Earlier this year (2013 – for those reading this in years to come!) I spent a lot of time on a project using Microsoft virtualization, using the Labs Management feature of TFS 2012 to deploy (replace) virtual and physical “multi dimensional” (i.e. volume x type) test labs on demand. Lab preparation time went from hours to seconds, on both Microsoft and non-Microsoft hypervisors (as well as on ‘tin’), greatly increasing the test throughput.
A subsequent project then looked to use the same capability to deploy flexible “build” (i.e. compilation) capacity. The ability to turn on and off (virtual) machines “on demand” (in this case “build” demand, but the demand could be web or LAN/user traffic) is a powerful one that helps leverage better value from infrastructure spending. Add to this suitable monitoring and scaling tools (such as Electric Cloud) and you have the option to expand automatically and “burst out” e.g. to cloud providers to satisfy extraordinary peaks in demand.
Recently I’ve been using this experience to advise a client on a strategy for virtual desktop infrastructure, another hot topic and one were the experiences described above start to add value. Virtual desktop infrastructure (VDI) is something Microsoft are really getting into and can bring real benefits. VDI allows for users to have “clean machines”(i.e. laptops) of any type with minimal deployment footprint. Everything the user subsequently needs to do – access corporate systems such as email, SharePoint, etc., co-work/virtually meet with colleagues, access a “personal” space – is done virtually, on virtual machines.
For this particular client the ability to tear up and down development environments (across many stacks/frameworks) was key – and demand for this changed depending on release cycles. Buying infrastructure to support peaks in demand just didn’t make sense.
Bringing all this together, a labs management-esque approach to deploying clean machines and productive virtual environments, supported by flexible and demand driven infrastructure (i.e. in-house and Cloud) based on DevOps (infrastructure-as-code) approaches can really bring performance, reliability and scalability benefits to any organization deploying large volumes of desktops (but don’t forget single sign on!). Oh, and disaster recovery and back up becomes easier too.
A final benefit of this strategy worth mentioning is the proliferation of virtualization access applications on modern smartphones and tablets. The ability for staff to access their productive corporate (virtual) machines from any device and from any location more or less at any time brings organizations a multitude of benefits – more comprehensive out of our support and hotfixes, home/commute/café working, instant idea capture and realization, to name a few.
At CAP Project Services we offer impartial advice on new and emerging uses of virtualization, using Microsoft and non-Microsoft (i.e. Linux) technologies through rapid and continuous deployment and availability processes and tools. For further information or to get in touch email info@CAPprojectservices.co.uk