Prospero All American 11662 Posts user info edit post |
not as secure as everyone thought... among other open-source projects affected. http://www.dailytech.com/article.aspx?newsid=11869 http://www.technologyreview.com/Infotech/20801/page1/
128-bit, 16-bit FTW! 5/23/2008 4:57:04 PM |
Prospero All American 11662 Posts user info edit post |
ok, to clarify:
Quote : | "Be careful. This whole article implies that the problem is more widespread than it really is.
Only distributions (like Ubuntu) that use the Debian repositories were affected. NONE of the commercial vendors and most of the other major distributions (RPM-based, source-based, etc.) are completely unaffected. This also only affects keys generated on Debian derivatives.
Further, Ubuntu is distributing with the updated OpenSSH packages a key blacklist and vulnerability assessment utility. Users who have bad keys are being notified at the time of update that their keys may be compromised." |
http://www.ubuntu.com/usn/usn-612-1
sorry for the partially incorrect title
[Edited on May 23, 2008 at 5:03 PM. Reason : .]5/23/2008 5:00:31 PM |
scud All American 10804 Posts user info edit post |
partially incorrect title?
more like extremely misleading title 5/23/2008 5:10:52 PM |
Prospero All American 11662 Posts user info edit post |
ok, replace linux with openssh
linux distros = ubuntu & debian
do millions of linux distro's out there (people running linux) only have 32,767 keys at their disposal? yes.
it essentially means that of the millions of people running debian / ubuntu, you could brute force their OpenSSH key in like seconds.
[Edited on May 23, 2008 at 5:16 PM. Reason : .] 5/23/2008 5:12:49 PM |
gs7 All American 2354 Posts user info edit post |
Ummm, even if I were running Ubuntu/Debian ... no, you can't brute force my OpenSSH key in any useful timeframe. It's multiple words and 36 characters long. Good luck.
[Edited on May 23, 2008 at 5:35 PM. Reason : .] 5/23/2008 5:34:51 PM |
Prospero All American 11662 Posts user info edit post |
um, that's not how it works.
you're passphrase is not the key, the key is a randomly generated key that's used to send your data, so after 2^15 attempts, the key WOULD be compromised since that's the maximum # of keys it has to generate. once someone has the key they can view your data being sent, so any "passphrase" you use for any other website.... no matter how long it is, would be compromised.
[Edited on May 23, 2008 at 6:08 PM. Reason : .] 5/23/2008 6:04:40 PM |
smoothcrim Universal Magnetic! 18966 Posts user info edit post |
this only matters if you're using 16bit hashing in addition to passphrase-less auth (pub/priv key pair) which most people do NOT set up. read the article instead of going off what you posted.
this still isn't a huge deal without a lot of other conditions met
[Edited on May 23, 2008 at 9:18 PM. Reason : I haven't used debian/ubuntu, does it setup pub/priv ssh by default or something?]
5/23/2008 9:07:59 PM |
philihp All American 8349 Posts user info edit post |
http://metasploit.com/users/hdm/tools/debian-openssl/
^ Good explanation of the bug and its impact.
5/24/2008 12:12:48 AM |
gs7 All American 2354 Posts user info edit post |
^Thanks, that explanation makes a lot more sense. 5/24/2008 12:55:09 AM |