Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

What program do you use for detecting Rootkits? How do you know what to trust?

share

locked by HopelessN00b Dec 5 at 6:35

This question exists because it has historical significance, but it is not considered a good, on-topic question for this site, so please do not use it as evidence that you can ask similar questions here. This question and its answers are frozen and cannot be changed. More info: help center.

6 Answers 6

up vote 5 down vote accepted

On Unix-based systems, Tripwire is a good general "what changed on this machine today?" solution. There are other, more specific rootkit detectors out there, but I've always thought that it was a matter of playing catchup with the bad guys; you'll never be sure that your rootkit detector is up-to-date enough to catch all of 'em.

share

I have been using OSSEC and have been really impressed by the results

share
    
+1 AFAIK OSSEC is more complete that other solutions like Tripwire because not only does file integrity checking, it also monitors logs, analyze security configurations and has a specific module for rootkit detection –  chmeee May 2 '10 at 20:22

None of the available free rootkit checkers for Linux are very good. They can't even detect all the publicly available (most for many years) rootkits that you can find source code for on sites like packetstorm. Not only do they not do a good job with the known threats, but there is good reason to expect that they will perform worse against any intelligent attacker, who has the opportunity test against the detectors in advance of using his rootkit.

Furthermore, there is no reason to assume that a rootkit will have an on-disk footprint that can be detected with a tool like Tripwire. Kernel rootkits can deceive any software running on the infected system; even if you're running a trusted binary from read-only media it might be getting falsified information from the operating system. Furthermore, memory-resident-only malware is not just a theoretical worry any more; its use has been documented by security firms such as Mandiant. Against such a rootkit, even offline analysis would be ineffective, although on the plus side the malware might not survive the reboot.

There are some solutions, but not free or cheap. A large corporation or government entity can spring for a memory forensics tool like Second Look from Pikewerks, or one could hire a security consultant who might employ such a tool as part of their arsenal. Others, go ahead and use the available anti-rootkit programs. They are not 100% useless. Just do what you can to avoid being rooted by someone with a customized, non-public rootkit. How large is the risk? Unfortunately, there are no statistics I know of to go by. Clearly Linux does not suffer from the enormous volume of "mass malware" that Windows does, but how common are stealthy, targeted attacks?

(Disclosure: I am an author of the Second Look product mentioned above.)

share

I think the standards are chkrootkit and rkhunter.

I would use both, and run them daily. I know chkrootkit has an option to only notify you if anything changes (avoiding daily false-alerts).

Running both helps to 1) not need to "trust" either, and 2) protects against attacks that try to specifically hide from one or the other.

share

Rootkit checker should not be installed on your target machine. I do not know either chrootkit or rkjunter but if they required to be installed on the end machine they do not protect you much. A rootkitcheck software that runs on the target machine runs the risk that the rootkit (or the person who installed it) would compromise it and thus it would not provide you with the protection you seek.

Personnally I go with Tripwire. What Tripwire does is take a hash (fingerprint) of all the file on your system and it lets you know when a file change. It allows for a remote host to be your "trusted" machine and have it scan your target machines for any file change. If a change is detected then you know something went wrong. Of course, you need to have some change control so regular updates on your system does not get flagged as an intrusion.

To be on the safe side, at regular intervals, you want to go and make changes on your target machine to make sure that tripwire will report it. You also want to cut the connction between the target and the trusted machine running tripwire to make sure that it will be detected. This is as important as making sure you can restore your backups.

share
2  
AFAIK sophisticated rootkits can modify system calls and thus spoof the hash of the file system. Taking the hard drive out and putting it in a 'clean' machine as a secondary drive and taking a hash of it from there would be a more comprehensive check I'm guessing. –  Jonathan Parker May 8 '09 at 6:28
1  
This is why running Tripwire from a "trusted" machine. This is so the tripwire executable and the hash database are "safe". Of course they are as safe as the trusted computer. –  Pierre-Luc Simard May 8 '09 at 16:31

I generally run external virus/rootkit scans from a Linux live disk. It's the only thing you can really trust. It does still involve keeping up to date with detection programs though.

share

Not the answer you're looking for? Browse other questions tagged or ask your own question.