Virtual computing - I'm finally catching on

At work we have been using virtual servers to manage multiple installations of Institutional Repository software on a single physical server. It's like we have our own room full of computers. I should have been doing this years ago.

I had been planning to write something here about the joys of virtual servers, but two better qualified people from the RUBRIC technical team beat me to it.

1. Caroline Ayers writes about her experiences regarding the RUBRIC project on her blog. Caroline writes that virtualisation has worked out very well for us in hosting complete self contained demonstration systems for our partners:

   This has turned into being a beautiful solution to what we’re trying to achieve. Not only do we have the flexibility to offer as many instances as we need, but we’re now also able to create these new instances in just a few minutes. If a request for an instance is put in in the morning, we’re able to (usually) have a running functioning instance by the afternoon. This just would not have been possible if we were relying on physical infrastructure. Stay tuned for more on Virtualisation and Irs.

   http://sophiaca.wordpress.com/2006/04/21/virtualisation-and-irs-–-part-1/

2. And Corey Wallis has added a technical report to the RUBRIC site. Won't be hard to find it's the only one there for now – lots more to come soon.

By coincidence Rick Jelliffe just wrote about his experiences. Rick notes that it now doesn't really matter what operating system you're using. I agree – I use Mac OS X, but the main applications I use, like Firefox and OpenOffice.org are available on Windows and Linux as well and like Rick I use Eclipse for programming. (I do like Apple's Mail application, mainly because of the fast searching but I hate that it and iCal will not play nicely with the Microsoft Exchange server used at work).

Rick speculates about where this is taking us:

It is interesting that virtualization in a sense reduces the operating system back down to a very small set of primitive intercommunication points. When the operating system is effectively so small, a lot of what passes for operating system today effectively is promoted to being session or application software. And this happening at the same time as the BIOS makers are revving up BIOS to be a little operating system in its own right squeezes the PC operating system further. Is that where we are heading: a superBIOS with PNP drivers hosting a thin basic OS (stripped down Windows, ReactOS, JavaOS, Linux, OS X, etc) with super fat clients running in virtual operating systems (perhaps packaged as virtual appliances.)?

http://www.oreillynet.com/xml/blog/2006/04/the_super_fat_client.htm

As for me, I have just dipped a toe in the virtual water with a (free-as-in-beer) VM Ware 'player' virtual machine running on a Windows XP box at home.

Why? I am considering getting an entire machine (either physical or virtual) hosted for me so I can host ICE repositories, and try out an idea I have for an online word processing to HTML conversion service. The service provider I was considering offered Debian Linux, although I would have preferred Ubuntu.

So to practice I started with a minimal Debian virtual appliance and let it install a full desktop system via the 'net. Took all day, including me doing stupid stuff like killing the network connection in the middle of the install and the resulting virtual computer was stupidly slow until I added an extra 512MB of RAM to the feeble 256 I had in the physical machine.

Now it turns out the ISP does offer Ubuntu – must have just added it – so I'll try again with Ubuntu. If all goes well, then I can probably turn off my $25 Linux computer soon – it's sitting here heating the home office and not doing very much – why not just run it off the Windows machine that gets left on all the time anyway and give the old computer to someone else? Someone else with extremely modest computing needs, that is.