Is it in the terms of conditions you have with the client to allow this?
If not well....
We sometimes open up the firewall just for an agreed period, say we set
a temp rule that expires in 5 days automatically....only from that
specific IP to a specific IP.
Of course our protection is the firewall, so in effect it passes...logic
should tell the security firm this.
We also find that lots of these customised scanners look for version of
software on a port and if its is < the latest it gets a flag, of course
Linux distro's backwards patch and version control is different so in
effect a redhat mailmain 2.5.1.64 (say) could be patched to a "real
mailman 2.5.3 and be safe, but the scanner says not.
My answer in these circumstances is, if the client has paid for this
work with the 3rd party security company we work with them, if not they
get told of the cost and it is up to them to pay us or not.
Also in your terms of conditions there should be something about routine
patching, say every 3 months, important patching done in say 7 days,
critical patching done in 3 days, critical with a known exploit 8
hours....and a disclaimer that you don't take responsibility for a
critical patch breaking the system. We have a rolling program every 3
months, so the client(s) know when the server will be patched and
whether to object or not, or test or not.
Regards
Steven aka thing
-----Original Message-----
From: Ademar Gonzalez [mailto:ademar.gonzalez@gmail.com]
Sent: Wednesday, 14 December 2005 5:37 a.m.
To: webappsec@securityfocus.com
Subject: PCI DSS Compliance
A shared hosting client needs to get his site PCI DSS certified.
He forwarded us the following request from the company doing the
assessment.
"Your site could not be certified. Your site appears to be running
scan detection software, that has prevented a reliable port scan. This
test is inconclusive. Please add our scanner ip: ##.##.##.## to your
scan detection software exclusion list to allow our scanner to make a
complete assessment of your system."
Is this request plain stupid or what ? Comments ?
I have deal with this kind of requests in the past and most of the
time the people running
this automated scans knows nothing at all about security nor anything
else and it becomes a pain dealing with the client on one end that
wants his website certified and the other guy on the security company
that wants you to open your firewall so hi can run his nmap or
whatever it is they run. It looks like the client runs the risk of not
being certified 'cause his website is over-protected. How would you
proceed in this situation ?
ciao ciao
ademar
