New P2P botnet infects SSH servers all over the world
arstechnica.com
arstechnica.com
The idea of using SSH servers (from the perspective of whichever group or individual(s) wrote the given botnet that uses them) makes a lot of sense (in an evil way), because there's a lot of SSH servers out there on the Internet.
To dig a little bit deeper (if I were an investigator), I'd investigate all of the historical security issues that SSH has had, in its history...
Were some of those intentionally engineered?
?
If so, then whichever groups or individual(s) were involved -- might be -- or might lead a security researcher a step closer to the actual authors of the botnet...
Also... one needs to ask "who would need the botnet, and why?"
While I'm on the topic, I think it would be a great idea to research the historical security issues that not just SSH, but every piece of server software (Apache, etc.), in order of the software's popularity.
In other words, the more popular a piece of server software is -- the more likely it is a target for intentional botnet "security modifications" (that is, intentional weakening of the server software's security to support future botnet activity...)
Also (and especially) if those server software packages use common shared libraries of code in their codebases, then those common shared libraries are prime targets...
So, here's what I would do for sleuthing this:
1) List/enumerate all known pieces of software that act as servers on the Internet.
2) Order the list created in #1 by popularity.
3) Look for all libraries/code bases -- shared across those.
4) Look at the historical security issues of those codebases.
and, for skilled investigators:
5) Audit these codebases for changes/commits that look similar to other historical changes in other codebases where security was compromised.
Anytime someone sees weird or not-understandable code in a shared codebase -- it should be flagged for peer review. The more eyes, the better.