| |||||||||||||
Cryptographic weakness on Debian systemsCryptographic weakness on Debian systemsPosted May 13, 2008 14:14 UTC (Tue) by neiljerram (subscriber, #12005)Parent article: Cryptographic weakness on Debian systems
How do we know that these latest announcements have really come from Debian? (I'm asking out of genuine ignorance, and I expect that that is probably a very dumb question, and that there's a simple answer. But consider, with --paranoia=max, what would you do if you wanted to attack a massive organization like Debian? Forge some emails, claiming that all of the keys with which DDs are familiar (recognizing fingerprints and such like) are now invalid...?)
(Log in to post comments)
Cryptographic weakness on Debian systems Posted May 13, 2008 14:33 UTC (Tue) by bcl (subscriber, #17631) [Link] Because the announcements are GPG signed. http://lists.debian.org/debian-security-announce/2008/msg... gpg: Signature made Tue May 13 05:03:24 2008 PDT using RSA key ID 02D524BE gpg: Good signature from "Florian Weimer (HIGH SECURITY KEY) <fw@deneb.enyo.de>" gpg: WARNING: This key is not certified with a trusted signature! gpg: There is no indication that the signature belongs to the owner. Primary key fingerprint: C8D3 D9CF FA9E 7056 3F32 FA54 BF7B FF04 02D5 24BE If you don't have the key, you can import it like this: gpg --keyserver pgpkeys.mit.edu --recv-key <keyid> Since there is no web of trust between me and the owner of this key there is still no way to guarantee that it really belongs to Florian Weimer other than checking it against other posts to the list.
Cryptographic weakness on Debian systems Posted May 13, 2008 14:39 UTC (Tue) by elanthis (guest, #6227) [Link] "How do we trust this announcement about weak keys is real?" "We look at the keys they used in the annou... shit."
Cryptographic weakness on Debian systems Posted May 13, 2008 14:44 UTC (Tue) by maks (subscriber, #32426) [Link] read the announcement gpg is not affected. openssl is bad enough!
Cryptographic weakness on Debian systems Posted May 13, 2008 17:43 UTC (Tue) by neiljerram (subscriber, #12005) [Link] Thanks everyone for your answers. I see now that GPG keys are in a separate space from the ssh keys, and unaffected.
Cryptographic weakness on Debian systems Posted May 13, 2008 21:03 UTC (Tue) by lab (subscriber, #51153) [Link] Hmmm.. Can I just ask a stupid question - how come the OpenSSH package in Ubuntu is affected, but not in Debian? http://www.ubuntu.com/usn/usn-612-2 "A weakness has been discovered in the random number generator used by OpenSSL on Debian and Ubuntu systems. As a result of this weakness, certain encryption keys are much more common than they should be, such that an attacker could guess the key through a brute-force attack given minimal knowledge of the system. This particularly affects the use of encryption keys in OpenSSH."
Cryptographic weakness on Debian systems Posted May 14, 2008 0:16 UTC (Wed) by cjwatson (subscriber, #7322) [Link] It's affected in exactly the same sense (i.e. only as collateral damage) in Debian too; unfortunately problems due to the advisory itself have made it difficult to publish an OpenSSH update in Debian, but it should be on its way soon.
Cryptographic weakness on Debian systems Posted May 13, 2008 18:49 UTC (Tue) by nix (subscriber, #2304) [Link] Even if those keys were compromised, that's OK. If the announcement is legitimate, then the fact that the keys are compromised is not problematic, because it was actually signed with the secret key. If it's *not* legitimate, and has been signed by a forger, then... why on earth would they tell us that the keys were weak, destroying their own strong point? (And if the message is forged and the sender is lying, being a nasty forger and all, well, er, a forger telling us that the keys are weak when they're actually *strong* seems really rather implausible.) I'd worry much more about an announcement coming out of the blue saying `hey, our keys are OK, keep using them!' because that *is* an announcement that an attacker who'd nicked the keys might want to give out (if he wanted to make people like me suspicious, anyway).
Cryptographic weakness on Debian systems Posted May 13, 2008 20:48 UTC (Tue) by man_ls (subscriber, #15091) [Link] Maybe the keys are good, but the attacker wants to make you think you have to get new keys -- which he will somehow forge and supply to you. In this case you should scrutinize the ways to "sanitize" your supposedly bad keys. An example: (s)he has discovered a weak point in GPG keys, so a method to generate "good" SSL keys from "safe" GPG keys is really a way to generate "compromised" SSL keys from "unsafe" GPG keys.It is a modern-day version of the old ploy where a fake detective comes and says: "here, your house is bugged, let me sanitize it for you", thus gaining your confidence and at the same time getting an excellent chance to install his own spying devices. You should watch him like a hawk. On second thought, even if you follow the guy he may be clever enough to deploy spying devices even if you are watching him all the time. Or in our case: the GPG vulnerability may be subtle enough that it is hard to catch the attacker. I hope some really clever people are watching this story unfold.
|
|||||||||||||