Wednesday, December 26, 2012

Yes, I get it!


When it’s an Indy racing car driver that expresses so succinctly the value proposition that comes with embracing standardization and commoditization, the message for the NonStop community is hard to miss and forward-thinking vendors are responding ...


For several years, well, decades really, we have been collecting wines that were sold in magnums – essentially a double helping of your favorite wine in a really big bottle. There’s a couple of Penfolds, several Silver Oaks, Duckhorns, a Joseph Phelps and even a ’93 Bordeaux from a famous vineyard picked up cheaply at Heathrow airport when the US dollar was a lot stronger. However, they are mostly Australian reds and when the holidays come around, it’s a great time to pull a cork and see just how well they have developed.

As many of the bottles are from the early ‘90s, it’s always a mixed blessing – sure we like the wine but now the cellar looks the poorer with the absence of a really good examples of wines we like, but the response from Margo is always the same; get over it! The running joke in our family is that it makes absolutely no sense to leave wine as part of an inheritance. What’s the point – you acquire wine to enjoy it not to let it gather dust in a chilly basement!

The picture above is of a magnum of ‘94 cabernet sauvignon from the John Riddoch collection by the South Australian winery, Wynns – and even though it was just on the edge after a few hours resting on the kitchen countertop, it tasted great. Readers of previous blogs posted at this time of year may recall a very checkered past when it comes to picking the wines but yes, I get it. There are few points to be won from not drinking some of the best you have, as the fall-out from choosing poorly can be considerable as will be the reminders. If you missed those earlier posts then you may want to revisit “
NonStop? Spreading the word ...” and “Product Roadmaps! Still required?

But I do get it – special occasions call for special wines! And a go
od wine is always a great lead in to opening presents, not to mention fodder for a good story line. Last year my good friend Brian surprised me with a subscription to Racer magazine and it was only a day or so ago that I was able to find the time to catch up on my reading of older issues. And among the editorials that grabbed my attention was one by the successful Indy 500 racer and winner and two-time Indy car champ, Gil de Ferran.  The focus of his column was the demands to reign in the costs of race cars while ensuring innovation would be allowed to flourish.

“Look, I get it,” de Ferran began. “Try to control costs, restrict options, standardize on certain parts, increase production, amortize investment over a wider base, increase value for money … It all makes sense and works particularly when the supply of investment into our sport has been dented by the economics of this era.” It is really easy to substitute our technology industry for sport and everything else remains relevant. Indeed, when it comes to technology relevance, Racer magazine was kind enough to include in its own definition of technology relevance with the bracketing (1) of the application of scientific knowledge for practical purposes with (2) machinery and equipment developed from such knowledge together with (1) the relation to the matter at hand and (2) practical applicability.

Just as deFerran acknowledges, I too do get it. And I am talking about the HP NonStop system, obviously. Control costs, restrict options, standardize on certain parts, increase production, amortize investment over a wider base, increase value for money … all for the sake of technology relevance. What we have seen over the past couple of years are NonStop systems that are commodity-hardware based, sharing almost every component with other systems of the Business Critical Systems portfolio. We have seen standardization on a scale we would never have expected to see only a decade ago with the prices coming down even as the value for money has climbed. To many enterprises, NonStop is on threshold of crossing over to being bargain priced.

Discussions with solutions vendors providing products for financial institutions are revealing how compelling an argument for adding more applications onto the NonStop system these lower prices are generating. “Once you start relying on clusters of Windows or Linux servers, throw in the cost of the operating system, database system and the necessary tools in support of D/R as well as monitoring,” acknowledged OmniPayments Inc. CEO,
Yash Kapadia, “let alone factoring in the people resources needed to watch over such a complex environment; the value that comes with the latest small NonStop systems, including the most recently released NS2100 NonStop system, and of having everything you need in just the one box is inescapable. From what we have seen we have moved an application from off of Linux / MySQL and ported to NonStop with NS SQL/MX for a lower overall solution price. A circumstance I wouldn’t have believed possible just a few years ago.” 

Probably of more value to software companies is the support of other standards – the Java language and NonStop SQL/MX. It has also become quite routine to externalize applications via SOA and Web services, and increasingly, applications being ported to NonStop find it easy to bring with them enough of a HTTP server to ensure modern protocols including HTTP, HTML (including the latest HTML5 iterations), XML, etc. to ensure NonStop is no stranger to the needs of the latest PCs, laptops, tablets and smartphones. In other publications you will read of where I connect the dots between the growing integration of social media with business, the increased reliance on mobile devices, the value from getting transactional data into Big Data frameworks, and the cost-effectiveness from leveraging Cloud resources – private as well as public. None of this would have been conceivable without HP fostering an environment where embracing standards and leveraging commodity components was encouraged at all levels. A situation not lost on vendors like IR that have embraced Big Data even as they add support for mobile devices or Infrasoft and comForte as they take a much closer look at Clouds.  

However, what I simply don’t get is the lack of any real marketing “noise” about all of this coming from HP, and indeed, the full ecosystem surrounding NonStop. There are many incidents where NonStop customers talk openly of NonStop as if it were just another Unix box – you have to look very hard to even find the references to NonStop. Likewise, Nonstop vendors who have embraced platforms apart from NonStop for sound business principles designed to develop complementary revenue streams seem to stress the alternate platforms over NonStop relying on the customer to ask the question before NonStop is introduced into the conversation. Yet, even as I look back on 2012 I have to admit, many solutions and middleware vendors who added support for platforms apart from NonStop, have become much more optimistic about the future of NonStop and are more inclined to reference NonStop and for that, the NonStop community needs to be particularly appreciative.

Yes, the year has come to an end and the sales success of NonStop continues. Without giving too much away it only takes a few calls to those involved in selling NonStop systems to come to appreciate that the oft-quoted “another quarter of double digit growth” with respect to sales of NonStop systems is a reality and after three years of such growth, shrinkage in the NonStop community seems to have
quiesced, a situation that I plan on covering in more detail in 2013. Who could have forecast just a few years ago that a working NonStop system, complete with an SQL database, could be purchased for under US$100K? And surely, growing popularity of such a system might just prove enough for there to be a lot more marketing noise in 2013 as well!

Then again, I do get it. Just as
de Ferran emphasized value for money even as he acknowledged the need to control costs, restrict options and standardize, so has NonStop returned to emphasizing almost the exact same list of attributes. And with that, I honestly believe, we have here all the elements that have gone into why today NonStop remains a relevant technology, important to so many businesses. All I can hope for now is that more of the NonStop community gets it as well and about that, I will remain every bit as optimistic as I have ever been.

Friday, December 14, 2012

maRunga, in the sky

Everyone picked up on Google’s clouds misfortune, but back on July 2, 2012, New York Times reported: “On Friday night, lightning in Virginia took out machines that were part of Amazon Web Services, which hundreds of companies use for data storage and computation. Well-known sites like Netflix, Pinterest and Instagram were not accessible for hours.”

On October 22, 2012, GIGAOM reported: “Here we go again. Problems with Amazon’s Elastic Block Storage (EBS) service have brought down Foursquare, Reddit, Heroku, and other popular websites. Once again, Amazon’s U.S. East data center in Virginia is ground zero for these issues”

Back in 2011 Earnst&Young published a paper “Cloud computing issues and impacts”, as part of their Global Technology Industry Discussion Series. Interesting reading:

http://www.ey.com/Publication/vwLUAssets/Cloud_computing_issues,_impacts_and_insights/$File/Cloud%20computing%20issues%20and%20impacts_14Apr11.pdf

So, unguarded clouds may get us all in trouble? I think that. What about the guarded ones? HP Clouds program does not address NonStop participation in the solutions. Not yet. We hope to change that with maRunga!

We spent a few hours last week discussing the name for a new InfraSoft product that will provide a shield against unexpected failures of the computers that are providing the resources as part of the cloud. It is no longer “pie in the sky”; it is a well-defined project, with plans, deliverable and a name!

Choosing a name is always fun; it needed to be an Australian Aboriginal word so that it is clear that the product comes from the same company that brought uLinga to market; yes it will share the underlying framework with uLinga, of course. The Team voted, and it will be maRunga, a word based on the Aboriginal word for cloud.

Building maRunga will not automatically promote NonStop systems to playing a prominent role within enterprise clouds and may only attract a select few – but after reading the reports being generated following the outages we have all seen recently, there certainly cannot be any harm in equipping NonStop to play a role in Clouds, so stay tuned and watch for more details in the coming year.

Wednesday, December 5, 2012

Big News on Big Data?

What has kept NonStop entrenched in our data centers for many years – its unique attributes when it comes to running mission-critical transaction processing – ensures that it will not be overlooked when it comes to building out Big Data frameworks!


Having the RV as our “command center” when we travel is giving us new freedoms when it comes to being able to be flexible with our travel plans. Any time we need to head back to the west coast, there’s now few issues preventing us from stopping by good friends at HP and to catch up with former colleagues from our times at Tandem Computers. With the recent move to new facilities, within the HP campus at Palo Alto, we were able to enjoy an unexpected pleasure – visiting for the first time the well-maintained original offices of both Mr. Hewlett and Mr. Packard, and the photo above was taken of us alongside the desk of Mr. Hewlett.

Big occasion and a big highlight! The street and city were one of the first US addresses that I came across – anyone who worked with IBM mainframes back in the 1970s should recall that manuals came out of an IBM distribution center with the address of Page Mill Rd, Palo Alto, and when the opportunity came for a drive down from Edmonton, Alberta, to California I went out of my way to find this street and to drive past the IBM building which, if I recall correctly, was on the opposite side of the road to where HP now hosts its current campus.

The more things change the more they stay the same – yes, all those decades ago I didn’t flinch at an 8,000+ mile road trip – my route from Edmonton to the Bay area was via San Diego and Dallas / Fort Worth. Big occasions, big trips, and of course a visit to the Big D itself! For someone from the Big Country, what else would you have expected? Whenever we label anything as being big, it goes without further explanation that it is out of the ordinary, sometimes even beyond superlatives, and possibly so immense that its significance just cannot be overlooked. It’s important.

All of this is a lead to the topic that has occupied much of my time over the past couple of weeks. Big Data. In the time I have been looking at what it means for the NonStop community – and yes, there are still those that question the relevance and likely impact on NonStop – I can’t escape the feeling that this holds great potential for even greater NonStop visibility and I cannot dismiss the intentions of those advocating for the inclusion of NonStop in any plans for Big Data. It’s all about the need to include transactional data, of course, and the developing acceptance that simply capturing unstructured data without any connection to what was actually going on at the time, leaves a lot on the table in terms of understanding customers. Knowing they purchased a jacket, based on a discount coupon delivered on a smartphone, holds a lot more significance if you know that the purchase transaction occurred within half an hour of the coupon appearing on your smartphone than knowing only the date.

In my last three “Musings on NonStop!” features, published in the Tandemworld.net eNewsletter, I have focused on Big Data starting with the observation in September of how there is a lot going on for the NonStop community when it comes to Clouds, but there’s also a lot happening when it comes to Big Data – it’s how companies will be able to better compete in the marketplace and to provide business unit managers with even greater insight into all that is happening within specific market segments. I followed this up in the October musings with the observation of how
industry experts seem only too keen to write-off the contribution being made by NonStop, even when they reference transaction processing. And yet, when you look at the data, it’s hard to overlook the presence of NonStop. Yes, I wrote rather tongue-in-cheek, critics of NonStop seem to be fond of promoting other platforms directly interfacing with customers even when their tolerance of failure seems pretty weak.

In the November musings I again referenced the work I had been doing for Attunity, in addition to that for IR. And they aren’t my only clients with an interest in Big Data – of late I have become associated with a start-up developing new ways to integrate data from many platforms to provide a better mix of relevant data for analytics programs and they are starting with the transactional data coming from NonStop. Surprised? I was, but as I talked to them, it made a lot of sense. The intersection of data generated by mission-critical transactional applications with the data coming out of emails, social media networking and other sources of unstructured data just makes sense. And if you think the Tandemworld.net eNewsletter is just a collage of vender promotions, you may want to take another look as there are interesting commentaries scattered throughout this publication well worth checking out.


But it is the work Attunity has been doing that has kept me focused on Big Data. Even as my previous opinion paper on Big Data became available as a download at
http://resources.attunity.com/HP-NonStop-transactional-data-important-ingredient-for-big-data-success I have complete a follow-on opinion paper that hopefully you will be able to download before year end. Whereas the central theme of the now-available paper looks at the challenges facing companies, as the landscape of servers that typically populate a data center continues to include a diverse mix of heterogeneous systems, the theme of the just-completed paper looks at the impact of an increasingly mobile client population, initiating transactions at any time, will have no plans to ensure transactional data is a part of Big Data. Moving from “what do you know” to “you know too much” to “yes, you can take it with you” provided an interesting backdrop for the material provided in these two opinion papers.

“We’re seeing Big Data being managed and processed in three types of Big Data stores,” Attunity’s Ankorion explained during the recent interview and covered in the previous opinion paper – again, follow the link above. “These include large-scale Data Warehouses that have been available for some time, cloud storage as we have today from vendors such as Amazon Web Services, and Apache Software Foundation’s Hadoop Distributed File System (HDFS), the primary storage system used by Hadoop applications. “For the NonStop community, the good news is that we have been actively engaged with HP’s Advanced Technology Center (ATC),” added Ankorion, “and we have already successfully demonstrated Attunity pulling data stored on NonStop including SQL and loading it to large DW such as HP’s Vertica - perhaps the very first time NonStop’s structured transactional data has been integrated with the essentially batch mix of structured and unstructured data that is in Vertica’s Big Data store.”


Equally as important was the comment posted by Justin Simonds following my initial post on Big Data of September 21, 2012, “Big Data? NonStop? It’s common knowledge … In that comment, Justin wrote “good EDW's and ODS's will continue to provide structured analytics around transactional data, of which most is useful. Hadoop and others will attempt to provide insights around social media data, most of which is not useful. Blending the two will yield better insights provided we get the proper filtering on the unstructured/social/sentiment pieces,” said Simonds. And on this point he is absolutely correct as all is said and done, there’s still no substitute for the smarts we can bring when the two worlds intersect as they are bound to do. Justin then makes the case for the Analytic Cloud where “Queries would enter the analytic cloud and either be responded to by the intelligent query router or direct the query to the EDW, ODS, Big Data analytic engine or unstructured analysis (Hadoop/Autonomy). Submitters would not know or care which backend system responds to the query." All pointing to a potential role for NonStop in the future that I am not prepared to dismiss out of sight, given the nature of recent conversations.

This then leads me to the Big Conclusion – there will be a lot of data moving around in the near term. Yes, there will come a time when we can better integrate systems, perhaps within a virtualized realm, but until then, we will be moving data. And with that, the investments being made by NonStop partners, like Attunity, hold my interest as I know they do the interest of others (including HP I suspect), and that shouldn’t be Big News for anyone who has been following the posts to this blog.