This article is more than 1 year old

Why 1.6 million people will miss Microsoft's Windows Server 2003 date with fate

You want to do what? Again?!

Don't expect a Windows-XP support lifeline

“Hopefully people realise this is not just Microsoft sabre-rattling; the date is fixed and it’s not moving,” he said.

Camwood's Shepley reckons on an upsurge in the second quarter of this year – as Microsoft’s deadline hoves into view. He says his firm has been asked to move individual apps rather than entire estates.

But Shepley reckoned the fact so many missed the Windows XP date and haven't yet been caught out on security has bred complacency about getting off Windows Server 2003.

“There was no feared hacker who sent out a big virus or security breach that couldn’t be fixed and that brought down every machine,” he told us.

So, let’s assume you’re one of those delaying migration – or else being forced to put it on hold. What are your options?

The default is to do nothing – continue running Windows Server 2003. Thereafter you have two choices: pay Microsoft for a custom support deal or go it alone.

The former option relies on the hope Microsoft offers custom support – as it did for the many laggards on Windows XP still using the client after April 2014’s end-of-support date.

Microsoft was dragged into Windows XP custom support and – so far – has made no such commitment of a repeat for Windows Server 2003 holdouts.

The reason is simple: Microsoft wants you to buy its latest software. Having its engineers building patches for old products simply diverts its time and resources away from that goal.

If Microsoft does turn, expect to pay through the nose: Windows XP custom support started at $200 per client, doubling every year thereafter.

Claunch reckons if Microsoft does U-turn, you can expect to pay two to three times what you paid for Windows XP.

"For some customers it [upgrade] will be relatively fast now. For some – they won't transition. If it’s doing what it’s doing and they don’t worry about support, they won’t change - they will risk it.”

“I would say that most of those machines that will remain running Windows Server 2003 will not take the custom support agreement,” Claunch said. “They will be run without support.”

Hewlett-Packard’s EMEA boss Ian Stephen said plenty of customers will evaluate the risk and calculate that running an unsupported Windows Server 2003 after July is worth it.

"For some customers it [upgrade] will be relatively fast now. For some – they won't transition. If it’s doing what it’s doing and they don’t worry about support, they won’t change - they will risk it.”

That brings us back to migration; you won’t hit that July deadline, so at this point it’s a question of priorities: which apps must be moved and which can go later?

Then it’s a question of: move to what exactly? New physical servers, virtualised servers, or bumping workloads up into this cloud thing people are talking about?

Stephen reckons big firms with lots of server-based apps are swapping hardware – going to yet more servers.

Mid-sized companies are swapping server-based collaboration apps for Microsoft’s Office 365. It is less clear the direction small firms will go, Stephen says.

According to Stephen, HP is working with Microsoft on that mid-market segment that is running lots of different apps on different servers.

He reckons the mid-market is most troubling: firms lack the full IT complement of the enterprise, which can more easily steward a move but has more IT to manage than a small firm.

“The process of doing migrations [for the mid-market] is more intensive,” Stephen said.

If you’re starting that move, the first step is to ascertain exactly which apps you’re running – numbers and type – to lay the groundwork on what can and should move.

Shepley says people seriously underestimate the number of apps they have running on their server estate – the product of years of updates as users and IT staff leave. Sometimes apps drop out of use or those who built them move on.

“Last time we did a software audit for one client they said they had under 1,000 apps. We found more than 2,000,” Shepley claimed.

Once you can see what you have, the next step is to assess what to move and the target environment: do you virtualise, buy new hardware or put the app in the cloud?

Cloud is no get-out-of-jail card

It can mean virtualisation or tracking down the online version or equivalent of your legacy apps supported on, say, Windows Azure or Amazon Web Services.

Zynstra has helped move Microsoft shops to a combination of Office 365 and Windows Azure. The answer for more than one-third of Windows Server 2003 users is to simply upgrade to "the cloud", but cloud isn’t that simple – you can’t simply turn off your Windows Servers and install an existing app on somebody’s cloud. Migration isn’t like swapping like for like.

Going to cloud is a strategic move – the kind of action that you should take having weighed the pros and cons of seeing off part of your IT infrastructure to a third party.

It’s like a more accessible form of outsourcing, so prepare accordingly.

The skills needed will seem slightly alien: it’s still very much a developer mindset up there, with apps built for things like elasticity and scale. Microsoft and partners like Finn have been running virtual academies for IT pros.

At this point, though, Gartner reckons the opportunity for the cloud has sailed, given the strategic decisions and architectural choices it involves. That might take a year to pull off.

“The time to do dramatic changes like moving to a hybrid cloud market was a year or two ago, and now it’s more of survival, get the job done,” Claunch said.

The message is simple: don't get seduced by the rhetoric of cloud. It's not a quick or easy answer.

That leaves moving from one server to another and - having audited your apps - it again comes back to deciding which are the priority before upgrading and moving in batches.

If migration is still in your future, so is missing Microsoft's deadline. Nobody who hasn't already started migrating will have made a clean break by July. How far you miss will depend on your appetite for risk and your long-term IT strategy. ®

More about

TIP US OFF

Send us news


Other stories you might like