Friday, February 21, 2014

So much to read – so many comments! Where to turn?

Have you been following comments posted to this blog? To discussions started on LinkedIn groups? One recent exchange caught my eye so I pulled it into this post and added my own observations …

My day may end with reading a menu, as illustrated above, but when my day starts it is always with a review of papers and reports from well-known daily publications, including a selection from Europe and of course Australia. The scope varies significantly but includes business publications, sports commentaries, some scientific articles as well as headlines from the popular press. Over coffee, I manage to skim as many as a dozen sources and there is no real objective in mind when I begin – a story or commentary just has to have enough “special ingredients” to catch my attention.

Following the early morning reading, I turn to skimming items on LinkedIn. While I check in with Facebook, I’m nowhere near as religious about my Facebook presence as I am about LinkedIn. Staying current with the 40 plus groups I support on LinkedIn – many associated with NonStop regional groups – my interests vary considerably, just as it does when skimming conventional publications. Colleagues have often expressed to me that they don’t have the time to check out the news, conventional or otherwise, and that’s a shame – so much is being written about NonStop these days that it presents opportunities to reach audiences that otherwise may not be aware of the merits of NonStop.

It’s not until early afternoon before I finally sit down at the keyboard and write my posts and features. The influence from what I have read in the morning is inescapable and much of what I write draws from incidents and opinions I have just come across. Of the LinkedIn groups that I follow, Mission Critical Systems Forum (facilitated by Oracle) – yes, “that” Oracle – is one I rarely miss. Indeed, many of the NonStop community seem to follow as well and for good reasons; who wouldn’t want to follow a group rallying under the banner of mission-critical systems! Along with folks from the IBM camp, it’s encouraging to see comments posted by these NonStop participants, and if this group is new to you by all means think about joining and help champion the message of NonStop!

There are 21,250 members and that of itself is remarkable. However, the discussion recently started by a simple question, In one sentence, what do you think is the most important element when it comes to designing a new data centre? and it generated numerous responses in record time. Among the simple, even one-word responses, were “Energy consumption”, “Redundancy”, “Availability”, “Cost and offering “, “Green data center design”, “Why not just use the cloud?" and even “Choose the right person(s) to do it!” 

Eventually though, lists began to appear and these attracted even more discussion.

The first list prioritized the important elements as follows:

This list was followed soon after by another:
Data center facility must be redundancy and use tools monitoring and minimum in tier 2 with strong people knowledge for the DC facility operation.
System Security
Discipline process in operation should be automation for EOD process, compliance with policy, regulatory and DC problem management
Discipline execution for Technology refresh to be green DC
Have a good leader for DC Operation management Head.

Then came another list:
Environmental Acceptance (GREEN)
Scalability (Client)
Utility Diversity

It doesn’t take a whole lot of time or reflection to see that there are a number of common elements and what’s really interesting is just how relevant NonStop remains. Having checklists like this – and yes, one came from an IBM member – helps all of us, as we are challenged almost daily as to why we continue to support (and rely on) NonStop. Furthermore, it gives us confidence whenever we face our data center manager or CIO. We remain well informed, and the consensus (among this very large group) suggests that the very same reasons why we justified NonStop usage remains at the forefront of most involved in data center design, even today!

Of course, too, with a topic such as this, I couldn’t pass up the opportunity to respond. Not with one sentence, mind you, as this isn’t my style, but with a couple of supporting arguments thrown in for good measure. Even so, I pulled together my own list for this group, observing right from the outset that there’s no one item here that surpasses all else, but there’s a clear starting point, availability.

Availability? Drawing on my own commentary already made to this blog, I noted how it was a pro-football coach who, turning to an injured player on the bench, remarked, “Your best ability is your availability!” If you are unavailable, you are out of the game. Literally. So yes, this is the starting point and here, even though my first two decades were spent working with IBM (Mainframes, 8100s, Series/1s etc.), it is the Tandem / NonStop that continues to excel today … with still, despite the complexity and costs (and skill sets needed), mainframes with parallel sysplex running second, a few steps behind.

So far so good – with a little extra added in deference to my friends still relying on IBM systems, of course. Once past the availability element, I elected to go down a slightly different, if indeed unusual, path. What about Value-ability, I asked? I saw references to value (in describing important elements, previously) and that has to be close to the top as well … we are not operating any longer in the “do more with less” model, but are being challenged to provide value. Elaborating on this I returned to availability, noting that like scalability (both up and out) it has to come as part of the package. You can’t bolt availability, and indeed scalability, onto the side as an afterthought – and yes, I like the comment about making sure you pick the right person from the outset. And yes, I said, notice the sleight of hand – solve availability and you address scalability when done right!

There was an immediate “Like” flag generated (and yes, from an IBM member), but this wasn’t the end as going down the path I was taking, I then became even more creative suggesting how too there’s Insight-ability? Particularly when it comes to security! When it comes to security, I called out, we are in an arms race and no, there’s no quick fix here or off-the-shelf solution. At best, we need to know (about attacks) as they happen and have the ability to neutralize quickly; for that, I need (operational and business) insight. I need to know the “person of interest” I should follow … and neutralize.

In my correspondence with my clients and from the exchanges I have had with users over the course of 2013, Insight-ability is a growing demand. Call it business insight – it’s important, as no system is an island and no system is immune to what is going on elsewhere in the data center. With the discussions I have had with those providing monitoring as well as security solutions, at best, we can erect strong walls to better defend ourselves.

While I didn’t include Security in my list of important elements – going as far as to ask in the post,
Yet three more wishes! if it is time to revisit these (NonStop attributes) and add securability? - when it comes to designing a data center, I firmly believe that with NonStop we have a foundation to build some impressive walls. Leveraging NonStop, we can utilize some of the best building blocks available today when it comes to fortifying the data center.

What about Clouds? And what about Green? Here my question was directed not so much towards providing an indisputable response as much as it was an attempt to draw our further commentary. Cloud; not so fast – it’s a technology we can turn to augment other resources, but for enterprises it’s no panacea. I have just watched as a client moved off MS Azure onto HP Cloud on their way to building out their own data center, this time using the new HP project Moonshot cartridges (yes, those that run cold).

Again, if your business revolves around a high ratio of “look to book” transactions then Cloud may be of value. I am working with another vendor developing Cloud Bursting solutions out of (Mainframes and) NonStop into Linux, Windows and the Cloud, just for this purpose. As an industry, we have gone from paper tape to cards to mag tape to disks, and cloud is just one more iteration of this process with the added benefit this time - it’s also got processing options. Again, as with the other comments, no push back just the “like” from IBM (and indeed, the reference to Cloud Bursting solutions out of Mainframes remains an idea more than a reality at this point but was included to reinforce some degree of objectivity)!

Finally – green. This cuts two ways. It’s a fact that we need to find a location (and a source) for cheap power but it also puts the onus back on the vendors we work with – we need more solutions like HP project Moonshot. I then admitted that I was pleased to read of how IBM and Dell were looking to take up the challenge to produce something similar (even if they were caught completely by surprise and will struggle to respond any time soon). Furthermore, the recent decision by IBM to sell off its low end, x86 server, business doesn’t bode well for anything low-end appearing from IBM in the near term. However, data centers that simply don’t need power (well at least, not on today’s scale) – that’s yet another step along the path I see us all taking, I wrote in closing.

When there’s so much to read and so many places to turn to for commentaries, it can appear overwhelming. However, as I come across interesting discussions I will always look to incorporate in posts to this blog. Where I see value in joining a particular group I will encourage the NonStop community to consider joining – we need to voice our opinions if we want NonStop to remain part of the conversation. There’s absolutely no reason at all for us to be quiet – and I am sure few of you will remain on the sidelines for long. Good reading! 

No comments: