[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: rkhunter / chkrootkit



Hi Rick,

> Why don't you make a copy of one or more of those binaries, then
> re-retrieve and install the Woody package of the same release, and
> compare md5sums of the resulting binaries?  (Note that you should make
> very sure it's the same release, or you'll get a different md5sum for
> entirely innocent reasons.)

indeed, I could do it. After an established contact to one of the 
maintainer the previous advice to --update the md5sum from the 
rkhunter server solved the problem and it was not an irregularity
within the debian server. So they've updated now which was required.

> >   Checking /dev for suspicious files...                      [ Warning!
> >   (unusual files found) ]
> Well?  What files?  The fact that rkhunter has an opinion is not, by
> itself, particularly interesting.  You either have to know rkhunter
> very, very well, such that you have a high degree of faith in its
> opinions, or need to investigate for yourself what it claims is
> suspicious.  Preferably both.

Don't know what files as there was no output and by the way it was
the first time I used rkhunter.

> > - ProFTPd 1.2.5rc1                                         [Vulnerable ]
> > - OpenSSH 3.4p1                                            [Vulnerable ]
> > - GnuPG 1.0.6                                              [Vulnerable ]

> Well?  _Are_ those actually vulnerable, or is rkhunter making bad
> assumptions?  If you are running a conventional woody system, then
> you're receiving backported security fixes -- which does not change the
> package version number.  Ergo, if rkhunter is stating the foregoing
> strictly on the basis of version numbers, then it is making a common
> elementary error.

Hm, to be honest I wasn't able to read the source code but I don't think
that my ProFTP is not vulnerable and I've to agree rkhunter is not
able to detect the correct version so you're right.

> > Incorrect MD5 checksums: 6
> Which ones?  And on what basis is it saying they're incorrect?  You
> don't say.

The binaries mentioned above.

-- 
Best Regards,

Mark



Reply to: