Internet and distributed technologies, can program in newer languages such as Java, and is bridging the gap between the distributed world and the mainframe by learning all about the mainframe and making it accessible to a new generation that has never heard of a 3270. SysPragues are people on a journey of discovery, which will lead to them becoming in-depth technical experts on the mainframe, as well as a keepers of its culture and history, and who are already effective today with newer technologies, interfaces and approaches that truly bring out the hidden meaning of "legacy": "it works".

In other words, a SysPrague is the kind of person you get when you take a brilliant new-generation person and immerse them in the ways of the mainframe, and then let them take what they already know about modern interfaces and computing and bring everything together into a whole new paradigm for effective management of the most important business platform on earth.

Of course, it follows that you don't actually have to be based in Prague to fit this definition - indeed, new CA mainframers in the Dallas, Pittsburgh and Chicago areas, for example, could just as easily merit this title.

So, here's to a new generation of SysPragues, moving mainframe management to a new generation as well!

Published: May 20 2009, 07:52 AM | no comments
by Reg Harbeck

If you've been reading this blog for a while, you'll know that one of the subjects I often mention - and often give presentations about - is regulatory compliance (for example, see http://community.ca.com/blogs/execio/archive/2009/04/02/pci-and-other-compliance-related-tla-s.aspx).

In a way, it almost seems odd that a mainframer should spend so much time on this topic, when the mainframe is clearly the ideal platform for running in a manner that is compliant with all regulations - so what's the big deal?

Well, there are a few problems:

• Many regulations that have explicit technical requirements were written with PC's in mind, rather than mainframes, so it's necessary to translate their requirements to fit this context.
• Many auditors don't know the mainframe well enough to be able to figure out whether or not it's compliant, except to the extent that the very technical people who are responsible for the mainframe explain it to them.
• Even if the mainframe is being run in a compliant manner, proving it has been a very challenging task in the past.
• To complicate matters further, a new generation of mainframers is arriving, and they have yet to take the same journey of figuring out what is and isn't appropriate on the mainframe that we experienced folks took a few decades ago, so we can expect them to bump into and possibly occasionally knock over a few things on the mainframe while they figure things out. That can lead to accidentally coloring outside of regulatory lines.

So far, when I've talked or written about these issues, I've mentioned the value of cleaning up obsolete IDs and permissions, validating that your mainframe security directory contents support compliance, ensuring a compliant configuration of your OS environment, and encrypting data sent off-site. But that's all stuff that could be thought of as "below the line" - it supports compliance, but it doesn't actively pursue it.

That's why I'm so excited about something we're announcing today: CA Compliance Manager for z/OS.

This brand new mainframe product brings together your security, operations, systems and auditing groups with a single, real-time system that watches the behavior, use and configuration of your mainframe environment and security system (using any of the three external security manager products, or ESMs), and alerts you to any events or changes that could negatively impact your organization's compliance with relevant regulations.

And it does it all on your mainframe, so there's no "weak link" that could interfere with it. In fact, it even tells you if someone restarts your ESM. And all of it is so real-time that it doesn't even use SMF - it watches the system itself.

So, for example, if someone makes a change to an important PDS such as SYS1.PARMLIB, you'll know right away: when, who and what.

The idea here is, once again, to move from the technical to the business perspective. The business is the computer. You want accountability for changes that are made to your environment, measurability of what things have happened, and transparency, so there are no hidden behaviors that could provide unpleasant surprises later.

Best of all, with its web UI, it's so easy to use that less-technical personnel can be effective right away - including those auditors and managers who don't have a mainframe background, not just those that do.

Can you tell I'm excited about this?

But don't take my word for it - check it out for yourself. Your local CA mainframe team will be glad to tell you more - and see our press release (http://www.ca.com/us/press/release.aspx?cid=206928) for more details, too.

Why put up with complaints from your management, auditors, and potentially customers and business partners: now you can prove that you're compliant, and stay that way!

And, once you've checked it out, let me know what you think.

Published: May 05 2009, 08:34 AM | 1 Comment(s)
by Reg Harbeck

It's happening! I know because I've seen the video: http://www.ca.com/files/videos/mainframe-recruit_205295.asx.

A new generation is arriving on the mainframe!

This is some of the best news I've heard since I wrote my original whitepapers (http://ca.com/us/whitepapers/collateral.aspx?CID=64111 and http://www.ca.com/us/whitepapers/collateral.aspx?cid=78289) and article (http://zjournal.com/index.cfm?section=article&aid=248) about the need to get a new generation in place.

Over the past four years, I've seen signs that the mainframe world was starting to wake up and begin to deal with the need to get a new generation in place, starting with such things as the SHARE zNextGen project (see http://www.znextgen.org/), very ably led by 20-something new mainframer Kristine Harper Neely (see her blog at http://www.neonesoft.com/blog/blogs/kharper/default.aspx).

I've also seen CA establish new initiatives, such as hiring new mainframers in North America and around the world, including at our Mainframe Center of Excellence in Prague.

In fact, just today, CA has even issued a press release about the important role that new 20-something mainframers are playing in developing our Mainframe 2.0 initiative and other key things we're announcing this month as part of May Mainframe Madness.

Clearly, we're well on our way in the journey to get a new generation in place and effective on the mainframe, and that's great news!

Of course, that's just the beginning. Now it's time to use all this good news to start waking up the management of mainframe shops around the world to the value of their mainframes and the people that run them.

How about you? Does your management realize the importance of their mainframe environment and people? What will you do to make sure they do? Feel free to start by sharing this blog and the above links. Let's get the word out!

Published: May 01 2009, 07:43 AM | no comments
by Reg Harbeck

It's May first! May Day! And, for the first time ever, CA is beginning our month-long celebration of the mainframe that we're calling "May Mainframe Madness."

Now, if you don't live in the United States, or maybe just aren't a big basketball fan, you may not catch the reference to a popular American tradition called "March Madness." Suffice to say that it's a celebration of excellence that leads to the emergence of a single, best team at the end of a month of amazing efforts and results.

Which has inspired CA's mainframe team to likewise spend a month celebrating excellence on the mainframe with great educational and informative webcasts, new product and product enhancement announcements, a world "Mainframe Infinity Tour," and, most importantly, the official release of our Mainframe 2.0 strategy including the first cut of our Mainframe Software Manager, which will enable mainframe people, both new and experienced, to far more effectively and easily manage their CA mainframe software.

Of course, there is more to this than just a few announcements and webcasts. After all, the mainframe is the computer that keeps the most important organizations on earth running, and it's important to keep this platform alive, well, and moving forward.

That means adapting it to the emerging needs and opportunities of today's businesses that use the mainframe, beginning with the need for business-enabling simplification and cost-effectiveness. So we're standardizing the SMP/E installation of our mainframe products in a manner that not only makes it easier for both new and experienced people to maintain all their CA software, but also enables us to create and automated graphical user interface which can handle those tasks for you.

We're also taking advantage of IBM's Health Checker for z/OS, having written over 100 health checks already, with many more to come.

Then there are our other initiatives, such as Releasing Latent Value - a set of initiatives that span from our Mainframe Value Program and the Mainframe Solution Center to our Releasing Latent Value Articles and CA Know-How Knowledge Center, and now even a book titled "Releasing Latent Value." (If you'd like a complimentary copy, ask your CA rep or send me a note through this blog.)

And, if all the webcasts, announcements, new products and innovations, and Releasing Latent Value initiatives weren't enough, we're also doing something we call Out Tasking, an initiative to allow customers to get new business value without increasing their staff workload.

As you can tell, I'm pretty excited about what we're doing on the mainframe, and how we're letting the world know about it this month of May. And the even better news is, next year, and the year after, we're planning to do it again.

In other words, to complement the tradition of excellence inherent in the mainframe and the CA software that manages it, we're establishing a new annual tradition, this month of amazing announcements, efforts and celebration of the mainframe: May Mainframe Madness.

Enjoy!

Published: April 26 2009, 10:27 PM | no comments
by Reg Harbeck

"Lean IT" - makes me think of that song "Lean on Me" - though, of course, it's a different meaning of the word "Lean". Still, we do lean heavily on IT, and we need it to be responsive but still responsible with such things as resource consumption and manageable complexity.

And, particularly during these lean times, we need IT to be cost-effective.

That's why CA is making a number of announcements today about how we're supporting Lean IT across the enterprise, including the mainframe.

Which works perfectly for me - remember, my motto is, "The Business is the Computer." So, if the computer isn't enabling the business with all of its features, including its affordability, then it's not meeting that requirement.

And, I have to admit, I'm rather proud of the many ways that CA software helps non-mainframe computers behave in a more lean, business-like fashion. After all, we did it first on the mainframe, then built on that great example.

So let's not forget the original lean, green machine when looking around at all these other examples of taking responsibility.

I suppose I'm preaching to the choir if I spend more than a moment talking about the small physical and energy footprint of the mainframe, or how it can run consistently around 100% utilization rather than the 5 - 15% that is often typical of other platforms. Not to mention, of course, that mainframe staffing has amazing economies of scale - there is over four times as much mainframe computing power installed world wide compared to the beginning of this decade, yet apparently fewer people managing it all.

Still, let's not forget, for all its strengths, that it's you and I, the customer and the ISV, that take it to the next level.

After all, it's how you build and maintain your mainframe environment, taking advantage of solutions such as we offer, that enable it to continue to be more and more cost-effective.

For example, think of the word "Legacy." By now, we've all figured out that it's code for "it works." But there's more to it than that. There's decades of optimization, but there's also the opportunity to take advantage of new, even more effective features that new versions of our software offer (see my blog from this past December about Hidden Treasures).

Then there's getting the most out of your very valuable people's time - whether that be through tightly-focused business-result offerings such as CA's Out Tasking or our Mainframe Software Manager that handles so many of the otherwise obscure and time-consuming aspects of mainframe software maintenance for you. (See http://ca.com/mainframe2 for more about both of these.)

And there's the value of automation and integration of multiple solutions, working together behind the scenes to get results that bring maximum value to you with minimal manual involvement, such as CA OPS/MVS for automation and our performance solutions such as CA SYSVIEW and CA INSIGHT for DB2. Of course, low CPU usage is a significant advantage here too.

Finally, there's the fact that doing something right the first time saves vast amounts of time and money, so the quality of the mainframe, the management software and the dedicated people all come together to make a lean, highly-functional environment like none other.

I'd say you can't get any leaner than that - except that it looks like the mainframe will just keep getting leaner and more effective, setting a great example for the rest of IT and the business world to look up to. Who knows - they might all be leaning in our direction soon!

Published: April 23 2009, 12:08 PM | no comments
by Reg Harbeck

Well, I finally did it: I've now been to every continent where there are mainframes (unless there's one in Antarctica that they haven't told me about - if so, please let me know)!

And what a great experience. As I've mentioned previously, wherever I travel in the world I find some of the same things true about mainframers: intelligent, hard-working, concerned, interesting people, and the folks I met in Brazil certainly upheld this expectation.

The occasion: Chris O'Malley (CA's Mainframe Business Unit Executive Vice President), Gary Hundemer (Vice President of Mainframe Technical Sales) and I had the opportunity and honor to present to audiences in Brazil about what we're doing on the mainframe.

We began in São Paulo with a presentation to an impressively large audience, complete with simultaneous translation into Portuguese. After an opening and introduction by the local team, Chris started out with our Mainframe 2.0 strategy, then I talked about the technical aspects of Mainframe 2.0, then Gary Hundemer talked about achieving maximum value from your mainframe ISV investments, and I finished up with a presentation on regulatory compliance and the mainframe.

The audience and the local CA team were great, and we had good discussions afterwards over lunch. Chris was also interviewed by the local press about what we're doing on the mainframe. Then we finished the day with a meeting with an important CA mainframe customer (of course, "important" is implicit in "CA mainframe customer"), and flew to Brasilia.

The audience in Brasilia was even larger - these are some of the best mainframe audiences I've seen - and even more interactive. We gave the same presentations, got some great response, and had some very good discussion over lunch. Then we met with another mainframe customer, and the following morning we met with one more before I flew to Rio de Janeiro to meet with two more CA mainframe customers and discussed Mainframe 2.0 and what it means.

What a lovely country, what great (and hospitable) people, what a great local CA team, and what a great place to be a mainframer!

I just keep getting more and more optimistic about where the mainframe is and where it's going, and while there was general agreement that organizations have to start getting a new generation of mainframers in place, there was also great interest in the things we're doing with Mainframe 2.0 to help that new generation get effective right away.

It is so cool to be working with the platform that truly keeps the most important organizations on earth running (managing over 70% of business critical data) and to keep meeting the great people that keep that platform running wherever I go.

Obrigado (thank you) Brasil!

Published: April 22 2009, 01:06 PM | 1 Comment(s)
by Marcel Hartog

Mainframers,

Imagine you have 5 cars (I know... Just imagine..). The family car does 50+% of the work. Holidays, kids to school (or worse, pick them up after a party at 2 o'clock at night), shopping, garbage to the local pitt, things like that. They fit you and the spouse and the kids and the dog(s) easy, and there is even room for a portable BBQ. The other 4 cars are used for "other"...

At the end of the month, you compare the cost and find that the BIG car (the one that does 50+% of the work) eats away 30% of the car-budget. Does this mean that your BIG car is too expensive and that you should sell it? No!!! Funny enough, this is exactly what we have seen happening in the past years with our beloved Mainframe.

And now we have even more proof.... A recent CA survey in EMEA uncovered that, on average, 50% of all Mission Critical Applications reside on the Mainframe. For obvious reasons. The AVERAGE % of IT budget spent on Mainframe (according to this survey) is 19%!! But it gets better.... Of the companies who consider their mainframe "fully connected" with the rest of the IT Infrastructure, 64% of the Mission Critical Applications runs on the Mainframe.. And they spend a humble 24% of their IT budget on that same Mainframe...

Please do me a favor and read the previous paragraph again. Now read it backwards... I hope that the following conclusion is also sinking in: These companies spend 76% of their budget on an infrastructure that runs 36% of their Mission Critical Applications?!?!?

So what do we do with this data? Do we look at the Mainframer next to us and say:"Hmmm I knew that already... It's probably a bit less here, but not much..." and continue with that stupid CICS program that simply refuses to perform? Or do we DO something with it???

Chris O'Malley, CA's EVP for the Mainframe Business Unit told me not too long ago;"Never let a good recession go to waste..". And he is right.. Every customer I talk to these days is trying to save money. And suddenly, the arguments we heard in the past years for NOT using the mainframe are disappearing. People are open for new ideas again. This is the right time to arm yourself with data like the above and come up with new ideas... Make people aware of the REAL costs of the Mainframe. Use analogies like the one I used above. Maybe it's not the best one, but I have found that it works... Try to come up with short term savings, but also remind people that the reason that the (cheap) Mainframe is still there is because of thinking beyond the next 12 months. Something we all need to learn again....

I have picked only one item from the survey we did. It has more surprises... So if you can't wait, look for the survey on CA.COM/Mainframe. Or keep following this blog...  I promise you that it's more fun....

Keep your eye on the money. You are the one saving it.......

Published: April 09 2009, 09:22 AM | no comments
by Reg Harbeck

As anyone who's been around for 50 years would likely know (though I'm assuming here, as I'm not quite of that vintage myself), "L" is the Roman numeral for 50, as well as the last letter in "COBOL".  

Which is only relevant for one reason: the day after the mainframe turned 45 years old, COBOL turned 50!

That's right: according to Wikipedia (see http://en.wikipedia.org/wiki/Cobol), COBOL had its genesis on April 8, 1959.

Now, I have to admit, this isn't actually one of the pieces of trivia that I normally cart around ready to spring on unsuspecting colleagues. Rather, I was made aware of it thanks to the following article: http://www.guardian.co.uk/technology/2009/apr/09/cobol-internet-programming.

This is a significant anniversary in the ongoing history of computing. And, coming four years after the 50^th anniversary of SHARE and five years before the 50^th anniversary of the IBM System/360 mainframe, it is part of a time line that sheds light on the nature of business computing. First, the definitive user organization for high-end business computing was formed, then the original business-oriented programming language was created (COBOL, after all, stands for COmmon Business-Oriented Language), and finally, the ultimate business computer was born.

But here's a connection that many people don't seem to have caught. Having written both COBOL and IBM Mainframe Assembler, it is my observation that the two languages are like twins. It's like the same business principles that led to the creation of COBOL were also key factors in the creation of the System/360 architecture, and were expressed in how both handled data and processing.

Of course, COBOL and the Mainframe have been inseparable since the beginning, and while many other great languages have since been introduced to make the mainframe even easier to program (for example, CA Gen and CA Easytrieve®), there continue to be billions of lines of tried-and-proven COBOL keeping the world economy running (and great products such as CA InterTest^TM to keep the COBOL running).

So, here's to the real meaning of "Legacy" (i.e. "it works") and a half century of business-quality functionality: Happy 50^th Birthday, COBOL!

Published: April 07 2009, 04:28 PM | no comments
by Reg Harbeck

Can you believe it? The IBM mainframe turned 45 years old today! Reading the original announcement at http://www-03.ibm.com/ibm/history/exhibits/mainframe/mainframe_PR360.html, I'm impressed with how consistent the business-oriented approach, that made the mainframe strong, has remained through these past four-and-a-half decades.

I was exchanging email reminiscences about our mainframe careers with a CA colleague today, reflecting on this important anniversary, and he commented that this milestone made him sit back and realize how long he'd been doing this (he started in 1977). My reply was that, on the mainframe, he and I were still "young pups".

Of course, all that's about to change with the arrival of a new generation of mainframers, and the pending retirement of those who are a bit older than us "young pups" (as soon as their retirement savings recover, or, as is popular to observe in the U.S. these days, as soon as their 101K is a 401K again). I've even recently had the chance to get to know a number of new mainframers at CA, many of them in their early 20's, who are leading the charge to the future with everything from support to developing new Java-based interfaces.

It's going to feel odd, going from being the youngest mainframer for most of my career to suddenly being the oldest, with quite a few years left in my career. The good thing is, I may think of myself as being in the prime of life, but clearly the mainframe is even more so.

In fact, one of the cool things that those young mainframers have been developing is something we call "Mainframe 2.0". Chris O'Malley, the head of the Mainframe Business Unit at CA, did a webcast about it today (click here for a replay).

Mainframe 2.0 is CA's way of positioning the mainframe for a successful, business-oriented future, manageable by a new generation that won't have to inherit a rat's nest of obscure and obsolete configurations, maintenance practices and interfaces. That's important, as I pointed out in my z/Journal article called "Cleaning House for the Next Generation: Security Minus Obscurity" a while back, because what a new generation can't understand, they can't improve.

In any case, if there's one thing I've concluded during my career, it's that the mainframe will be around longer than I will. And I'm planning to be around for at least another 45 years.

So, happy 45^th birthday, IBM Mainframe, and here's to many more years of showing the business and IT world how to do it right!

Published: April 02 2009, 04:26 PM | 1 Comment(s)
by Reg Harbeck

Talk about TLA's - start talking regulatory compliance and you'll get TMI PDQ. Of course there's SOX, which refers to Sarbanes-Oxley. Then there's CIA, for Compliance Information Analysis, a feature of CA ACF2 and CA Top Secret in r12 and later.

But the compliance TLA that I've been most focused on recently is PCI, which is shorthand for two TLAs together: PCI DSS, or Payment Card Industry Data Security Standard.

It turns out that the average consumer is quite touchy about their personal information getting into the wrong hands, and the folks who issue payment cards, such as the credit card you likely use every day, have put together a set of regulations to respond to such concerns.

The cool thing is, on the mainframe, we've had solutions that can respond to these needs for a long time, and those solutions are growing and changing to respond even better (for example, the CIA feature I alluded to above).

So, when I had the opportunity to work with the folks at Swiss Colony Data Center, Inc. on a webcast about achieving PCI compliance on the mainframe, I couldn't refuse. It was a lot of fun, and I'm inclined to think that it was pretty informative too. I'd be interested in what you think of it - it's available at http://www.ca.com/us/events/item.aspx?e=199059&eis=1.

Once you've had a chance to view it (or even before if you wish) let me know your thoughts on regulatory compliance, PCI and the mainframe. I think this is a pretty important topic, and I'd value your insights on it.

Published: March 19 2009, 05:03 PM | no comments
by Reg Harbeck

Two weeks since SHARE and I'm still feeling the glow of information!

Strange, really, that it's taken two weeks to get to this one, as it's been on my mind since the System z Keynote at 10:30 AM Monday morning, March 2.

Moore's Law is over!

Well, not really: Intel's Gordon Moore originally (in 1965) was just observing that the number of integrated circuits that fit in a given amount of space doubled every two years (see http://en.wikipedia.org/wiki/Moore's_law). But since then, this observation has been dubbed Moore's Law, projected out into the future, and applied to everything from disk capacity to processor speed, and even changed from two years to eighteen months for the doubling effect to take place.

In any case, we have come to rely on Moore's Law to make computers exponentially smaller, faster, higher capacity and cheaper for most of the past few decades. So seeing even one of those trends "hit the wall" can be a bit of a shock.

As it turns out, the one it happened to first was processor speed.

According to IBM at their System z Keynote, processor speed has hit a ceiling of roughly 5 GHz, due to the sheer limitations of physics.

Before I go any further, I have to admit that this is a bit embarrassing for me: I'd written an article recently about what we were going to do when we reached the limits of Moore's Law, and didn't even realize that, in this one area, it had already happened (see http://www.cmg.org/measureit/issues/mit54/m_54_11.html). I mean, how would you feel about predicting something and then finding out it had already taken place?

To be clear: it is theoretically possible that this ceiling will be lifted. I've heard that IBM has a CPU running at 500 GHz - but it's bathed in liquid helium, a rather impractical coolant for most computing environments.

For now, however, we get to find out what happens when the apparently unlimited becomes limited. And I certainly have some opinions on the matter.

The first opinion is, welcome back to the concept of scrupulous computing, and the mainframe that embodies it! While other platforms have clogged their systems with more and more software demands, so even the most advanced seem to take forever to boot, the mainframe has continually been optimized for greater and greater performance.

As I understand it, distributed (i.e. non-mainframe) machines tend to run an average of 5-10% busy, and can't handle being much more than 30% busy at the best of times. That may seem reasonable if you can just keep getting faster and faster CPU's, but what do you do when you've hit a hard limit?

The first answer, of course, is to go parallel. But, as IBM also pointed out in their keynote, there are also diminishing returns in that direction. Get enough parallel CPU's and your performance actually starts to drop. Of course, IBM made it clear that they're working very hard to push that limit back, but it's a real limitation regardless.

So, the fact that a mainframe can run at roughly 100% busy in a sustained fashion suddenly looks even more attractive. If you can't add speed, you certainly want to take maximum advantage of what speed you have available!

Of course that also implies another thing I think this entails: greater optimization is going to be necessary. In fact, I'm quite curious to see whether there's a resurgence of Assembler (or at least "C") programming as people look to squeeze every last bit of performance out of their CPU's.

Two other things I'll be watching for the impact they receive are: emulation and virtualization.

We've been doing both of those for a long time, but they're getting a lot of attention these days.

Emulation is the one I expect to be hit hardest. Why use up significant amounts of one architecture's power just so you can pretend to be another architecture? When there's not enough power to go around, you want to run as close to the bare metal as possible.

I have more hope for virtualization. IBM has demonstrated that it's possible to do virtualization very well and very close to the metal - so much so that you can't really run a mainframe without some virtualization; it's optimized for it.

However, whether you're talking cloud computing or PC virtualization or any other way of introducing a layer abstraction between yourself and the hardware, I think we're going to see a real focus on getting closer and closer to the absolute minimum of insulating fluff (that's a technical term  :-) ).

Strangely, whenever I blog about some emergent opportunity or challenge, I always seem to end up here: I'm glad to work on the mainframe, the one platform that has the best prospects for dealing with this effectively!

Published: March 16 2009, 09:53 AM | no comments
by Reg Harbeck

Talk about an epiphany! There I was at one of the sessions I attended at SHARE in Austin, when someone mentioned something that hadn't really occurred to me yet, and it completely shifted my perspective.

Then, later that same week, someone else mentioned something similar - possibly something I'd heard before but had never connected the dots about.

So, later, I mentioned it in a presentation I gave and got general nods of agreement. Apparently, it's been obvious all along, but no one's been saying anything about it. And here's what it is:

You know all those new "distributed" applications in your organization that have been developed in preference to mainframe ones? Many of them are really mainframe applications that just have some distributed processing and distributed front-ends!

In other words, if you were to turn off your mainframe today, many - possibly most - of your critical distributed applications would stop working because they rely on the mainframe for so much of their important data and processing!

The story that triggered this for me was about a shop that had been exclusively developing "all distributed" applications for so long that management didn't see the value of their mainframe. However, one weekend, there was a bit of an issue with an upgrade on the mainframe that had to be backed out, so it was unavailable for an hour or two during business hours, and suddenly, critical distributed applications stopped working!

It turned out that they were more reliant on the mainframe than ever!

Check it out in your shop and see if it's not the same case: how many of your most critical "distributed" applications have no mainframe interaction at all? Now, how many are so reliant on the mainframe that they'd cease to function if the mainframe were unavailable?

Of course, the reason why no one seems to think of this is because the squeaky wheel gets the grease. I mean, the rarest part of that story is that the mainframe was unavailable - that almost never happens! As a result, nobody notices that their distributed applications are completely reliant on the mainframe... because it's completely reliable!

For me as a mainframer, that's exciting. But for me as a responsible business person, that's scary, because it means that IT management around the world may be uninformed about the fact that the mainframe is still the goose that lays their golden eggs, so they may be reducing their investment in it and even thinking they can move off of it, when in reality, if they did, their goose would be cooked!

Call to action time: what are you doing to make sure your management is aware of how important your mainframe is to the ongoing survival of your business? Since the mainframe is the opposite of a squeaky wheel, someone needs to squawk on its behalf in order to pre-empt some seriously bad decisions!

Published: March 13 2009, 11:41 AM | 2 Comment(s)
by Reg Harbeck

I just realized I should have included a link to the following cartoon from http://ca.com/knowhow to complement yesterday's blog:

http://www.ca.com/us/content/page.aspx?cid=199714