Migrating from Windows XP to VDI, Linux or BYOD. Is it a Non-Starter?

Microsoft may have given plenty of notice of its decision to stop supporting Windows XP, but what to do next is is still a major headache for many large companies and large parts of the public sector who have yet to migrate. There are a number of both Mic rosoft and non-Microsoft migration options available, in the first of a series of articles, we evaluate the choices.


According to NetMarketShare almost a quarter of PCs today still run Windows XP, down only slightly from last December’s figure of around 29%, suggesting that many organisations haven’t yet made the leap away from this unsupported OS.  Managing a migration from XP to Windows 7 (or even 8 or 10) is not for the feint hearted.

This is largely down to the complexity of the application stacks we have today. When organisations last undertook a major Windows migration it was to Windows XP more than ten years ago, in an era when the application load was far lighter. Faced with a challenging migration to Windows 7, some organisations may be tempted to look at Windows alternatives as an escape route.

While VDI, Linux and BYOD might look like an attractive alternative to trudging down the Windows Client path once again, they all come with considerable baggage.

Let’s start with VDI (Virtual Desktop Infrastructure), which uses a centralised infrastructure that administrators can host and control on a shared platform. In this scenario the desktop OS sits on a host solution and users securely access their virtual desktop from a range of different devices – thin clients, desktops, tablets. The organisation can choose to own all of the client devices, or can incorporate a BYOD (Bring Your Own Device) model - of which more later.  

The widely acknowledged benefits of VDI are economies of scale with regards to management and maintenance and reduced total cost of ownership. So far, so good. But there’s a Microsoft-branded elephant in the room: most of the VDI solutions are based on Microsoft technology, and of course most of the legacy applications used by organisations run on Windows OS.   

This throws up a whole new set of migration problems because a fully supported VDI solution simply offers a different delivery model but it won’t avoid the need to migrate important Windows-based applications to the new operating system.  And the way VDI provisioning works means that if the application is coded for  Windows it has to be published through Windows too.

Of course there are some browser-based applications that can be run on non-Microsoft-based VDI solutions that are platform-agnostic. These might work for the minority of organisations that really want to escape from the world according to Microsoft.

With this cloud-based approach to application provisioning you can publish applications that run in a browser within your organisation, allowing users to access that application through a URL behind the firewall. A rising proportion of applications work like this, but that doesn’t fix business-critical applications coded for Windows. All in all, this approach is unlikely to solve many problems in the vast majority of organisations.

Which brings me to Linux, and whether it is a viable alternative to a new Windows OS.  It isn’t. The same arguments I have put forward for needing to migrate Windows-based applications to a VDI platform hold true for Linux. We live in a business world full of Windows applications and Linux will only support these applications through a VDI solution.

So for the purposes of this discussion the fate of Linux is intertwined with VDI. You can opt to use a Linux-based PC to access a backend VDI solution but it is not a standalone alternative to a new Windows OS migration. There have been significant improvements over the years in availability of types of applications for Linux systems, and the support options allow a company to consider them as Enterprise solutions, but the orientation and retraining of staff will prevent significant uptake.

Turning to BYOD, some commentators are still touting this model as the way forward. And of course BYOD in some shape or form has been successfully adopted by a considerable number of organisations. According to a recent study of employees into mobile device usage in a BYOD setting by Ovum, “the percentage of all employees using their own devices to access corporate data is up from 56.8% in 2013 to 69.2% in 2014”.

But one of the issues that obscures what looks to be some rather rosy BYOD adoption statistics is that BYOD means different things to different organisations. For many, it can be as (relatively) simple as allowing employees to use their own mobile phone for business calls/emails; at the other end of the scale – and far less common – is requiring employees to cough up for their own laptop or tablet for use in the workplace.  

There are considerable, well-publicised barriers to adopting a fully-fledged BYOD strategy: it requires a complex strategy to implement, and opens up the thorny issues of data leakage and device ownership. For example, who pays to get an employee’s broken iPad repaired, and just how productive can they be if their access device is out of commission?

While effective strategies and policies can be put in place to minimise these issues, moving to BYOD does represent a major cultural shift requiring reviews of HR and IT & Security policies as well as the involvement of many departments, not just IT. Never underestimate the time required to plan for, communicate and fully embrace a BYOD culture.

Although implementing comprehensive security policies can secure the infrastructure and minimise data leakage, we are now starting to see a more worrying trend. Increasingly, organisations are ending up with BYOD by the back door because, for example, someone in IT has enabled fellow employees’ own devices to work on the corporate network.

Sometimes it can be as simple as the CEO wanting to use their tablet in the office. This happens much more regularly than any of us would expect and makes organisations extremely vulnerable. Saying a definitive no to BYOD is certainly a safer option than ignoring the fact that it goes on.

In summary, regardless of the destination platform, organisations cannot allow IT to risk business functionality unless there is a major shift in the way the company - and the (mainly Windows-based) applications they rely on - operate. And that is not something that will happen overnight. For now, migrating to another Windows operating system – however daunting it might seem at the start - is the only viable option.

In the next article, I will examine the post-XP Windows alternatives - and argue the case for migrating to Windows 7.

 Damian Dwyer is Practice Director, End User Computing Practice at IT consultancy ECS


"Recommended For You"

Seven reasons Windows 8 adoption in the enterprise is at risk UCL punctures Windows 8 hype with XP move