Tuesday, October 25, 2016

Oh Canada! NonStop RUG, CTUG, enjoys a packed house.

Every successful system and platform attracts an ecosystem; whether it’s groups of users, vendors, service providers and even media channels all of them provide a vital “service” and NonStop is seeing it’s ecosystem continue to grow …

As readers of the NonStop community blog, Real Time View, know all too well, when opportunity to drive somewhere to meet with the NonStop community presents itself and there’s time on the calendar to make the tip possible, you can count on Margo and me to put in the hours behind the wheel to get there. And the drive to Mississauga, Ontario is always one of the longer trips on our calendar we do our very best to make sure we make it. Last year was a little bit of a let-down as for one reason or another we attended MATUG outside Washington D.C. but found we had no time to drive to Canada. Oh dear …

But logistics aside, we missed listening to Mr. T – Jimmy Treybig – and I suspect further opportunities to hear from Jimmy will lessen in time so yes, from that perspective, electing to drive can be sometimes problematic. As for this year, choosing CTUG meant we couldn’t make it to NENUG that this year represented a combination of community members from New England and New York. On the other hand, being a participant at this year’s CTUG event was reward enough to make the time behind the wheel enjoyable.

The HPE NonStop development team is moving fast. This has been the message coming from its leader, Andy Bergholz, whenever he has the opportunity to address a large gathering of NonStop users. It’s not something that Andy simply throws out there on a whim but rather, truly reflects all that is taking place in support of NonStop. To paraphrase a popular commercial, “You don’t know NonStop” if you have missed any recent NonStop events as it’s front and center in the march to hybrid infrastructure with converged systems. And it isn’t stopping with this as at CTUG we were briefed on much more that is in store for NonStop that continued to surprise even the hardiest of NonStop supporters.

The good news is that from my perspective, I am seeing the team from NonStop development working more closely with the NonStop vendor community. When one question was raised late in the day about the potential for NonStop development to develop a capability, the response was unequivocal. With three competitive products from the NonStop community already in the market, this wasn’t a priority for NonStop development – there would be no competing with NonStop vendors on this occasion.

Whenever business looks at technology particularly with new product initiatives be they a major solution from a well-established vendor or simply the latest project coming out of the open source initiatives, these businesses want to see a thriving ecosystem actively involved with the technology or product. Today, NonStop may not attract the types of major vendors that are better known within the media but that isn’t to say, the NonStop vendors active in the NonStop market are any less important.

Again, personally, I would like to see many more vendors in the NonStop marketplace but the core group of NonStop vendors that are committed to NonStop ensure that there is little that NonStop systems cannot do today – and when the discussion turns to issues of modernization, staffing and integration with what surrounds NonStop today, all the pieces that business would need are present. When another question was raised about NonStop doing more to bring down the costs of NonStop the response from NonStop development was immediate.

“Have you really looked at the costs to run NonStop compared to a Linux solution?” When you look at the tab to run a reasonably sized cluster of Linux and throw in the utilities, databases, replication, etc. licenses and then look at the associated support costs, the true value of HPE providing an integrated stack in addition to the NonStop operating system, most businesses would be shocked to see the bottom line – NonStop is more than competitive with the best of the open source offerings.

Several years ago I did a study on the true cost of ownership (TCO) of NonStop and compared it to both Windows and Linux offerings. The report I produced can still be found on the HPE web site and has been referenced in a number of NonStop presentations and it didn’t mince words. If you aren’t aware of this research note, then check out
NonStop offers the lowest TCO  in its class for complex  mission-critical applications And this was before the work that led to the release of the latest family of NonStop systems, the NonStop X. Point is, HPE has done a tremendous amount of work that has seen the TCO for NonStop drop to its lowest level all time. However, it’s not just the costs for a NonStop system itself that needs to be taken into consideration but the ecosystem of NonStop vendors.

What is often overlooked is that within this ecosystem, questions about configurations, APIs, integration at the file and message level, etc. can all be quickly and easily addressed with a single call or email. And more often than not, for free. Going one step further, raising a question on any one of multiple LinkedIn groups focused on NonStop and there will be multiple responses produced in barely a heartbeat – that’s what the NonStop community does best today. NonStop has a thriving ecosystem and it’s not just a group of product companies but increasingly today, it includes managed service providers as well.

Not sure how to manage a new NonStop X deployment – then call in any one of these managed service providers and they will be quick to set your mind at ease. Scattered globally there is bound to be a couple of them that serve your geography and increasingly, your industry. And don’t forget too, several of these managed service providers can actually do all the work necessary to order and then commission a brand new NonStop system. When I was at BITUG earlier this year and sat down with TCM CEO, Tony Craig, I was really impressed even as I was taken aback by just how comprehensive managed service offerings have become.

My earlier reference to a research note I wrote on NonStop also touches on the subject of publications and commentaries. This month, a new monthly publication has been launched and it is one that has Margo and I actively involved. Indeed, Margo is the Managing Editor of the publication, the NonStop Insider. By now, you probably have been emailed with the links but in case this is news to you, then read the
October issue and let us know what you think.

This is just the latest addition to a growing list of publications now supporting NonStop and whether your preference is for a Connect publication such as The Connection, or Tandemworld or even the Availability Digest, the good news here is that there are multiple sources you can turn to for up-to-date information on all things NonStop related. A thriving, indeed growing, media ecosystem is every bit as important as the ecosystem we have in support of products and services and I think that this is one of the big surprises that has come out of the recent RUG events we have participated in – so many members of the NonStop community are now deriving their income solely from supporting NonStop!

The upcoming NonStop Technical Boot Camp is now only a few weeks ago and a lot more will be written promoting this major NonStop event in those weeks. At the recent RUG events there has been much that has been hinted-at by the NonStop folks giving presentations so expectations among the NonStop community are running high. CTUG proved popular and was a packed house for the entire time I was there but hints about what was to come were being provided on a regular basis.

However, whatever is eventually disclosed at Boot Camp I suspect will still be only a small part of what NonStop development is pursuing as for sure, they are moving fast. A lot of the attention will be on the data base and the big changes happening with NS SQL/MX but like everyone else, I am going to have to wait to see what is announced on the day.

Ecosystems surrounding popular technologies, products and services are always a good sign for any business evaluating solutions. When it comes to NonStop then there is an abundance of vendors making up a very strong ecosystem. And whether it’s a product, a service, a simple question or two, the NonStop community has never been better served than it is today and this is something we all need to share with our associates and colleagues as for sure, I have to believe, for many of them, “They don’t know NonStop!”

Friday, October 14, 2016

Fall is a time to take in change; RUG events likewise are time to reflect on change with NonStop

NonStop is evolving yet again as it heads down different roads. Just how disruptive a technology NonStop will become? Why don’t you participate in the next Regional User Group (RUG) meeting to find out!

In my previous post I wrote of my brief encounter with a simulated race track experience at HPE Discover 2016 and of how I didn’t much care for the simulator. This past week Margo and I drove through Colorful Colorado on our way to Scottsdale for the DUST Regional User Group (RUG) meeting and my earlier observation about not caring for simulation was justified. There’s nothing like the real thing and the display of color this year was as good as it has ever been. The photo above was taken while Margo was behind the wheel approaching the town of Durango, CO!

For this trip, we elected to skip the freeways and stick to paths less traveled. In so doing, we couldn’t help comment about the path HPE was now treading as it pushed ahead with dramatic changes for NonStop. And these changes are taking place rather rapidly – from a single NonStop system to a Hybrid NonStop with Linux (and Windows) to no system at all. NonStop running virtually on commercial off-the-shelf hardware, and while these transformations have been well broadcast before their availability there’s still considerable surprise over what we are now seeing coming from the NonStop development.

However, it just had to happen. For NonStop to continue to be a contributor to then bigger HPE, NonStop the system, the platform and the software had to change its ways. At no time did I ever consider that NonStop was a “club” where only those invited few would meet around drinks and talk of the good old days. One of the difficulties many vendor user groups have suffered from over the past decade is exactly this – as numbers of systems and number of installations declined, they closed ranks and simply enjoyed fellowship with each other. This is far from a healthy sign for such organizations and it’s so easy for the more established user groups to go down this path. For the vendors, it’s perplexing and not something that excites them any longer – why fund an organization that simply feeds itself?

I have been told more than once that user groups and the communities they serve are just one more example of legacy. We have always done this, so the commentary goes, and after four or even more decades, this is who we are. This is our culture and our comfort zone but unfortunately, it does very little to encourage newcomers to join the ranks. And this is precisely why I keep returning to the RUG events no matter where they may be held. NonStop RUGs generate such  positive energy for me that participation in them remains a priority and no, they aren’t becoming “exclusive” clubs. These days, I am pretty much restricted to those being organized in the Americas but the word I received following the recent VNUG event outside Stockholm seems to confirm that what I am witnessing here is very similar to what is happening elsewhere.

I have to thank
OmniPayments, Inc. CEO, Yash Kapadia, for giving me an opportunity to present on his behalf at DUST. In so doing, I was given considerable leeway by Yash to frame OmniPayments latest services option – OmniCloudX – within the context of the changing NonStop landscape. Not too long ago I posted to the ATMmarketplace blog, Dressing up for the occasion ... where I referenced the work being done by OmniPayments to deliver a hybrid system, all LAN connected, in the one box. “Yash anchored HP standard chassis for rack-mounted processors with enough processors to support the key components of his payments solution running on HP NonStop. He then threw in a couple of Atalla security modules before topping off with ProLiant servers running Linux and Windows.”

However, that was just the start for Yash as he worked to better leverage what he saw coming in the NonStop roadmaps that were being presented at RUG events. OmniPayments may be offering its OmniPayments suit out of a cloud that’s based on NonStop X systems, but looking further afield, there is little to suggest Yash will not be adding support for NSADI / Yuma for faster access, exploiting more of the “white space” he has observed when running on NonStop. Translation? More applications coming back to run on NonStop! There are two other aspects of what Yash is pursuing that truly intrigue me – centralization and build-your-own!

Centralization; what the arrival of hybrid infrastructures with mixed NonStop and Linux (and Windows) does is foster greater centralization. Aren’t public clouds nothing more than Service Bureaus and Time Sharing Options (TSO) revisited? If you trace the roots of clouds back far enough in time, you will be surprised to see that their initial popularity was derived from the reality that business didn’t have access to qualified IT staff. With few businesses prepared to invest in training any more, isn’t it reasonable to expect centralization is a return to pooling of expertise in a very highly structured environment?


How many times have NonStop vendors reported that a customer has asked them to send on site their best developers to resolve a problem when in fact, the best course of action is to send someone who can best communicate what they observe, leaving the best developers to work with their development and test environments difficult to reproduce in the field? There is real value in having a small number of centers around the globe, each with enough skilled personnel to support the population of end users 
depending upon them?

By build-your-own then this too is nothing new but rather a return to the days of VARs. As we see more information about vNonStop being reveled and HPE puts in place the processes necessary for anyone to order vNonStop there’s the small issue of ensuring the “NonStop-ness” of vNonStop and as much as NonStop development sets forth a reference model, it opens the doors like never before for smart companies to begin on-selling their favorite white-label x86 server with OpenStack and vNonStop. Differentiation based on superior levels of availability all while supporting your favorite development environment – tell me what’s bad about that!

However,
 this goes a little beyond what has been historically associated with VARs. We have already read about OmniPayments becoming a VAR for both NonStop and Atalla in select South American markets where it fills a need in those areas no longer supported by the HPE NonStop team. The new crop of VARs that I am anticipating emerge will also provide remote management as well as a range of services from simple upgrades to the operating system and middleware to capacity planning on through to what had always been performed by on-premise systems managers.

There really isn’t any hiding from the fact that today, it’s not only that businesses aren’t investing in training but that the pool of knowledgeable talent is shrinking as the years pass by – the only place where such investments will continue is at these new age VAR “centers of excellence.” One example of this is
TCM Solutions who I have been working with this year and have come to really like its model as TCM tackles these exact same issues. TCM CEO, Tony Craig, is almost evangelical in his belief that, just as clouds are a return to centralization, so too is the emergence of managed services providers. Better to have all the expertise needed to support NonStop under one roof with the tools needed to support anyone at any time anywhere in the world?

There remain a lot of unknowns surrounding hybrid infrastructures and clouds and, when it comes to the deployment of NonStop, as either part of a hybrid system or within virtual machine, leadership will more than likely come from solutions providers. They have the most to gain from mastering either environment as remaining price competitive while offering a better solution is a major goal for them.

It would be simple for me to make the observation that this will all sort itself out over time but the simple truth is there’s no quick and easy fix. Just a lot of perseverance, as different models are tested. NonStop is evolving from what we once knew NonStop to be and is heading down different roads, meeting different needs and, in the process, there will be disruptions. Yes, this new NonStop will prove to be highly disruptive technology and it’s been a long time since words like this have been associated with NonStop.  

The DUST RUG event turned out to be well worth the time spent driving to it  - and the opportunity to spend time taking in the changing colors that the Colorado landscape never fails to provide at this time of year making up for the weekend spent in transit. Shortly, we will be returning to the road for the slightly longer transit to Ontario, Canada, for the upcoming CTUG event. We may be a little late to see the fall colors along the foreshore of Lake Ontario but CTUG will more than make up for any disappointments on that score.

NonStop, on the other hand, is definitely changing its ways and for that, and the impact it will have on the broader NonStop community, we have to thank the HPE NonStop team and I for one, am pleased to see the path it has elected to take. NonStop a part of every businesses cloud? In time it’s bound to happen – there is still no better way to offer up services from within a cloud than knowing that they will always be there, 24 X 7.

Thursday, October 6, 2016

More milestones – this time, it’s the 400th post! And did I mention that it’s all about NonStop!


For all the time I have been blogging, it never occurred to me that there would be 400 posts written over such a short period of time. And yet, the basic premise for NonStop in the marketplace hasn’t changed – we just don’t want outages of any kind. But wait ... there's a lot more to this story!

Celebrations always include moments of reflection. When a milestone is reached and where there is energy to keep going, these milestones can sometime be nothing more than bumps in the road. It was only a few weeks ago that I wrote a post about this blog entering its tenth year but with this post, we will be celebrating 400 posts. I was thinking early on that perhaps interest in posts to Real Time View will wane considerably with time but, in all reality, thoughts like this couldn’t be further from the truth. There’s a growing audience for all things NonStop!

On the one hand, barely a month goes by without HPE being in the news. Likewise, the pace of development for NonStop has ratcheted up significantly over the past two years, so much so, should HPE deliver everything depicted in the current NonStop product roadmaps, NonStop will be available in different guises targeting many different marketplaces. Whether purchased as a complete system delivered directly to the data center or downloaded as software, NonStop continues to provide value and in a world deep in transition, bridging traditional IT with clouds private and yes, public – availability remains an issue that needs to be addressed. And NonStop continues to deliver.

As I look back and reflect on much that I have written to this point, I have never wavered in my confidence that HPE, in the NonStop product line, has a prized server capable of doing practically anything when it comes to transaction processing. NonStop inherently scales out even as its fault tolerance is once again becoming a talking point – how do large applications interacting with the general public, stay up and running 24 X 7? I have looked over countless publications and news releases by HPE for the past decade and it’s interesting to observe just how prominent the attribute “fault tolerant” has become – after a very long time of using almost every other term, it seems back in vogue within HPE.

Fault tolerant? It’s a story that once focused solely on the hardware. More recently, the key differentiator was the “integrated stack” that included the operating system, the database, the transaction monitor and all the logging that goes with it to ensure not only 24 X 7 operation for a single system but almost instant recovery and take-over by an “active” secondary (or even tertiary) off-site system. Yet the hardware, together with the integrated stack, isn’t the full story. NonStop is both a product, a technology and an architecture and it’s now heading down a path to give HPE customers and prospects a wide pallet of options from which to choose – all influenced of course by IT facing the difficult transition to clouds where availability and scalability and indeed security are of the essence.

It would be very easy to me to continue to look back at this time and recount the many highlights NonStop has enjoyed over more than forty years. However, it is the future that interests me most and while I was not privy to every “whiteboard discussion” held within NonStop development and product management over the years I have been pretty good at reading the tea leaves and making educated guesses. Had you been in the audience of SATUG back in 2008 you would have heard my presentation on just how important developing virtual NonStop was to the future of NonStop. And yes, if you had been in the audience of DUST this week, you would have heard even more!

NonStop as a product, a technology and an architecture – it is pretty obvious where NonStop is now headed. From a product perspective, HPE has made it clear that NonStop will continue to be available as a complete system, whether it’s one of the NonStop i family members (HPE is committed to taking Itanium chips from Intel through to 2025), one of the NonStop X family members or the hybrid product based on NonStop X with an InfiniBand fabric integrating Linux and even Windows. A product line in it’s own right and one I expect to see provided as a complete system at some point, like other products in the Converged Systems (CS) product family should market demand develop where such an investment makes sense – think Financial Institutions where payments solutions have argued for some time for a “programmable CLIM” supporting something apart from a Debian distribution of Linux.

Race to your digital enterprise! The photo at the top of the page was taken during HPE Discover 2016 and featured groups of “drivers” with VR goggles competing with each other in digital race cars. Of course something like this would attract my attention but being averse to using simulators (they don’t give you any indication of how much grip you have), I preferred to watch other. It is a race to a digital enterprise but not one that can be accomplished on one fell swoop - a ruthless and fierce undertaking not for the faint of heart! CIOs understand the risks of the pressures to succeed even as they are acutely aware of the downside risk. When it comes to global enterprises it only makes sense to enter such a race with incremental change – the heart of transitioning to hybrid infrastructures.

NonStop is proving to be more than just a product but a technology and the externalization of its interconnect fabric with an API that supports a direct to memory transfer between disparate stacks (i.e. NonStop with Linux) opens the door for applications running on one stack to be integrated with applications on another. For the NonStop community I suspect it will not be very long before the technology behind NonStop is used to provide an overarching mantle on everything connected to NonStop – the inherent property of NonStop take-over will be extended to application processes running on Linux or even Windows. Transaction Monitors on NonStop, in other words, will essentially wrap these application processes within monitoring capabilities – persistence, and hence even greater levels of availability, will extend beyond NonStop.

Technology is often described as being a “collection of techniques, skills, methods and processes used in the production of goods or services or in the accomplishment of objectives, such as scientific investigation.” In this sense, NonStop is a key ingredient in not just processes used in the production of goods or services but rather, is that additional “special sauce” that makes these goods and services available 24 X 7 whether the supporting methods are on NonStop or simply adjacent to NonStop. In essence with the hybrid, converged systems, I envision appearing in enterprise data centers, the shadow of NonStop will be indeed long and cover as much of the servers present in the data center that the CIO deems part of their enterprises mission critical operations.

NonStop, too, is proving to be much more than just a product and a technology as increasingly it is an architecture that is attracting more attention of late. As NonStop evolves to become a software platform, running in a virtual world, HPE development will be providing models, templates and yes, a “reference architecture.” One of the key messages coming from NonStop development is that with the release of vNonStop –a software only offering – this reference architecture will ensure configurations of vNonStop are truly fault tolerant.

With a reference architecture there will be a separation between HPE the architect, and HPE the general contractor. Maybe you will rely on HPE for everything – the hardware, the virtual machine including the hypervisor and vNonStop, but perhaps not. As software houses and managed services providers gaining confidence, a new ecosystem of general contractors will likely appear some of which utilize all the HPE offerings, whereas others will mix and match hardware, virtual machines and vNonStop. Maybe you will rely on general contractors apart from the HPE NonStop team?

In the time it has taken me to write 400 posts, NonStop has come a very long way. As a community, all associated with NonStop have held our collective breaths on occasion as we watched the changes unfolding at HPE but as we gave a collective sigh of relief when we heard of the sizeable investment HPE has been making in NonStop, we have witnessed firsthand the emergence of a very powerful and comprehensive response to the growing market need for fault tolerance along with scalability and of course security. This is a segment that NonStop alone can address and is doing so in a compelling manner. So yes, I look forward to all that is ahead and here’s to the next 400 posts!