Bladewatch.com

I wonder if cloud, grid, and virtualization are not all a very polite distraction from the core issue of delivery, consistency and stabilization of the core. Of IT and service delivery management.

If we abstract ourselves from the current data center, power, support cost and strategy and work in the here and now:

What is it you want from your IT?

That it works?

That it’s virtual or grid, or comes in blue?

I would suggest that the core objective is that it works, and all too often we focus on the future, we focus on it’ll get sorted as part of project z, (virtualiation, grid, hardware refresh etc), but understand that your support cost, your call out and outages could all be not erased, but reduced through proper investment in man time in stabilizing your platforms; the network switch configuration, the cluster setups, the windows server configurations and everything else. Just some effort spent on looking at what we have, making sometimes small changes to improve settings and configurations can provide not the stability you want, but need.

Let’s remember my DL360 web server with 18Gb disk space is never going to give you the space your application needs long term, but here and now, for the next few months, we can clear down space, defragment the drive, optimize the page file, the anti virus and the other settings, and at least make it the best DL360 you have in the estate at relatively no capital cost and marginal support cost. All you need is a check list, are these settings applied, and you could even have scripted/automated jobs running everything against each server one at a time to validate the configuration. Anyway, back to the article I was originally posting about:

Information Week

While last week’s Interop event in Las Vegas showcased a wide range of innovative enterprise technologies including mobility, WAN optimization, security, storage, smartphones, data center infrastructure, servers, and more, the biggest CIO-level activity was centered around cloud computing, whose acceptance and potential are growing rapidly, and virtualization, which has already become a cornerstone in 21st-century enterprise IT strategy.

Faced with tighter budgets and the relentless charter to find ways to do more with less, CIOs today are pushing harder than ever on their IT vendors and partners to help them uncover innovative approaches and technologies that (a) let them replace old, brittle, and expensive infrastructure with new gear that’s less expensive and is built for openness, speed, and reliability, and (b) give them new capabilities and growth potential in the online-driven global economy.

Is it cloud computing CIOs or CEOs want or is it like a CIO said to me yesterday:

“I don’t care what you call it, how you package or charge it, my budget is x, you need to meet it if you want the business, I need control or ownership of the data.”

Interesting, as an analyst and a server guy, my view is somewhat different, on a day to day business, we’re fighting fires, dealing with incidents, dealing with requests and changes to meet business needs. Strategically am I thinking about cloud computing? Well I am and I’m not, you see, my main concern revolves around the simple steps, you see the steps to success to get to the point where we can contemplate cloud computing, or virtualization or grid, or any other cool things are:

Understand the environment – who owns what, what range of servers, operating systems, applications, layered products and middleware we have – what we’ve got, where are and where we should be.

Deal with the immediates – I have 100 calls, 76 requests in the helpdesk queue, arguably, that’s 176 people that need help, that need dealt with first.

Focus on the hear and now – what steps can I take to resolve the helpdesk queue – the quick wins, fix them and move on.

Fundamentally though it’s about as my good friend taught me about 7 years ago, it’s all about ”STABILIZING THE CORE. By that he meant, first of all get your area, your core fixed, whether that’s the network, the server or the printers, once that’s done, everything else will follow. It’s not glamorous, it’s not exciting, but after years of server support, I can tell you it’s true.

By getting all the NT4 servers to SP6, fixing disk space issues, reducing those unnecessary tasks, upgrading to supported drivers and firmware, fixing share permissions and moving towards a consistent platform, we get stability, we get reliability and crucially we get an understanding of the estate. I know server19 lacks disk space by 19th of every month, so I can make sure just before then, I clear down some space whilst we look for a longer term fix.

So are we thinking about cloud computing, virtualization and grid? Yes. But I wonder if we might obtain the same business benefit at lower cost, by stabilizing the core, of course we need to move on, we need to adopt new technologies, but we need to make sure, that I’m not moving a problem from one platform to another long term. Unfortunately there remains that concept, that “we’ll fix it” as we go, a fine concept, but how long, and how much cost, outage or disruption are we prepared to accept until that club class event happens, that my IT is virtualized, that my applications are all grid, web or citrix based and everything just works. Can we not do both? We get fixated on waste of money of duplicated effort, but it’s only duplicated, it’s only an issue if it’s not adding value, and we have to balance the marginal cost of having a guy sort out permissions now, to the cost of having them done on go-live when we’ve moved everything to a virtual machine and the project manager is screaming “we didn’t have this problem with the physical server”.

No related posts.

Related posts brought to you by Yet Another Related Posts Plugin.