How easy it is to NonStop!

From video clips to social media posts to publication editorials, NonStop is gaining traction whenever availability is no longer an option. But the question remains, can you afford not to NonStop?                                                                     

We are now deep into spring and for us, living in the Northern Hemisphere, summer is in sight. For us Americans, Memorial Day isn’t all that far away and with this remembrance of times past, we begin a period of backyard BBQs and days at the beach. Yes, it’s nice to step outside to enjoy a timely glass of wine. Well, sort of; no one is prepared just yet to get too far ahead of themselves as we continue to be influenced by what is driving all of our plans – the global pandemic we know as COVID-19.

This past week has been a time where Margo and I have been focused on publishing the May issue of NonStop Insider. Have you read it yet? Have you checked out Margo’s editorial? Just as importantly, have you become a NonStop Insider subscriber? What we have come to enjoy reading are numerous opinion columns that are submitted as they provide different perspectives on NonStop the product and NonStop the community. It has been our goal to limit the editing to a bare minimum so that readers can enjoy the story lines “as they are written!”

It shouldn’t be a surprise for any reader of this latest issue of NonStop Insider to know that the focus of many of the articles is on business continuity and in particular, on the need for business continuity plans. NonStop systems offer levels of availability unmatched by any other vendor, but even so, this doesn’t negate the need to be prepared for outages manmade or natural. In these most trying of times, nothing can be taken for granted, but fortunately the NonStop community is well-served with many good products addressing business continuity available from NonStop vendors.

It is all too easy to imagine our current predicament lasting longer than authorities predict. Any forecast being given by any institution can be taken with a grain of salt as the unknowns do outnumber what we understand today. Given how so many of us are working from home (and doing a really good job as best as I can tell), we are all watching to see how our employers react when time is finally called off on self-isolation and we can begin to assemble, in numbers, in offices and cubes left idle for so long.

It shouldn’t be a surprise to read of how even now there are major enterprises that are still taking outages at this critical time. To be honest, aren’t you surprised that there isn’t more attention being directed at those businesses that have failed us – how long did it take you to get through to your bank, your travel agent, your government agency? And when we did get through, all we were told was that “our systems are slow” or indeed, nonfunctioning? In many ways, it is a source of frustration to the NonStop community that enterprises have an option not to suffer outages and yet for one reason or another fail to consider the opportunity to run 24 x 7 with an off-the-shelf product.

In a May 8 video HPE Cancels Coveted Customer Award, HPE NonStop Sales Specialist, Steve Kubick, suggests that perhaps it’s down to our own enterprise’s architects not realizing how easy and how inexpensive it is to deploy NonStop. Existing NonStop customers do not tolerate outages and yet, it still hasn’t dawned on these enterprise architects that they too need not suffer from downtime planned, or unplanned. Kubick makes it very clear when he advocates for NonStop advising these same enterprise architects of “how easy it is for you to NonStop!”

The antithesis is often overlooked by these same enterprise architects: How difficult and in fact expensive it is not to NonStop. Rather than going into the benefits of NonStop it’s better to paint a picture of availability using very broad strokes. And in doing so, let’s line up NonStop against Linux where Intel x86 architecture is exploited. When it comes to hardware today’s NonStop is built atop HPE ProLiant servers linked via a Mellanox (soon to be Nvidia) InifiniBand (IB) switch. But there’s no big price difference here as you can build a similar configuration using anyone’s x86 servers linked via a similar fabric based on IB.

Then again, continuing with the broad strokes, NonStop can run virtualized so any enterprise already deploying x86 servers where VMware is deployed, the fabric required for virtualized NonStop is simply Converged Ethernet where the interface is via RDMA over Converged Ethernet (RoCE). Again, no substantial price differences here and we have an option to run mission critical applications within our server farm that leverages the benefits of a fault tolerant virtualized NonStop (alright, I can live with some enterprise architects naming this their private cloud as yes, painting with broad strokes continues).

As for software here the Linux advocates begin to lose ground. On paper, there is so much open source on offer that surely, we can build out our platform free but here the first missteps can be identified. The enterprise wants full vendor support for any open source brought into the data center preferring to go with RedHat, for instance. Ouch, now you have the “configuration twostep” to master (and maintain). As you build out your stack with messaging and transaction support then add a database and move on to sort out replication and monitoring these separate products bring with them considerable feature overlap that requires a level of expertise (indeed experience) to sort out. Which product is responsible for logging, for instance; and the logging of what exactly?

By comparison, out of the box NonStop comes with a completely integrated stack with no feature overlap as NonStop developers talk to each other. Surprise? What this leads to is that the costs to get to a usable level of availability without considerable fragility through complexity is now in the same ballpark that you will experience with NonStop. So, where is that cost for NonStop enterprise architects seem so ready to promote? A decade or more ago, there was a price premium to be paid for fault tolerance, but no more.

One last item to consider as these broad strokes continue to be made on an almost complete canvas: the costs associated with staff, be they developers, systems managers or operations personnel. In case you missed the memo, most popular DevOps tools, utilities and libraries are supported with NonStop. Did you check out the earlier post to the HPE Community blog Modernizing the development world of NonStop applications where the topic of DevOps is covered? NonStop SQL is now so compatible with Oracle you don’t have to worry your DBAs about something unfamiliar. As for operations everything is done online – a NonStop system never has to stop; no need to plan for downtime whatsoever.

Where the scales swing back to favor NonStop however is the drop in staff required to manage NonStop versus Linux. DBAs can be reduced to a part time responsibility and operations? Well, so much is automated today that this too becomes almost a casual assignment. Too broad a stroke? Not really as all you need to do is check in with any large enterprise running NonStop and no matter the industry, there are plenty to choose from who have elected to run NonStop.

Business continuity in these times is of paramount importance. Having proven business continuity plans in reach at all times is a necessity. Deploying NonStop simplifies it further. These days, perhaps it’s even harder than ever before to ignore HPE NonStop as you will find today that it has become so much easier to NonStop. So, all that remains is to ask you; what stops you?