Wednesday, March 28, 2018

How did that happen – I have hybrid IT?

Hybrid IT is the new normal, according to one HPE blogger, for the NonStop community it may only be a matter of time before the plans of both NonStop and hybrid IT intersect …

Let’s face it. There are many disruptions that can interrupt our routines. Not that long ago I was a regular at the local fitness center but now, having moved, I haven’t found a new place to go exercise. I have always loved to walk and there will be some of you from my time at Tandem Computers who may recall that most mornings I could be seen with fellow walkers striding the sidewalks of Cupertino. Even when I lived in Simi Valley I would devise a routine that had me cirling the shopping center where my Starbucks was located so that I could eke out a 20 minute walk. I don’t know what it is but I am horrible at improvising and unless I have a set time and place to do something, I find ways to duck whatever task or activity I had set out to do.

On the other hand, we only just coming out of winter here, along the Colorado front ranges, and so to some extent, I can blame the weather. More to the point, however, it has become hard to ignore that I have put on a little weight over the winter and almost by accident, I can no longer wear the same clothes I wore last spring. Perhaps accident isn’t the best choice of words as nothing happened overnight and yet, here I am, wondering how I can return to my regular exercise program. Working all day at a keyboard demands a price but I hand’t thought that would include a few extra punds!

Talking of accidents leads me to remark somewhat candidly about a tweet I happened to come across by chance. Gary Mitchell, a fellow HPE independent blogger who I happened to catch up with late last year while at HPE Discover, Madrid, tweeted “A new report from @Forrester finds 2/3 of businesses end up w/a hybrid estate by accident – not design.” The link takes you to the HPE newsroom and to the article,   Survey Says: Composability and Continuous Delivery Accelerate Business Initiatives and Deliver a Competitive Advantage. “We were curious to better understand how IT is dealing with this transition, so we commissioned Forrester Consulting to conduct a survey to evaluate the current state of Hybrid IT,” said HPE blogger, Gary Thome.

“While Hybrid IT is the new normal,” Thome noted, “two-thirds of businesses ended up with a hybrid estate by accident – not design.” Now, there are all sorts of ways to accidental IT these days. Sorting out IT following a major merger (or breakup), changing the IT leadership and with it, changing the applications and perhaps more commonplace these days, migrating to open platforms. Whatever the cause, IT suddenly wakes up to find it has inherited a completely foreign mish-mash of technologies with the directive to get it all working. And now! On the other hand, clouds and embracing clouds, shouldn’t be among the accidents IT inherits unless, unhappily, IT is deliberately kept in the dark. And there are plenty of instances where change was not only an accident but a deliberate sidestepping of normal IT oversight.

I am often caught up in discussions about what the data center of the future will look like if, at all, it still exists within Global 1000 companies. A very strong case can be made that when your business isn’t centered on IT either as a manufacturer or a vendor, then perhaps you shouldn’t even have an on-premise data center. Today, there are many enterprises rethinking their strategies in this regard and I know of one very large retailer, for instance, that is electing to move it’s IT support for payments to a cloud offering.

And this just makes sense – back in the 1970s when working for a Caterpillar Tractor distributor, the enterprise first consolidated all of its IT within a new organization and then sold it to a New York consulting company where it became a major service bureau, but accidental hybrid IT? It does sound a bit far-fetched even as it might have us looking to the horizon for the imminent arrival of storm clouds!

Time sharing; service bureaus; clouds – they have a familiar ring to them in one sense. It is ultimately just another perspective of relenting control and yes, funding, of the data center to someone else. I can recall all the passion and enthusiasm for embracing best-of-breed solutions freeing enterprises from having to place major bets with select few vendors.

But it “growe'd like Topsy” and even with the best defined infrastructure in place, an awful lot of silo-ed solutions emerged that even today require additional infrastructure to ensure viable participation with the rest of IT. When I look at what is being addressed by hybrid IT, it is starting to look a lot like the big data centers of the past as increasingly, enterprises argue strongly that the applications deemed mission critical really should run on-premise, in a private cloud!

I recently wrote to my clients about an in-depth article published late last year in Virtualization and Cloud Review,
The Composable Datacenter of the Future, where writer James Green starts out on familiar ground:

“IT Infrastructure is always evolving. The next big change will be back toward an old, familiar model, closer to the days of the mainframe than the era of silo-ed servers that has dominated for decades.”

However, possibly of more interest to NonStop vendors is a subsequent observation Green makes:

“What Does It Mean for Infrastructure To Be "Composable"? Software that enables the abstraction of infrastructure resources transformed the modern datacenter—most notably, the abstraction of physical machines into virtual machines (VMs) via a hypervisor. Spurred on by the benefits many organizations realized by abstracting the machine construct, some datacenter thinkers have been working toward building a ‘hypervisor for the datacenter.’”

Accidents? Familiar models? And now, hypervisor for the datacenter? I am not so sure that what is being referenced here can be considered haphazard or even random as there is an underlying plan working itself out among enterprises. For those enterprises that deem oversight of the business logic and data as being crucial to the ongoing success and indeed expansion of their business it is important to stay on top of the technology and the best way to do that continues to be the big boxes in the air conditioned room. Or vault, as it is increasingly becoming given the unease among enterprises executives about security.

There is a lot being written by HPE about composability even as there is a lot of head scratching among the NonStop user community as to what it all means. There is a very solid line connecting composable to virtual machines and that’s one place to start. But for now, you will not read anything from HPE suggesting that future virtualized NonStop workloads will be able to run on any hypervisor for the datacenter supported by infrastructure that is composable and yet, it wouldn’t be an accident if such a situation develops. It is common knowledge that the push to support NonStop on VMs came from the NonStop telco customers and there is every indication that shortly, HPE will take the wraps off a telco up and running with virtualized NonStop so perhaps it’s all about to happen for NonStop.

That is, NonStop being software-defined as HPE would like all of its core products to become – yes, think first up how advantageous that would be in development shops and where QA, testing and PoCs are the order of the day. Think too of the multi-tenancy on offer for NonStop SQL/MX and how this plays into a software-defined world. It’s no accident that HPE has let its own IT group drive requirements for NonStop and it’s no accident either that NonStop was never a consideration for the spin-merge of non-core software products.

Fault tolerance, as it is embodied in NonStop, has enormous value add in the world of hybrid IT and it would be foolish for any enterprise to discount the benefits that come with levels of availability only NonStop can deliver. After all, we may be in hybrid IT by accident but there really isn’t a positive spin that you can put on an outage of any kind these days – accidental or not!

I know that there is work ahead for me to get back into shape after a prolonged absence from our local fitness center. At least, I have a good excuse and can quickly address the situation. Enterprises everywhere know that they are entering the world of hybrid IT and that software-defined everything will rule the day. While I cannot get HPE or the NonStop team to make any statements about future product directions I have to believe, based on recent conversations and email exchanges with those close to HPE, that NonStop isn’t oblivious of HPE’s strategy and is paying a lot of attention to possible future synergies between NonStop and hybrid IT.

Clouds are now all around us and whether you see them as the harbinger of stormy days ahead or not, recent NonStop developments suggest that there are plans afoot for NonStop systems to better leverage cloud computing. From where I sit there is nothing accidental whatsoever with the journey NonStop is on and as someone with longevity in IT, I have to say, I can’t wait to welcome NonStop to hybrid IT!

Sunday, March 18, 2018

Yes, I read it in a blog!

With HPE CEO Antonio Neri posting and tweeting – HPE strategy is becoming clearer and it is all good news for the NonStop community!
Most of my background in IT is well known to the NonStop community but this isn’t the time to dig into the archives and reflect on past accomplishments. However, and just an observation in passing, for someone who was often told that he would find it hard to stick to doing just one thing well, maybe there is a story here after all. The NonStop community is witnessing change, but change a little different from what the rest of IT is observing. Yes, we are embracing hybrid IT in one form or another and yes, we are coming to terms with virtualization, provisioning and software-defined-everything. And yet, for the NonStop community, change is happening in unforeseen ways as it wrestles with where NonStop is headed amid concerns about “what is NonStop!” 

Expressed differently, are we happy to be given more options as to how we deploy NonStop or are unhappy about NonStop apparently responding to questions we never asked? My background in IT has exposed me to many turns of the technology wheel and from where I sit it’s as if we have circled back to topics I thought we had addressed long ago. Decades ago, what fueled the argument centralized versus distributed was the uptake of online systems and the transaction volumes they generated, but today, it’s not so much about transactions per se as it is about the data. If we continue with the systems and deployment topologies we have relied upon for two plus decades then we will be inundated with so much data that we won’t be able to handle it.

While a lot of attention is being given to the edge, when it comes to NonStop any consideration of running NonStop at the edge is probably an exercise in futility. At least for the next decade! For starters, there is greater potential for high performance computing running analytics and AI to make it out to the edge  than NonStop, but that isn’t to say some solutions provider may build a product that exploits NonStop out on the edge at some point. On the other hand and a little more realistically, with discussions about hybrid IT and converged / hyper-converged becoming more pronounced, NonStop has a very realistic shot at providing value in the data center. Centralized may not be the same as it was in former times, with multiple data center sites all seamlessly integrated constituting an enterprises modern data center, but the expectations will remain – business critical logic and data need to be accessible 24 x 7 x forever.

But what is the strategy of HPE? What are HPE’s plans for NonStop and do they dovetail nicely with HPE’s overall strategy? Before delving into HPE’s strategy, it is good to start with how new HPE CEO, Antonio Neri, views the world and what HPE considers is the next big thing, the next big challenge we all face, and where HPE sees itself playing an important role. In the March issue of NonStop Insider, in a new column we headlined
HPE: News from the very top … reference was made to a post by Neri, The Next Big Thing is … The Intelligent Edge published in the LinkedIn blog, Pulse:

We live in a hybrid world – the IT estate has to span from the edge to the core to the cloud because the power is in the data and you have to connect and secure all of it.
The challenge in this future is that the volume of data being created will soon outstrip our ability to capture, analyze and act on it. This intelligent environment will require a platform that is open, simple, automated and secure. This is exactly why we’re advancing our vision of Memory-Driven Computing.

What can you expect to see next? Remember the first time you heard about the possibility of an autonomous vehicle? Or how revolutionary AI and blockchain could be? Now think about what could happen when IT leaders apply real-time intelligence to retailing, sports arenas, medical facilities, college campuses, transportation systems and smart cities. No doubt, we’ll achieve higher productivity, mesmerizing customer experiences and change the way we live. In every industry, every city, everywhere.

There will be an intelligent environment. AI and yes, Blockchain will be important and yes, HPE will apply real-time intelligence to retailing, sports arenas, medical facilities, college campuses, transportation systems and smart cities. Memory-Driven Computing? Yes, we are seeing it being introduced into the marketplace already as the SuperDome Flex – it’s definitely a memory-driven computer and it is definitely foreshadowing the potential that is an extension of the first prototypes coming from The Machine. But wait, there’s more – Memory Driven Computing is at the very heart of the converged and hyper-converged models HPE foresees the market requiring to process the workloads that will arise from all the data being ingested.

Beyond simply viewing the world this way, in a subsequent post to the HPE web site, Neri gave more information about HPE’s strategy and of the markets HPE will be focused on as it addresses meeting the described needs of an intelligent environment. In a recent tweet by Neri that takes us to a blog post by him,
Market Share and the Tale of Two Strategies, he fills in some of the blanks that have existed within HPE strategy as articulated in the past:

Within the compute market there are a number of different segments and they are not all created equally. We think about them as volume markets and value-and-growth segments. At HPE, we are aggressively segmenting the market to ensure we are making the right investments in the value-and-growth areas where we can bring differentiated value like hyper-converged, high-performance compute and Synergy, while also profitably supporting our customers in more commoditized parts of volume segments like rack and tower. But, as I have said many times before, we are not going to chase market share just for share’s sake. We are going to drive profitable share.

HPE announced last fall that we will no longer pursue the low (or zero) margin, custom-built, commodity server business with a very narrow set of Tier 1 service providers – Apple, Amazon Web Services, Facebook, Google Cloud, Microsoft Azure in the U.S. and TenCent, Alibaba and Baidu in China. While this segment of the market is large and growing, there is simply no profit to be had by HPE or our competitors because there is very little value we can bring. We still do substantial business with these companies, but we are focused on delivering solutions that provide differentiated value for them like our value-and-growth compute, storage and networking offerings.

This is the strongest language to date concerning what HPE is not going to do. There should be no ambiguity here when it comes to understanding HPE’s strategy – if it is a high value, high growth opportunity then HPE will address it. Otherwise, look for other vendors who may be interested only in market share – a game HPE no longer wants to play, and for good reason. It is in this high value, high growth market where change will be needed as it is inclusive of the Global 1000 enterprises HPE is already working with seeking to “apply real-time intelligence to retailing, sports arenas, medical facilities, college campuses, transportation systems and smart cities.” And I got the scoop on all of this from reading a blog – did you read them, too?

I have covered some of this material in a couple of my latest private emails to my clients but let me spell out what this all means for NonStop as best as I can. There it is for all of the NonStop community to see: hyper-converged, high-performance compute and Synergy. The investment HPE has made in virtualized NonStop means that in response to where it provides high value and high growth, virtualized NonStop has the potential to become just another, albeit very important, virtualized workload in a hyper-converged system. They are all x86 and they support converged Ethernet so the basics are all in place. In talks I have had with various engineers on the NonStop team, it appears that it is only a matter of time before enterprises make a request to run their mission critical transaction processing virtualized NonStop workloads on the latest hyper-converged systems.

As the new CEO of HPE Neri covers a lot of ground in these two recent posts and while specific systems and platforms aren’t named (apart from Synergy), the door is left wide open for NonStop. Nothing I have read from HPE this year leads me to believe that anything included as part of HPE’s strategy precludes the participation of NonStop. Ultimately, it comes back to the imagination of the enterprises involved and the creativity of the solutions vendors working with them, but it is so tantalizingly close to think that the most important of all of mission critical applications, no matter where they run – on scale-up or scale-out systems – can opt to turn to NonStop and in so doing, bring to HPE the commensurate amounts of high value and high growth it is so focused on providing. There will always be a need for the levels of availability NonStop alone can deliver – so make you plans early and by all means, get to L-Series as fast as you can. There are opportunities all around us and we simply cannot afford to let them pass us by; it was Neri who said it best “we are not going to chase market share just for share’s sake. We are going to drive profitable (market) share.” And what better way to succeed other than with Nonstop!

Sunday, March 11, 2018

Our strategy for new applications is to build and deploy off-premise!

It is becoming all too clear. For the NonStop user the world is changing rapidly as we move to hybrid IT and look to a virtualized NonStop. Where will find the help we will more than likely need?

Only this week I was visiting our new off-premise storage facility where we are now storing vehicles we can no longer accommodate on-premise, at our home. The company command center – yes, our faithful RV that many of you have seen at recent RUG events, our motorcycle, our trailer and even our fifteen year old Corvette “track toy!” It is not a cavernous storage facility by any means but it does come heated and with a number of different power outlets including one 30amp circuit which should be enough to keep the command center’s refrigerator working and the drinks chilled. Now, all we need is a couple of chairs and we have a new weekend retreat! Well, perhaps not …

Fortunately it is all rather simple and we don’t need to have someone present looking after the facility. There is a management company lurking somewhere in the background that watches over the site, but to have someone on hand to consult when it comes to what more we can push off-premise well, that too is fortunately another instance of, perhaps not. In today’s world where we live, work and relax, there are so many instances where you do need to reach out for a helping hand. That it’s as if we are being viewed as totally helpless and yet, if something needs work to be done, I still call for assistance whether it’s a leaking faucet, a misbehaving HVAC or yes, the tires need replacing on the car. As the television commercial promises, “we have people!”

I came across a link this week that took me to the site of and to a Q&A about IT and Business Strategy. DXC having been created late last year with the merger of CSC and the Enterprise Services business of Hewlett Packard Enterprise. You may have seen this link to too as it was being promoted on LinkedIn. “Moving applications to the cloud requires a sound strategy that takes into account technology issues, business needs and embedded IP. Assess your organization’s readiness level and ability to take on this important work …” The topic of cloud always grabs my attention and when it comes to IT and to the data center, there is nothing more topical than cloud computing when it comes to off-premise offloading of both processes and data.

The link on the page took me to a self-assessment page where the first question stopped me from going any further, “How much has cloud affected your IT and business strategies?” For the NonStop community this is currently the topic that is most hotly being debated leading to us all asking the question, “How much has cloud affected my NonStop deployment strategy?” And by this, with options to run on x86 servers, either from HPE (as NonStop X) or third parties (as virtualized NonStop – vNS) with the option to run essentially on-premise as a private cloud or off-premise co-located on a public cloud?

When asked the question above, “How much has cloud affected your IT and business strategies?” you could say you were pursuing a position of leadership by, “using cloud to create new digital business opportunities to better serve customers and to improve employee efficiency and effectiveness.” Or, you could back away from this just a little and say that your, “apps strategy is to build all applications on cloud (such as Cloud native, PaaS, SaaS). Perhaps it was more a case of you “provide cloud services, and you have made cloud a key element for many of our applications and solution architectures.” To be honest, for many members of the NonStop user community I think the correct answer may be, “We don't have a cloud strategy. IT doesn't support cloud. Use of cloud is up to the requestor.”

It is this last observation that truly raises that all-important question – do we need help? Or, put another way, do “we have people?” I may not need a facility manager to help my off-premise storing of vehicles but when it comes to IT and NonStop, from where I sit there is a very strong argument to be made that sure as heck, we do! Language aside, irrespective of the resources available from HPE, any discussions of NonStop, clouds and IT and business strategy leads me to believe that we will be seeing the birth of a rejuvenated managed services ecosystem in the very near future.

And with this still fresh in my mind, I went to TCM’s Daniel Craig for his impression on the options the HPE NonStop development team is providing today. “TCM is very keen to explore the space that is being created by HPE's NonStop development: vNS, Hybrid, DBaaS etc. We have been working towards a NonStop-as-a-Service solution for some time now and we are confident that very soon we will have the right blend of technology and willing users to make this a reality.” When questioned further about vNS, Daniel said that, “Commercially we've held many discussions, but more from hypothetical than practical viewpoint. It seems a lot of Customers, especially those not already setup for private-cloud operations, are holding off to see what HPE is going to do in this space. Certainly the L-Series release with support of vNS with VMware has TCM's attention as we suspect that this may be a more readily deployable solution for most NonStop users.”

TCM is a provider of managed services and was the first to enter this market targeting the NonStop community. Now TCM is the largest managed services provider in the EU, having pioneered NonStop service solutions where mission critical systems and the solutions it supports were of paramount importance to the enterprise. It is only natural then that TCM is anticipating seeing even more business headed its way as TCM definitely can say, “we have people!” and the HPE NonStop team is continuing to advocate to users that they better be well-educated in all things virtual before they begin pursuing a vNS deployment in their data center.

However, the cloud continues to loom large for even the most ardent supporter of NonStop. Whether you follow the path established by HPE IT department with NonStop at the head of their operations serving up NonStop SQL to all and sundry on the basis of DBaaS or simply look to go off-premise and to a side running a combination of x86 servers with VMware as the hypervisor, the challenge remains. Where is the expertise coming from and how do you imagine recruiting the staff knowledgeable in both VM and NonStop? TCM will be the first to acknowledge that this will be a challenge for most enterprises, but it is also quick to highlight that this is precisely the area where TCM is investing today!

“Let’s just look at what we have done in Scandinavia and at the investment we have already made,” said Daniel. “TCM has just taken over from DXC in providing services to their former Scandinavian clients. It has been a lot of work but the TCM team, DXC's staff and the Customers have all been extremely professional, making it all the easier to pull off without a hitch. Great collaboration and teamwork all round!” Furthermore, and with this in mind, Daniel wrote of how, “It has also been a good opportunity for us to try out something new from a services perspective. For the last few years we have been concentrating on our cloud-based services via our Centre of Excellence. Indeed, this was how we first began providing service into DXC (CSC at the time). It is now all working out incredibly well and harkens back to TCM's willingness to adapt, try new things and remain flexible - basically whatever best suits the NonStop user. Yes, we certainly do have people!”  

Whether it has become your strategy for new applications to be built and deployed off-premise and in the cloud, it really is only a case of rethinking your NonStop solution as being just another virtualized workload. Running on the L-Series operating system, it can run anywhere there are x86 servers, some Ethernet deployment (capable of supporting RoCE), and a hypervisor including KVM and now, VMware. I fully anticipate NonStop users to first deploy on-premise and with the help of managed service providers like TCM, gain knowledge of what it takes to successfully deploy NonStop with the same fault tolerance as is supports today running on NonStop X, and then look at the business benefits of moving a lot or perhaps just a little to an off-premise public cloud. And if this is indeed your goal, then perhaps it is time to have the people and the way forward may very well lead you to folks like TCM. You need people? Sure do! Then we can help you – “Just give us a call,” said TCM’s Daniel.

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...