Timeline for Protection of Keys/Passwords on Virtual Hardware (XEN, KVM, VMWare, etc.)
Current License: CC BY-SA 3.0
7 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Mar 9, 2012 at 8:51 | answer | added | Yoav Aner | timeline score: 4 | |
| Mar 9, 2012 at 6:27 | answer | added | minopret | timeline score: 5 | |
| Mar 9, 2012 at 5:46 | comment | added | Dexter | Thank you for your reply - you seem to be one of the few that has given some thought to this. I agree with your rationale about cpu-access, but trying to decode cpu-instructions is of course a much more complex problem, with many issues of impracticality, than "just" reading/modifying values in RAM. Also the problem is that "legal and economic" protections are often expensive to enforce, and sometimes even sketchy to begin with, depending on who your adversary is. | |
| Mar 9, 2012 at 3:22 | comment | added | minopret | I don't have a real answer as I don't know the research or the experience. But the principle I'd apply is that if a CPU (whether real or virtual) can access your secrets, then someone who controls that CPU can access them. It's within the ability of a determined amateur to make these kinds of profound changes in the behavior of a program, such as a virtual machine host program, by modifying its machine-language code. Someone who pays for a virtual private server could usually hope to rely on legal and economic protections rather than technical protections. | |
| Mar 9, 2012 at 2:38 | history | tweeted | twitter.com/#!/StackSecurity/status/177946345057951748 | ||
| Mar 8, 2012 at 15:43 | history | edited | logicalscope |
edited tags
|
|
| Mar 8, 2012 at 13:37 | history | asked | Dexter | CC BY-SA 3.0 |