Monday, October 10, 2011

When you push the buttons, it better be NonStop!

There are still many situations where the reliability of NonStop saves us from misfortune and discomfort. At a time when so many NonStop users aren’t sure of what to do to modernize, there remain very valid option in the marketplace.

I read, I travel, I observe, and I participate in events. I also provide commentary and opinions in many forms. This shouldn’t come as a surprise to many, and it represents the very core of my business pursuits these days. It has been noticed by many of us that comForte CTO, Thomas Burg, has been travelling as well, as anyone following his many tweets, Facebook updates, and LinkedIn texts can attest to. Contact with the wider NonStop community is essential to fully appreciate all that the NonStop community has achieved.

It’s also a very important element in determining what’s important to the NonStop community. While the marketplace for NonStop always generates serious debate from within the NonStop community, there are many who argue strongly that there are many market segments, not traditionally associated with NonStop, where NonStop can more than adequately meet the needs of the businesses operating within those segments.

It was just this week that Marty Edelman, well known to the NonStop community and formerly a key contributor at sites like Home Depot as well as SIAC, posted to a Yahoo Groups community that “I do not only 'tune' myself with HP. I don't think what they are doing is perfect. I think they still have some work to do but I do know that for a mission critical application there is nothing more reliable than the NonStop, nothing! Go ask Amazon, Ebay, Google, ... they do not run 5-nines they would kill to get to 1-nine.”

Later in his post, he added quite succinctly, “I don't think a 911 system or an ATM system or the mobile phone system can afford anything less than 5-nines, do you? Would you like to be the one laying on the floor having a heart attack and you discover that your mobile phone isn't working and even if it was the 911 system was down? I know I would not!!!” Even in our very modern society of today there continues to be very important roles where the NonStop Server provides enormous value.

This week I revisited a paper “Legacy Software: Rebuilding the Ship in Rough Waters” first published in 2009 by Aberdeen Research, an organization that pulls data from a client base of over 500,000 contacts. What caught my eye were two simple observations, the first being “when it comes to a software development environment, companies essentially have three options to respond to market pressures. First … stick with the status quo; second, software modernization through migration, upgrade, or other means; lastly … ‘rip and replace’ methodology … typically, the most costly option.” Aberdeen Research that added that “companies who have embarked upon a formal strategy for legacy modernization are seeing demonstrable performance improvements over those sticking with the status quo.

Readers of this blog shouldn’t be surprised by what Aberdeen Research reference as this has been the subject of several recent posts. Modernization has been a reoccurring theme and the value that comes with modernization has been emphasized repeatedly. As far back as the post of July 8th, 2011 “Hiding in plain sight!”, I wrote of how modernization is all about reducing costs and in the process, better aligning with the company’s business and IT strategies. I observed too how the process of modernizing helps bring back focus onto innovation, and with innovation to become more responsive in a changing global economy. Yes, as Aberdeen Research had found just a few years back, software modernization through migration, upgrade, or other means is producing the most demonstrable performance improvements.

However, it was the second observation in the paper by Aberdeen Research that really grabbed my attention, as what they say is so often overlooked these days. In a reference to an earlier benchmark report, “Modernizing Legacy Applications: Maximizing the Investment”, it was shown that “best-in-class companies were largely taking a tools-based approach to legacy modernization and were leaning on the domain expertise residing in third-party organizations.”

Again, readers of this blog may recall the post of August 18th, 2011 “Modernize? It requires more than a re-set!”, where I suggested that to transition successfully to modern IT it’s not just products but the leveraging of experienced partners and the services they provide. Yes, selecting vendors as partners and “leaning on the domain expertise” that have accumulated remains the most effective way we have today to modernize remaining legacy applications. comForte continues to tune its marketing messages, and in the coming months visitors to the comForte web site will see more attention being given to modernization – including “migrations, upgrades and other means!”

Yes, I read, travel and in general stay close to what the NonStop community is doing and what they view as a priority, and all the while the continuing trust that the NonStop community places in the NonStop Server platform shows little signs of waning, despite the outcry’s of the naysayers. As Marty so well expressed it “would you like to be the one laying on the floor having a heart attack (only to) discover that your mobile phone isn't working …”


  1. ATM systems tend to be Crit 2. If an ATM is down you can get your cash out later. If a POS system is down you'll clog up the supermarket lanes, abandon your shopping trolley and the store will have to bin the perishables and pay overtime to restock the canned goods. Thus POS systems tend to be Crit 1.

    You don't need to have a NonStop to get that level of reliability but it does make it easier. The question is whether that ease is worth the pain of running a non standard platform.

  2. Yes, agree - and it's this "easy to do" that needs a lot more emphasis in my opinion. Not only initially, but then maintaining it all? When architects and main technicians move on who really is expected to look after it all ...

    This is becoming a little more appreciated as we continue to see more systems, clustered or otherwise, come down (routinely) so yes, let's become a lot more vocal about this and not shy away from keeping our management well informed!

  3. Not sure why this came up as annonymous but yes this is me, Richard, author of the original post!