Re: Tracking the bad guys
Eric Brunner-Williams is slightly incorrect in his description of the blog-spammer's attack, because he's misinterpreting whois. He states that based on the spammer's entry in the whois entry, the spammer "claims domicile" in whatever location. Whois records don't make any claims about domicile, legal jurisdiction, True Name, National Identity Number, Retina Prints, likely sleeping location, likely location of hardware, ICBM coordinates, or preferred subpoena acceptance location, though ICANN would certainly like it if they did. They're strictly indicating some postal contact information, and for the billing address, they're indicating where to send a paper bill. (Keeping them current is certainly good practice, and I'd recommend that Eric check nic-naa.net's whois phone numbers, which appear to have suffered from some helpful spreadsheet doing arithmetic on them.) Meanwhile, while it's annoying to have to do self-defense, rather than getting the miscreant's ISP to do it, if Eric's wife's machine is self-administered as opposed to administered by some hosting company, adding the miscreant's IP address to the firewall or routing table can take care of the bandwidth problem, and while collateral damage is a bad thing for ISPs to do, it's not unreasonable for personal machines. Bill Stewart, bill.stewart@pobox.com
Eric Brunner-Williams is slightly incorrect
that happens.
Whois records
if you read my note, the only whois data of interest is the registrar and the ns providers (and their ns providers). other data of interest originates from rir public rwhois servers.
Meanwhile ... the miscreant's IP address ...
this instance was interesting in its unsophistication. from a related writing: The insertion network is is single address [151.42.235.185]. The subscriber network is is single property [paxil-medication]. More generally, multiple robo-hosts comprise the insertion network (attack side), trailing, but following the same technical trajectory as SMTP spam, and multiple URL payloads (benefit side), and commit only a few ad inserts in any discrete attack over a larger range of targets.
I'd recommend that Eric check nic-naa.net's whois phone numbers,
that was the one useful item you wrote. core-50 may have a problem, and it may be the case that the core-srs whois server may have a problem. thanks for the data point. incidently, in addition to post-detection persistent blocking, temporal approaches (interstitical gap management) for a single attack address are available, and a nanog reader has mentioned an implementation of a baysean approache in private mail. eric
participants (2)
-
Eric Brunner-Williams
-
Stewart, William C (Bill), RTSLS