ADVERTISEMENT

[OT] log4j (Dungeon IT Geeks)

UCFBS

Todd's Tiki Bar
Gold Member
Oct 21, 2001
31,412
12,019
113
USA
First, for non-IT people, a mainstream article ...



Now, for us Dungeon IT Geeks (don't get me that 'professional'), This is not good ...




Someone else even claims a Black Hat talk from 2016 identified this issue, which wouldn't surprise me.




Maybe I'm missing something, but I don't know why log4j's devs seemed to think it's wise to pull and execute any plugin -- essentially any arbitrary code -- from any LDAP store (and seemingly DNS, among via one other protocol) ... and, most of all, enable it by default. That latter was just ... well ... chronically stupid. How this escaped all the eyes of open source for over 5 years -- although maybe it didn't if the Black Hat talk comment is true -- is beyond me.

I'm not an anti-Java guy, but as a colleague I greatly respect put oh-so-well ... "Java is the new COBOL." I've never liked 'stack machines' (I won't go further into my EE bigotry against CS), and we've reached that point. Rust the Internet. Even Linus sees value.

I'll quote the Tenable (e.g., Nessus) for issues related to the Proof-of-Concept (PoC) log4shell and other, known information as of Fri 12/10.


Assuming the prior PoC from March isn't the same (although looks similar enough), the attacks have been in the wild since 9 days before disclosure. But given how much of a massive oversight this is, and how much LDAP store lookups are used for almost everything -- including to Active Directory (it's also LDAPv3 based, although it's had similar RFC exploits, but nothing so generic like this across Java/non-Windows) -- this is really bad.

 
  • Like
Reactions: OldClay83
ADVERTISEMENT
ADVERTISEMENT

Go Big.
Get Premium.

Join Rivals to access this premium section.

  • Say your piece in exclusive fan communities.
  • Unlock Premium news from the largest network of experts.
  • Dominate with stats, athlete data, Rivals250 rankings, and more.
Log in or subscribe today Go Back