On Monday, June 17, 2013, 3:24:26 PM, Brent Bice wrote:

Hey guys. I was just looking over the service (sounds promising) but

before I start using an RBL (or similar service), especially one we have to pay for access to, I always run some tests first. In this case, I'd take a few months worth of DNS query logs and run that through a perl script to see which hostnames would be caught by the RPZ.

Obviously, I'd only do this if I had the data mirrored locally so 

I'm not beating up on anyone else's DNS servers. :-)

So... Is there a way I can get one-time access to the RPZ and 

rbldnsd formatted data just so I can run some sanity checks? If I can quantify the number of URLs the users here queried for that the RPZ would otherwise have blocked (we run our own in-house RPZ too), I could then do a cost/benefit analysis for the folks who pay the bills, assuming it was catching stuff our own RPZ isn't.

Possible?  I'd only run it on one, non-production DNS server here 

for testing and would toss the data when I was finished, unless we decided to use it in production and pay for access...

Brent

Hi Brent, Thanks very much for your interest in trying SURBL data. We're always happy to let folks give them a try. The simplest way to do this is to sign up for data feed access at:

http://www.surbl.org/df

and we'll ask a SURBL reseller to start a trial with you.

BTW, the RPZ and rbldnsd (and really BIND too) data are trivially similar transformations of the same data so any of them should be usable for testing. The rbldnsd version is by far the most commonly used for DNS service, and also non-DNS uses, so it perhaps could be thought of as the reference format. But we make the data available in the different formats for different applications.

Cheers,

Jeff C.