On Thu, 4 Mar 2004, Michael Torrie wrote:
Mar 3 21:58:15 pentane sshd(pam_unix)[19464]: authentication failure; logname= uid=0 euid=0 tty=NODEVssh ruser= rhost=213.157.189.xxx user=root
Mar 3 21:58:19 pentane sshd(pam_unix)[19464]: session opened for user root by (uid=0)
j00 is teh ownz0r3d.
While an unskilled attacker can be expunged from a system with some careful work, there's no way to distinguish an unskilled attacker from a very skilled attacker. And there's virtually no way to get rid of a skilled attacker.
So if you really want to be safe, do this:
1. Shut off the machine, instantly, right now.
2. Mount the disk in another, secure machine ONLY FOR THE PURPOSES OF GETTING YOUR DATA FILES OFF. Mount it somewhere like /mnt/unsafe with -noexec, make sure not to run any programs on that disk, and copy off any data you want to keep, like text files and email. Don't copy any programs. You should even avoid things like configuration files, since they can be hacked to open up vulnerabilities.
3. Reformat and reinstall, then copy back any files you kept.
-J
Recovery of the data will not be terribly difficult, nor risky. There isn't a whole lot of data there (almost all visually and programmatically inspectable for correctness) for starters, and the configs are straight forward and easily re-done from scratch. While it would save some work if we did recover the data/configs, it wouldn't be a major loss if we didn't (MRTG logs, and a development trouble-ticket system).
I will be wiping and re-building the machine from scratch when the time comes. The main goal (at this moment) is not to get this install working again.
Finding the method of attack is more important to me, because I have a window of opportunity to do so before I need to get the system operational again. This was a very up to date machine with regular human monitoring of the log files, so I feel an obligation to do some forensics and report the method of break in, if that is in fact what happened (and yes, those two lines lead me to believe that is the case).
That is why I need to find a useful interpretation of that kernel error, to find out how they got in. At first glance, it looks like sshd and the kernel are involved, and if so, then that represents a serious risk to the community, which is pretty important to me to address since I am in a (time-limited) position to do so.
Thanks,
Adam Augustine____________________
BYU Unix Users Group http://uug.byu.edu/ ___________________________________________________________________
List Info: http://uug.byu.edu/cgi-bin/mailman/listinfo/uug-list
