Philosophical about virtualisation

May 4, 2009 – 5:28 pm

I stop and ponder things from time to time. Get philosophical about why we’re all charging like lemmings after buzzwords without stopping to wonder why or if we’re really going in the right direction. I’ve seen people end up in dead end careers because they were too near sighted. I dont want to end up like that.

This article reminded me about a particular topic i ponder from time to time; Virtualisation.

Google fires back at VMware

I’ve known about VMware since it began yeeeeears ago. back then, i saw it as a developer tool, for people who needed to try various operating systems but having multiple hardware boxes wasnt practical.

fast forward to the here and now, and virtualisation is all the rage, with VMware being top dog.  I’m from a telco core systems background, which is pretty much pure unix, of varying flavors.  vmware is almost exclusively used by SME, who are all windows shops.

My opinion is that the dominating reason for vmware being such a huge success is the nature of windows and the apps that run on it.  windows server apps dont play nicely with each other, thus the resulting plethora of servers, each running one or two apps, all under utilised, with headroom for peak usage.

In the unix world i’m used to, multiple apps all run happilly alongside each other.  the first big production systems i worked on were a pair of Sun enterprise 3000 servers. they ran mail, news, web, proxy, dns and some other minor apps. all on a single instance of Solaris. the apps were installed on external SAN disk and could be failed over between the two servers. this was in the late 90s.  ten years later, vmware is bringing to the windows world the efficiencies and high availability features that has been available for a long time in the unix world.  that said, even the unix world lagged behind the mainframe world.

that got me thinking, why the move from mainframe to unix to wintel?  its about $.  when “open systems” aka UNIX started to beat the pants off mainframes, it was because of cost. the systems were good enough for what they were needed for and a lot less expensive. the fact that the technology wasnt as impressive didnt mater. the business case, weighing everything up tipped in favor of unix.   later, for the SME, the same thing happened between windows and unix. windows was “good enough” and required much less skill to work with. many companies couldnt afford the skilled sysadmins. their wages would cost the company more than the losses due to system crashes/outages.

vmware helps companies run their windows environments in a much better way, without changing the way those companies’ apps work.  google helps companies by offering a whole new way of using software. new applications. gmail instead of Exchange/Outlook.  Google Apps instead of the MS Office Suite.

as a result, system/hypervisor virtualisation is not needed by Google, they virtualise the apps.  the fight is between Google trying to convince companies to consume IT apps in a new way and VMware trying to convince companies their existing way is fine, but that vmware can make it run better.

pass the popcorn 🙂 this should be fun to watch

NOTE: VMware and Google are just examples. there are other players in each of their respective markets, they’re just the leaders in my opinion.

  1. One Response to “Philosophical about virtualisation”

  2. One of the reasons for the move from mainframes to unix, was that you only had a very small number of mainframe vendors.

    If you didn’t like what your mainframe vendor was selling you, you were SOL. In the midrange world, there was competition because unix was “open”, or certainly more open than MVS or Z/OS.

    Expensive stuff is just par for the course when you had to buy from the one supplier.

    Windows changes that in so many ways because it’s so much cheaper to develop on and the customers are way less demanding.

    VMware helps customer get on top of the issues that arise from using software that’s so busted.

    And yes, as you say, Google is trying to change the paradigm in which all of this activity happens. It’s, in a sense, teaching organisations and individuals that you can outsource the whole kit-and-kaboodle.

    Plenty of food for thought.

    By Francis on May 6, 2009

Post a Comment