Thursday, December 26, 2019

For the NonStop community it’s time to reflect.

With the year rapidly winding down there is much for the NonStop community to consider going into 2020 …

The traditional Christmas dinner is over and for some it’s been a time to revel in even more traditional Boxing Day fare. For Margo and me it’s been a time where our thoughts have drifted back to where this time last year we were busy celebrating the season in Sydney. When it comes to traditions then yes, it was a feast that included grilled sausages (on the nearby BBQ) together with Pavlova for dessert. Sorry, my Kiwi friends, Pavlova is an Aussie creation / tradition. On the other hand, so too is watching a cricket test match beamed live from Melbourne, even as locally all eyes are on Sydney Harbor for the start of the annual blue water sailing classic, The Sydney to Hobart yacht race.

Here in Colorado it’s been more a case of celebrating Christmas Eve with the family where Polish traditions have dominated the conversations. And the food! There will be deviled eggs along with special Polish cakes and much more. Our grandkids are now 6 and 4 and continue to develop right in front of our eyes – after a troubling time for the twin boys following birth routines have been established and they seem to be making good progress. There is nothing quite like ripping the paper off presents to set a positive upbeat tone for any gathering. And did I mention, laughter?

For the HPE NonStop global community the holiday season as such takes many forms. It would be hard for me to catalogue them all so suffice to say, in whichever language or custom it takes, Margo and I wish you all the happiest of times as we head towards a New Year.  When it comes to 2020 there are plenty of commentaries being shared highlighting that in years like 2020 we do think about how far we have come. Automotive executive Bob Lutz just wrote of how:

“There's something about a year with a zero at the end that gets our attention. It seems more important. It makes us recall the last "zero" year, causing us to pause and reflect—not only on our own lives (slipping by ever faster), but also on society, politics, the world.”

Lutz also wrote that:

“It's been quite a decade, marked by arguably more change than in the previous five decades put together. I'd like to highlight some of the more notable ones, whether they're caused by creative engineers, manipulative marketers, government regulations, fickle buyer preferences, or, in some cases, a complex blend of several initiating factors.”

Indeed! A year ending with a zero sure does get our attention even as the changes we all witnessed were caused by creative engineers, manipulative marketers, etc. or simply by a complex blend of several initiating factors! I am pretty sure that this rings true for many of us. Not everything that came to fruition was by design nor did everything that was designed resonate with everyone. Where’s my flying car? Where’s my robot housekeeper? Ooops, that’s right – we are almost there on both counts, I have been informed. The sad thing about making longer term predictions is that they have a habit of coming about in ways not imagined. Who would have imagined that I could order paper towels and have them delivered the same day?

A year ending in zero reminds me too that it was back in 1970 that I turned up at Wollongong’s steelworks (John Lysaghts and not nearby BHP), to start a two year cadetship in computing. Yes, back then Australia’s universities did not consider computing an academic pursuit so to be part of this new industry you had to become an apprentice. I had sat for the apprenticeship qualifying tests back in the summer of ’69 – that’s the Australian summer of ’69 but that’s a whole different story – even though my interest in computing dated back to ’65. That was when my father, Roy Buckle, was responsible for bringing to Australia’s shores a computerized Mergenthaler “line-o-film” typesetting system and the first such system to be setup in Australia. The data entry keyboard for such systems is depicted above where the photo could just as easily have been taken at my dad’s place of work.
As I look back at that apprenticeship, we were taught IBM 360 Assembler in excruciating detail – our online system in support of the different steelwork machinery (from cold reduction lines to pickle and plating lines) only accommodated programs that could fit into 4K of memory. Yes, 4K. And I am sure that this will bring back memories about the early days of computing for many members of the NonStop community looking at the improvements in productivity over the past decades! At the time, I was given the task of writing the disk access code – as yet we didn’t have enough memory to support logical I/O; it was all at the physical / channel level where sorts could be executed directly within the disk channel processor.

For some reason, I recall the concept of channel command block / channel command words (CCBs /CCWs) and in particular, the Transfer-In-Channel (TIC) command. And I am not sure why that is! Even so, here we are writing about DevOps and the breakthroughs we are seeing with Open APIs and technologies like REST and JSON. Wow! So likewise, this has taken five decades to reach this point, but productivity today is incomparable to anything we witnessed in the very early days of IT. And yet, here we are. Whereas I was learning 360 Assembler way back then today it’s more the case of assembling a process / program from libraries of code that are then used to flesh out a basic code stub pulled from somewhere else. And I thought it really cool to have an approved global macro accepted by IT and to have it saved to disk for other programmers to access.

Fickle buyer preferences? Have to believe that without the push to accelerate the way we develop code we wouldn’t have key social media channels that dominate today’s conversations. Just think what Google has brought to the table or before Google, think of what Netscape achieved? In many ways I am glad to report that I never made it to where programming was all about the manipulation of objects, and more, or where a terminal was anything other than a simple keyboard. On the other hand, it would be unreasonable to expect an industry so forward thinking as IT would have ever stood still over any prolonged period. Yes, those fickle buyers of computers always wanted something better, faster and naturally enough, cheaper.

For the NonStop community there has been so much talk this past decade about the need to train more programmers on NonStop. Just as importantly, there has been equally as much talk among CIOs about the difficulties in finding qualified NonStop staff. Unfortunately, both discussions missed the point – with the steps that the NonStop team has taken this past decade, you shouldn’t be too concerned about either, but more so, leverage today’s modern development environments and platforms to build applications for NonStop. Again, deployment shouldn’t be confused with development and with the options available today for development teams to choose from, there shouldn’t be any need to look outside the skillsets already on-hand inside any IT organization. The above slide was used recently by HPE’s own Franz Koenig and it has already began turning heads and in so doing, has begun framing the discussion for the coming decade – why not NonStop?

I may not be able to describe in any depth the majority of tools and utilities listed on the above slide. But I do know where to go to obtain more information should I need to do so – talking to the NonStop vendor community and almost all of them are leveraging one or more of the products included on the PowerPoint slide. What stands out most of all is that we are no longer thinking in terms of bringing programmers to NonStop but rather, quite the opposite. We can now talk openly of bringing NonStop to any programmer and thereby leveraging the current skills of all programmers. Key point? Program today’s business logic and have the deployment option inherit NonStop attributes transparently!

This decade beginning with a zero will prove to be a fascinating decade. Starting with 2020 we will see even more discussion about clouds and service models with many new ways to consume technology. However, by the end of this decade, when it comes time to write about expectations of those entering 2030, will the cloud be even part of the discussion? Will traditional computing models even be relevant? Will we even have to pay for a chip or worry about memory – 4K bytes, or even 4Y bytes (Yottabyte) it won’t matter; the cost will essentially be the same.  What then? Will we be throwing out ideas to an assistant buried deep within our car whatever that might be and see programs come to fruition automatically?

We may yet be still waiting on our flying cars but suggesting ideas and having them transformed into programs well, probably not too far away. Let one more zero show up on the calendar – I am ready for it, are you? And with that, Margo and I wish you all the very best for the coming year and yes, the coming decade!

Wednesday, December 18, 2019

Top of the mountain? Wide valleys look even better …

The niche NonStop occupies is about to widen to include all mission critical scenarios – and it’s all because critical CIO concerns have been addressed …

Living alongside Colorado’s Rocky Mountains certainly has its moments. There are times when it’s all about rapidly changing weather conditions even as there are times when the view to the west can be awe-inspiring. Driving across the top of the Rockies is always a wonderful distraction from the daily routine and to cross the continental divide well, spectacular is the most descriptive word that comes to mind. Having driven the Peak-to-Peak Highway many times and traversed the length of Trail Ridge Road almost as often, Margo and I never shy away from the prospect of driving it one more time.

Estes Park anchors both drives – one heading south along Colorado Highway 72 while the other heads west, up and across Rocky Mountain National Park along US Highway 34 – but choosing either way highlights how clinging to a ribbon of road may have many advantages when it comes to the views provided, but leaves few opportunities to accommodate much greater traffic flows. There is much to be said about the benefits of the wide open spaces. When it comes to the landscapes that are present between mountain ranges it is apparent to us that perhaps what’s in these broad valleys offer a more diverse vista – with its wide open prairies to the east of the Rockies or the high plains to the west, driving along a sinuous ribbon of road might just be the exception to the rule!

In conversations with clients this week, the common thread turned out to be the need to know more about NonStop penetrating markets beyond financial services. Almost from the very beginning of NonStop’s presence in IT, NonStop has been all about front-ending networks of ATMs and POS devices with some presence in branch banking and more recently in support of mobile devices used in payments. It has been a lengthy journey, skimming the tops of peaks as it were than it has been a case of enjoying popularity across a broad and clearly diverse landscape. Occupying niches to the point of dominance has its upside but only insofar as it can be turned into a reference base for finding even more niches where NonStop’s value proposition can be leveraged.

At face value, supporting mission critical applications suggests something more than a very focused outlook that is so specialized that few in IT are even aware of its presence – no, step outside of the financial services marketplace and you will only get puzzled looks should you suggest your CIO consider NonStop as a solution. Of course, every member of the NonStop vendor community would like to see NonStop’s presence become more widespread as this is in their interests to see increased diversity. However, it’s not too hard to imagine that there are many more markets for NonStop than simply financial services even as there are sizable niches that once identified could provide many opportunities for NonStop. Is it our lack of imagination holding us back? Is it messaging that’s missing the mark? Or, is it something more basic to which we need to pay more attention and if that is the case, can such a basic need be one that can be overcome?

For as many years as I have been associated with NonStop here have been many discussions about the need to port this or that application to NonStop. If only we could provide a modern healthcare solution or even an airline reservation application or perhaps take a couple of baby steps that see us building on partial success in manufacturing and transportation. Why isn’t NonStop a major force in the world of container shipping where having documentation available, 24 x 7 is a must! With all of these observations the talks turn on how best to pull together a new application and present a working model in an abbreviated timeframe. Isn’t it possible to do something along these lines without pursuing a training program of staff needed for such a project?

NonStop has come a long way over the past decade – NonStop SQL now is as compatible with major SQL offerings like Oracle than at any other time in its history. When it comes to Java so much work has been done to simplify ports to NonStop and now, with the commitment coming from NonStop product management to support Kernel Level Threading (KLT) in a way that’s compatible with the NonStop architecture, 2020 clearly has the potential to be a breakthrough year for any IT organization looking to port a Java / Oracle solution to NonStop. But having said that is it enough? And more importantly, is this taking NonStop down from the mountain tops and into the broader valleys where opportunities abound for a platform that is fault tolerant and scalable?

When it comes to the conversations I have had over the past week, what has stood out is that there is a growing awareness that the new NonStop may indeed be compatible with all other systems you can name that can be found in any data center. Compatible, that is in terms of ease of development using standard development platforms. There is now a consistent message of development versus deployment resonating within IT – a message taken up by NonStop product management. And for good reason – the NonStop no longer needs anything special in order to develop a new solution. In fact, perhaps thoughts of porting an application may prove to be old hat and a reflection of what was done in the past. Let’s look at rapid development of new applications on NonStop that can be as easily done as for any other platform. Bigger yet, it can be done leveraging the skill-sets on hand as possessed by the latest college graduate!

Tall order? Not possible? There are many questions that arise from observations like those described above but does that necessarily mean that it’s wrong to think this way? When Franz Koenig, head of NonStop Advanced Technology Center (ATC), gave the presentation “NonStop – The Art of the Possible” at this year’s NonStop Technical Boot Camp (TBC) he included a reference to what the NonStop team is hearing from CIOs. Among those listed there were two standouts that caught my attention. “Leverage common technologies and standards” together with a call for “High productivity of staff. No steep learning curve.”

Considered as challenges to the NonStop team, it led to a short list of answers (to the CIO challenge) that highlighted just how far NonStop has come in addressing the needs of application developers. Again, two answers provided stood out. “Build new services like on other platforms” and “Deploying on NonStop gives your services automatic scalability and fault tolerance!” While Franz had more to say on this topic the point he drove home in no uncertain terms was that any barriers to developing new solutions on NonStop are being dismantled to the point where, as I wrote in a client newsletter recently, almost anyone can program today’s NonStop. DevOps ready? Wonderful! Applied to more than financial institutions and retailers? Terrific!

Knowing more about applications available on NonStop as its appeal widens to include niches beyond what are considered traditional NonStop niches may prove to be the wrong question to be asking. Rather, greater penetration into all mission critical environments may be realized more quickly than we thought if it becomes common knowledge that everyone can program for NonStop! The evangelism coming from the NonStop team is  admirable and a pursuit that has been percolating for some time is now coming to the surface. Those executives present at TBC 2019 have gone away with many questions on their minds – why didn’t we know more about NonStop’s answers to our challenges? I suspect that in 2020, there will likely be even more attention given to just how easy it has become to develop applications on NonStop.

Margo and I will continue to drive the peaks of the Rockies as we explore new ways to cross the continental divide. The views to the valleys below will continue to be spectacular and yet, what opens up in the wide valleys below will be every bit as spectacular as what is found high up on the narrow ridges. Anyone who has ascended Utah’s Highway 12 Scenic Byway and driven along the narrow confines of its Devils Backbone will appreciate that while the road itself is a marvel what opens up to the east is the amazing Capital Reef National Park. A reminder to all drivers that there is a lot more to take in than a narrow strip of bitumen, given so much beauty on display beneath the pavement! And for NonStop, being part of that much wider niche – all mission critical environments – is now within reach! Taking in this much wider vista will likely be the highlight of 2020; yes NonStop can do it all!   

Tuesday, December 10, 2019

Clouds don’t bother me, anyway …

Whenever a technology, product or solution appears to be the only answer then it’s time to look beyond it …

At this time of year we can expect to be the victims of weather systems that Mother Nature cares to throw at us. Living alongside the Rocky Mountains simply means it can be cold and windy and the skies can be dark and cloudy. On the flip side, I don’t have to venture too far from my desk and given how I have just finished a major upgrade to my systems – a new HPE Pavilion laptop connected to multiple monitors, one of which is vertical and is the one I am using right now – it’s all still at the fun stage. Of course, the fun factor will diminish in time as patterns set in and I realize that it is still all work. 

This week it’s been all about publishing. The December 2019 issue of NonStop Insider has been the priority and given how close we are to the holiday season Margo an I have been pleased with the responses we have had to date from the NonStop vendor community; should be another good issue as it not only covers what transpired at the recent NonStop Technical Boot Camp (TBC) 2019 but expectations for 2020 as well. It’s been quite a while since I last was able to say that these expectations all center on a year devoted to upgrades as the NonStop X finds its way into every enterprise data center.

At this time of year too we rarely see the night sky as the snow keeps falling. It is a stark reminder that winter is fast approaching and even as there was a time when we looked forward to the first snows of winter, we are now very much over it. When will the clouds clear and when will we be bathed in sunlight again? In the mid-1970s I lived in Edmonton, Alberta, and it was around this time of year when the sun didn’t make an appearance much before 11:00am (to vanish some four or five hours later) and I am reminded of those times when I see darkness descend well before dinner time. However, there is more to this than just clouds, darkness and cold; whereas neighbors are busily placing skis on roof racks while others are decorating Christmas trees and installing lights outside of their homes, there are those dreaming of warmer days spent closer to the equator. Then again, no matter our preferences we cannot escape looking up at the sky only to see more clouds.

It wouldn’t be a blog post of any significance these days if it didn’t include references to clouds. It’s been hard to ignore the ongoing battle between tech giants Amazon and Microsoft over providing cloud services to the federal government expected to generate US$10 billion in revenues over a ten year period but then again, there never has been a big contract that didn’t create animosity amongst the bidders. Indeed, HPE’s acquisition of Cray had some connection to the big High Performance Computing deal with the government that HPE lost, according to some reports I have read. But that really isn’t the big story here; what’s been captivating has been emerging commentary on the post-cloud era. What are we doing to prepare for a time when clouds become well, last year’s news?

There is a tendency in our industry to fall head-over-heels in love with the latest technology offerings only to find a few years later that their significance was fleeting at best; distracting and costly at worst. For many this isn’t the case with cloud computing but then again, are we so sure of this? Are we still prepared to make the big investments in something that really doesn’t change the parameters all that much – the cheese still has to be pushed closer to the goal line, but does a cloud contribute to any forward movement to this end?

No, clouds don’t bother me. They are just the latest trend to capture our imagination and to introduce us to language that is well, kind of new. I have often written about the progress of technology being much like the tide - waves advancing and then retreating in a continuous march up the beach. Where clouds do unsettle me more than a little is when I read accounts of many clouds being relied upon to the extent where the complexity that arises masks any advancement in meeting critical business objectives. Cloud computing of itself is just another model that taken to the extreme can become so complex that whatever savings might have been anticipated have been squandered trying to oversee it all!

From the very first presentation on clouds that I saw, the word that stuck with me was “elasticity of provisioning.” This is an attribute I can warm to, winter weather notwithstanding. In any given data center there is more compute power on the floor than is typically required to meet the expectations of business, but are present in order to meet the needs of “Black Friday!” Wonderful; but there have been other options for many years – outsource to service providers to meet peak demands. Yes, traditional service bureaus have proved adept at meeting peak requirements for many businesses.

And yet, elasticity of provisioning is far more universal than we may think – everywhere we have compute power there are advantages of being able to assign resources on the fly. To this end, cloud computing shouldn’t be considered an end game but rather an attribute of all systems. What we see today being promoted by cloud services’ providers should be “standard equipment” of any new operating system! Whether native or via virtualization or the presence of containers, the disconnect between what’s real and what’s needed should be transparent. And the flexibility inherent with this model doesn’t mandate a separation of hardware from software – no; it’s likely to all move into the hardware freeing up resources to run the applications business deems necessary to staying in business.

In the post cloud era – or, dare I say (repeating what HPE’s CEO Antonio Neri spoke of this year), the cloudless era – this elasticity of provisioning we associate with clouds will become so ubiquitous that to purchase any new platform or system that doesn’t inherently support such a property would be very shortsighted. In the post cloud era we are heading towards, the trend will be to on-prem the big stuff and perhaps engage with others for the secondary stuff. But what about the OpEx versus CapEx arguments – surely, this is an unstoppable force with respect to how business sees future investments in technology? 

Unfortunately, governments worldwide aren’t stupid. Well not entirely. When every enterprise has moved to the “everything-as-a-service” model and shifted to reliance on external cloud service providers such a universal shift will go unnoticed? Governments have a history of changing models so would we be prepared to go along with all of this if there was no difference to accounting between OpEx and Cap Ex. How brilliant will we all look if with the stroke of a pen, it was all changed so governments could still count on revenues as they have done for eons? Just saying! No, business should never build a case around a current anomaly as these can disappear overnight.

Should clouds be ubiquitous and there’s no tax benefit from OpEx over CapEx, whereto then? No, clouds don’t bother me as already we are counting down the days to when the hype simply fades away and we are onto the next big thing! And what would that be exactly – well, for starters, the service model that is evolving in front of our eyes where everything we ever want is only hours away from being on our doorstep, is going to continue accelerating to give us new industry models. We talk about the edge but what happens when the Edge isn’t an access point but is us! We are the edge … ten billion people or thereabouts all hitting the enter key at once.

No, clouds don’t bother me even as the incoming tide matters little. History has taught us that when something has overwhelming support, the next thing is about to appear. In all that you are doing with respect to technology are you preparing for a world beyond clouds? Shouldn’t you be concerned or at the very least, be more than a little bit curious? For me it’s simply more stories to write and more interesting discussions to be held. More important perhaps and indeed more relevant to what will likely transpire in 2020 – don’t think for one moment that this is the last big thing we have to sign for and deploy. There is so much more to come – are you looking forward to that day?

Looks can be deceiving! HPE NonStop; when being the best still matters!

For the NonStop community, we know what looks good may not only be deceptive but borderline dangerous; mission critical applications are bes...