Whitelist Vendor IP range from Paloalto IPS

[mnadeem]

Hi All, I am looking for more effective way to whitelist a vendor on IPS without whitelisting at the FW as well.

I am looking for traffic from vendore ip range to be completely exempted from Vulnerability / antivirus / Anti-spyware without creating any firewall rule and security profile.

Is it possible ?

[OtakarKlier]

Hello,

Not without other potential complications, but a Policy is the best approach. Create a policy with the vendor IP's as the source and then do not perform any scanning on it. 

Its simple and effectve. It's how I allow these actions.

Regards,

[mhowsmon]

We have a scheduled security scan coming up and I need to perform this step as well. However I do not want to introduce openings to the scanner that are not open to the Internet. If I were to use this method, I believe I would have to mirror each policy with the source IP of scanner and have same destination rules / ports and not do scanning. Does that sound right?

[Kris.Waddle]

I've had this same question.  Building several shadow rules to exempt an ASV from IPS only is not a road I want to start down.  Allowing a separation between IPS and Host/Port would be 100% Helpful.   

[OtakarKlier]

Hello,

Sorry for the late response, but the answer is no, you would not need a shadow policy for every one you have. Just one before the others that has source of the vendor, destination, your IPs, and the rest allow any/any. This will only open the firewall to the vendors. We do this but internally, datacenter A scans data center B it doesnt report on every port and application because a full tcp handshake was not established. 

Example:

Internet -> PAN -> webserver over ssl/443 only.

The scanner will not pick up port 80 since its not open on the server. Just make sure they disable syn only packets for a full connection.

Hope that makes sense.

[Jamshid]

Like that we do not need to duplicate each rule!  But not sure about the recommended solution:

Allowing any/any from the source IP will fully open the network to that IP, when we only want to allow them Vulnerability/Compliance Scanning.  Isn't there a tighter/cleaner solution, or am I misunderstanding your suggestion?

Thanks!

[Jamshid]

Allowing any/any from the source IP will fully open the network to that IP, when we only want to allow them Vulnerability/Compliance Scanning.  Isn't there a tighter/cleaner solution, or am I misunderstanding your suggestion?

Thanks!

[OtakarKlier]

Hello,

If I'm understanding your question correctly, the end result is to allow an external vendor to scan your external perimeter without the PAN blocking it.

If that is the end result, this is the cleanest way I know how with one policy. There could be others out there that have done other things.

Regards,

[mhowsmon]

I want the scanner to be able to scan without being blocked but I want them to only see the ports that are exposed to the Internet, not "any/any". That is why I went with shadowing of each rule.

[Jamshid]

Thanks for your quick reply!  This allows full scan, but not validation of existing rules.  Duplication of each and every rule would be a nightmare, as we have 10 pairs of firewalls, and many hundreds of rules between them.  Ideal solution would be a single Panorama pre-rule to disable IPS for one single IP, from which the Vulnerability/Compliance Scanning would take place. Is that possible?

[OtakarKlier]

Hello,

Yes this does validate existing policies. If its say ssl, the vendor will try ssl over port 443 and if its allowed the vendor will show it as open.

Regards,