Wednesday, December 3, 2014

Niches open and niches close and yet the versatile NonStop prevails!

A chance posting to Facebook had me looking at the attributes of NonStop and whether you argue for availability or perhaps scalability, in the end, just having the discussion suggests NonStop is still very much in the game …

Versatility! How many times have we heard a sales pitch that highlights the versatility of a product? In many ways, this is an approach to mitigating potential future obsolescence and it extends the product’s useful life.  Flipping through pages about NonStop on the web, I came upon a reminder of days past when NonStop was considered the Swiss Army knife of computing – versatile in its own right. This knife featured in a First-Friday skit by Jimmy Treybig that just came to my attention following a post on Facebook by long-time Tandem supporter, Maria Olivero. However, the original fault tolerant message has been replaced with new messages on availability, scalability, data integrity, security, networking, manageability and so on. So what really does propel NonStop to the fore in the twenty-first century? 

Looking around at the NonStop users present at the recent NonStop Technical Boot Camp it was pleasing to see a lot of familiar faces. Once again, MasterCard took home the NonStop Availability Award – a reminder too of the time I spent in Product Marketing as part of the NSA team under Dr. Tim Chou. However, this time MasterCard went home with a perpetual trophy as, after 20 years, the program has come to an end and it seemed only fitting that MasterCard be rewarded in this fashion having won the award more times than anyone else by my count. Seems that when it comes to availability, too many corporations are loath to talk about the levels of uptime they achieve and view the stats they may have as being strategic to the business. Even so, while some big name customers might be reluctant to talk about what they’re achieving, as far as SLAs go, I think that the bigger issue is talking about the processes that they have in place in order to achieve those SLAs. Wouldn't it be cool if we could get that discussion out into public forums! NonStop today may be about a lot more than availability but its importance to business shouldn’t ever be discounted. 

Measuring NonStop availability leads me into the subject of this post – there have always been markets where the attributes of NonStop have been greatly prized. Sometimes considered just a niche or a submarket, nevertheless, the expectations of what NonStop could provide propelled it to a place of preeminence. When it comes to niches, NonStop has a rich history of being able to dominate. Over the 40 years of NonStop systems being used, niches have proved to come and go with the fortunes of NonStop closely tied to each niche – it’s just good fortune smiling on the NonStop community that NonStop found one of the longest-lived niches in finance, for instance. For anyone who has followed postings to this blog, my fervent belief in niches, in tapping cool applications and in NonStop being at the heart of some very hard-to-do processes, have been hard to miss.

Across the NonStop community and from customers and vendors presence at numerous user group events it’s well known that German manufacturing giant, Mercedes Benz, depends upon NonStop systems at all of their plants and that these systems oversee manufacturing. In a recent editorial published in the January 4, 2015, issue of Motor Trend, Angus MacKenzie comments on how today, “If there’s a niche, fill it. And if there isn’t a niche, create a new one. That’s the mantra at Mercedes Benz these days.” I can’t think of any other way to express where I see NonStop heading than what Mercedes Benz states here and while that may not sit well with everyone in the community, there are few other viable options for NonStop if we want to talk about the next 40 years.

When you talk to the team at WebAction about their inclusion of NonStop as part of their Data Drive Apps program, the story is very much about a niche. The only reason the executives at WebAction were keen to keep NonStop in their program is that key mission-critical applications run on NonStop and as their goal is to “build and deploy real-time data driven Apps in days, not months or years” it made little sense to ignore the NonStop marketplace. WebAction, Inc. Cofounder, Sami Akbay, summed it up best last year when he said, “Niche means distinct, specific, and distinguishable. Niche doesn't have to mean small or dying. When it comes to the bigger picture of computers worldwide – then yes, NonStop is a niche.”

Along similar lines, it was OmniPayments, Inc. Yash Kapadia who told me that, working as long as he has on NonStop, that “There are times when we do struggle hard to make NonStop sound like a normal system and try our best not to call it a niche. However, when it comes to the bigger picture of computers worldwide – then yes, NonStop is a niche.” Both Akbay and Yash agree that once you appreciate that niches are indeed sub-markets, then when it comes to a sub-market as big as payments, NonStop is the predominant player.

However, is it the availability or the scalability value proposition that today fosters renewed viability for NonStop? And does it even matter if NonStop can continue to carve out niches best suited to its superior ability to support mission-critical applications? In the final analysis it still boils down to cost and perhaps, “cost – scalability”. Or, should that be “low – price availability”? According to DataExpress President, Michelle Marost “scalability to me means being able to start small, and grow.”

Marost than acknowledges that as of right now, this “is not evident in the NonStop environments, both hardware and OS, and until they are made more affordable by HP they won’t be considered ‘cost-scalable’. Like many within the NonStop vendor community watching for more news on the NonStop X product family, to exploit a niche or submarket, the price for newer models in the NonStop family have to be affordable to more submarkets and any improved viability of NonStop will likely be snuffed out, according to Marost, should there be any “elimination of a small (i.e. cheap) starting point”.

Scale-out and even, scale-up; have these attributes begun surpassing the availability story in terms of importance for the enterprise? “Guess what, we can scale to the wazoo and do you know what else? It doesn’t ever fail … cool, eh?” Will conversations like this become more prevalent in the future even among NonStop users? Just as importantly, are there even more niches for NonStop to occupy in terms of scalability than there are of availability? Is this the future for NonStop and if indeed scalability attracts the spotlight, will we likely see a greater presence of NonStop in the future? Possibly, even inside enterprise clouds?

There’s much anticipation over NonStop X and its price points just as there is over the arrival of a more industry-standard packaging of NonStop. No matter what transpires over the next few months – hopefully, a situation better clarified at HP Discover, Barcelona – NonStop is firmly entrenched in the twenty-first century with many more anniversaries to follow. Availability! Scalability! Indeed, demonstrable versatility! Pick your favorite attribute as you may want to do but in the end, just having discussions like this suggests NonStop systems remain a viable option and even as we read of companies looking at alternatives, we also hear about those that elect to stay with NonStop. And to look for new ways to leverage NonStop! 

2 comments:

Scott Healy said...

Given that a good many new applications start out on a commodity Linux server (or desktop/laptop) cost/scalability is a bit of a conundrum. NonStop will likely not come out with a product that competes with Linux on a laptop, but should definitely seek to make the migration from there to a fault tolerant, scalable NonStop easier.

Stephan Amsbary said...

Excellent article Richard! While NonStop continues to survive in several niche markets, Microsoft continues to convince high -availability and -scaleability COTS offerngs to migrate their applications. The current belief that "just good enough" is the state of the art is replacing the experience Tandem/NonStop and other vendors have worked so hard to develop.

I am now in the utility sector (CTO) and the mission critical applications I require (and am purchasing) are no longer available in UNIX, Linux, zOS or of course NonStop.

To further Scott's comment, the "new" applications developed today are being built with a (Microsoft) PC mindset. Hopefully the next generation NonStop will offer not just a good entry point, but also bridging tools and quick-start programs to exploit native NonStop attributes.