Skip to content

History of Xen – Architecture – Part 4

The final installment of the History of Xen – Architecture involves the ultimate question, where does the name “Xen” come from? It is clear that Xen comes from the XenoServer project at Cambridge which is the research that the Xen hypervisor emerged from. The name “Xeno” for the XenoServer project is specifically mentioned in the Controlling the XenoServer Open Platform (Nov 2002) as footnote1:

The name derives from the Greek word “”o&” (xenos), which means
foreign or unknown, much like the tasks that XenoServers accept and safely

So, the name Xen comes from XenoServer with a Greek origin; but who was the first person to claim the name “Xen”?  If anyone knows or wants to make the claim, here is your chance.

Be Sociable and Share!

Posted in Community, Xen History.

Tagged with , , , .

3 Responses

Stay in touch with the conversation, subscribe to the RSS feed for comments on this post.

  1. crowcroft says

    we were standing by the coffee machine in the netos corridor (ian, steven, keir, me) and discussing the title for the SOSP submission
    on the art of virtualization – so I said – xenoservers – xen – ah,
    xen & the art of virtualisation. (sp :-)

    it stuck.

    i have no shares in citrix or vmware:)


  2. crowcroft says

    so the idea was due to thinking about Isolation (a crucial property of any good VM) – I was thinking, originally, of Zen and the Art of CPU Cycle Maintenance, since the paper in progress at the time (around jan-feb 2003) had more about schedulers for splitting CPU amongst guest OSs – this derived from earlier thinking in the Nemesys (pegasus II) project/system where Isolation (hard multiplexing of resources) was a key requirement – indeed this is still “work in progress” in xen:)

    The name was of course from Pirsig’s book (essential Geek reading) Zen and the Art of Motor Cycle Maintenance (along with Cryptonomicon, by Neal Stephonson, which has an ace spoof “Business Case for a Startup”, and Alice’s Through the Looking Glass, an endless source of Computer Science ideas for 100 years). The Zen/Pirsig book has some philosophical meanderings (Chautaqqas) on Platonic Ideals and realities, which matches well to “real” versus “virtual” and “para-virtualization”).

    Why was I there (I do Nets, not really OSs, although in the Systems Research Group, we all mix and match)? Well, we all supervise lots of students in a grand melee and we all read each others draft papers to get an idea of what each other are up to and help keep the quality up – several of my PhD students were doing Xen related work – indeed, Boris Dragovich ended up a co-author on the paper – julian chesterfield and tim deegan both ended up at Xensource too, plus I was on the microsoft research cambridge lab’s tech advisory board, so knew about the heavy lifting paul barham was doing to get the windows xp kernel to speak xen’s paravirtualized VM API (which he had to do without looking at any of Xen’s GPL’d code, due to Microsoft’s lawyers’ paranoia – a heroic job), plus I also worked with Intel research and knew something about Rolf’s work on said scheduling (for isolation )….

    the Xen grand plan continued to unroll from then with ian pratt’s customary military precision, including truckloads of HP and other folks showing up for master-classes in OS porting and driver virtualising

    recently, I was happy to see Ian doing the keynote at NSDI 2008, where also, a best paper award went to students and one Andy Warfield at UBC, another phd output of the SRG, who worked on virtualisation of storage, networks, and more recently, awesome performance for OS migration….

    the rest is, as they say, history…

Continuing the Discussion

  1. » Xen Name History Finally Solved linked to this post on May 7, 2008

    [...] For those of you tracking my “History of Xen” blog series, I am proud to announce that we finally have the story of where “Xen” came from – check out this comment.  [...]

You must be logged in to post a comment.