Virtual Desktop Infrastructures (VDIs) are hype. But are they really a strategic element of IT? Or are they just a deployment option? I think that the answer is influenced by two major aspects:

• Time and the maturity of Desktop Virtualization
• The functional breadth of VDIs

With respect to the first aspect, VDIs today are more sort of a more expensive, more complex alternative to Terminal Services. Less users per server, the same (sometimes a little bit more advanced) protocol for remote desktop access, very limited capabilities to run the VMs locally on a hypervisor – VDIs aren’t really mature yet. However that will change. We will see more deployment options, improved management capabilities, some improvements regarding performance (however, VDIs will always be expensive in terms of compute power at the server), and so on. And especially with different local deployment options (streamed, synchronized), the need for remote desktop protocols will disappear, mobile users will be fully supported and less servers will be required – without giving up advantages like the (relative) independence from hardware and some centralized management aspects (which are, however, not that different from other deployment approaches).

The other aspect is about management. Is isn’t sufficient to integrate the management of server and desktop virtualization – and even adding storage virtualization management to that is not enough. Application virtualization has to be integrated as well. But even then we have some lack of capabilities:

• There will most likely be other types of desktops for a pretty long time – the more specialized ones for “power users” and “knowledge workers”, for specific user groups like engineers or stock brokers, and so on. It is not only about the 50% or 80% of desktops which fall into few standardized categories. The main issue are always the remaining 20% or 50% of not-that-standardized desktops. And they have to be managed centrally as well.
• That requires configuration management and software deployment beyond building few standard images. Image management in reality is far more complex than just having few standard images. And not every application can be virtualized. Beyond that, we need several other elements which typically are found in Client Lifecycle Management today: Think about inventories and License Management. With other words: You will either need Client Lifecycle Management (CLI) or VDIs have to fully integrate that in the future.

In the future, a more complete VDI stack with full CLI support and optimized support for local deployments and mobile users might become the standard – even for older operating systems and non-Windows platforms. For the meantime, it is probably the better strategy to understand VDIs as one deployment option amongst other and to integrate all these deployment options under centralized management system. At least it is a good idea to be realistic about VDIs and not too enthusiastic.

So I’m a believer in VDIs – but I’m a sceptic regarding their short-term value for most use cases. What is your opinion on this?

One of the IT market segments I’m observing for a quite long time ist the System lifecycle management market, including software distribution, OS installation, inventory, patch management and some other technologies. There are few segments which are that crowded. If I count the vendors/brands which compete in the central European region I end up with something aroung 20 at least. Given this number of competitors it is obvious that not all of them will survive. There will be the big ones to survive – and there will be the smart ones.

Read the rest of this entry »

This afternoon I had an analyst briefing with one of the vendors which offspring the market segment formerly known as system management which is usually defined today as client or system lifecycle management. This change has been definitely necessary because system management covered a very broad range of different technologies.

But the system lifecycle management segment, which exists for a while, is as well pretty heterogeneous. There are vendors which still mainly support software distribution, OS installation, patch management, and some other administrative functionalities. There are vendors which are moving towards the security market, like LANdesk with their NAC products or Symantec. Many vendors are adding license management capabilities and move towards the ITSM (IT Service management market) or at least to some part of this market – Enteo/Frontrange as one example -  whereas others set their focus on compliance and related topics. The borderlines aren’t always clear. There are many vendors which claim to support license compliance. But there are few which really cover all the details of licenses and which integrate their license compliance tools as well with asset management and automated inventory services as with contract management. In this area you’ll find Managesoft as well as Brainware.

But even for these vendors, there’s the question about how long the niche will exist. The approach of ManageSoft is pretty interesting. They are providing a strong technical integration with the inventory and asset management as well as a dashboard for the business user and IT management. Thus, they might move towards more controls they support in this “compliance dashboard”, they might add risk management functionality or they might do both.

But ManageSoft, like every other vendor who has successfully done the step beyond the administration-focused system lifecycle management, will always have to find new niches fast - because other companies will enter interesting market segments and because the big BSM players always will try to position their solution as the “swiss army knife” you can use for everything.

My observations of the vendors in the system management space over the last years are, that there are some vendors which are able to reinvent themselves. There are vendors who try to grow through acquisitions – not always successful. There are the big ones which sometimes struggle when it comes to the details and still need support of smaller specialists who are able to fulfil the customer’s demand in the context of an enterprise framework. And there are many companies which are neither able to reinvent themselves (at least not fast and innovative enough) nor to grow through acquisitions. In a market segment like system lifecycle management with more than 20 active competitors in Europe – not counted the ones in other areas – the ability to move forward is one of the most important aspects for product decisions. It is because the ones who aren’t innovative are the ones who will in the best case become acquisition targets and in the worst case just will disappear.

Thus, it isn’t done with re-positioning in a newly “invented” market segment which is just a new name for something existing. It is about re-inventing the market segment.