soren

There's a paper on live updating Xen to be presented next month at a cloud computing security workshop: https://www.trust.informatik.tu-darmstadt.de/publications/pu...

Similar work was done in the OpenTC project (http://www.opentc.net), i.e., running legacy OSs in isolated compartments and multiplex their visual interfaces using a "secure GUI", which labels the windows/interfaces according to certain properties. It also supported OpenGL in the "AppVMs", which is currently omitted in Qubes OS.

The isolation of drivers in separate VMs and enforced isolation using VT-d is definitely interesting. A paper on the disaggregation of dom0 for improved security is http://www.xen.org/files/xensummit_fall07/22_DerekMurray.pdf

An example for a microkernel running a Linux OS is L4Linux (http://os.inf.tu-dresden.de/L4/LinuxOnL4/overview.shtml).

There is a lot of controversy about the distinction between VMMs and microkernels. A view from the microkernel side is given in http://www.ertos.nicta.com.au/publications/papers/Heiser_UL_...

There's a list at stackoverflow: http://stackoverflow.com/questions/38210

I contacted someone at AWS through a friend and asked him basic architectural questions, like if he can confirm that they are running Linux in Dom0. However, AWS can not give such details as long as you do not have a valid business need and sign a NDA, which I declined. If they are so closed about basic questions about their architecture, I doubt they will give in-depth details.

I totally agree. The intention of writing this post was not to provide a precise documentation of Amazon's architecture (only Amazon can do this), but to get feedback if I am on the right track to understand the underlying architecture.

"more text in an eye fixation" is not necessarily a good thing for reading code, because code is semantically more dense than regular text. I think one will not be able to absorb all information and therefore not get a faster reading experience due to re-reading. I find the fixed-width font example easier to read, because it appears not as cramped as the proportional-width example, and I can easier get an overview of what the code does.

I uploaded a tool I've written in SML for generating space-filling curves (Hilbert, Peano, Sierpinski) with postscript output on github: http://github.com/soeren/snippets/tree/master/spacefilling/

The difference is that they completed the computation of the rainbow tables, which are required to make this attack practical, and made them public (although the webserver with the torrent files is currently down: http://reflextor.com/torrents/). Of course theoretically this attack is known for quite some time, but making it practical by providing the equipment and rainbow tables is new.

Interesting enough they planned to do a workshop today, where you could bring your GSM data and they wanted to try to decrypt it. However, due to legal reasons they had to cancel it (http://events.ccc.de/congress/2009/wiki/The_demonstration_is...).