But that’s the problem: you can’t detect even one thing because the attack traffic looks like normal traffic 💯 . It is, in fact, normal traffic in the sense that it’s exercising the service, such as HTTP, on the victims network. What makes it a distributed DoS attack is that many attackers present a huge load to this victim server while not showing any particularly intensive stream on the consumer attacking network. The consumer attacking network is then directed to attack many distant victims, but that just looks like more normal traffic to a lot of different web hosts. In extreme cases, the attacker may max out the upstream capacity of the consumer proxy network. But once again, that’s not terribly surprising because some consumers just hit their limit based on gaming or whatever it is, they’re doing. It’s not an AUP violation to use all the band with you’re paying for. -mel via cell On Jan 17, 2026, at 11:22 PM, Suresh Ramasubramanian <ops.lists@gmail.com> wrote: Yeah that’s why my focus was on isps detecting this outbound - if only through feeds like shadow server - and cutting off infected customers. --srs ________________________________ From: Mel Beckman <mel@beckman.org> Sent: Sunday, January 18, 2026 10:44:34 AM To: Suresh Ramasubramanian <ops.lists@gmail.com> Cc: nanog@lists.nanog.org <nanog@lists.nanog.org>; Tim Burke <tim@mid.net>; Corey Smith <cosmith80001@gmail.com>; Roland Dobbins <Roland.Dobbins@netscout.com> Subject: Re: ISP Operators AISURU/Kimwolf botnet Suresh, Aye, there’s the rub. It’s very difficult to classify DDoS traffic at the ASN level unless you can see it across ASNs on the Internet backbone, then and correlate it using sophisticated pattern recognition. This what services such as Black Lotus, CloudFlare, and Fastlick do. Especially for small ISPs, this is impossible to do in-house, at least today anyway. Hackers quickly learn how all these DDoS recognition services work, so it’s a constant battle or whack-a-mole trying to stay ahead of them and their information hiding techniques. There are CPE devices — NG enterprise firewalls — that can detect and block some large output streams. But customers aren’t willing to pay a $2000 setup fee and turn over the shelf space and endure the noise footprint for these products. But you’re right about one thing: ISPs, where they can identify abusive outbound traffic, can turn off those customers and leave it to them clean up their home networks. -mel On Jan 17, 2026, at 5:50 PM, Suresh Ramasubramanian <ops.lists@gmail.com> wrote: Write it and submit it to the NYT and that crowd would just call it mass media lies. Whatever happened to all the walled gardens Comcast and others were working on? If you see this traffic just cut the connection based on the customer bumming free tv with a knockoff Chinese device, if not for the malware. --srs ________________________________ From: Mel Beckman via NANOG <nanog@lists.nanog.org> Sent: Sunday, January 18, 2026 9:41:54 AM To: Tim Burke <tim@mid.net> Cc: nanog@lists.nanog.org <nanog@lists.nanog.org>; Corey Smith <cosmith80001@gmail.com>; Roland Dobbins <Roland.Dobbins@netscout.com>; Mel Beckman <mel@beckman.org> Subject: Re: ISP Operators AISURU/Kimwolf botnet You should write the article and submit it to the New York Times technology group. I believe David Pogue still works there, and he is a tech guy, so maybe he would be a good vehicle to get it published. I used to work with DAVID at Macworld magazine. But it’s not the job for an ISP, or even something an ISP could get the major media to publish. -mel via cell
On Jan 17, 2026, at 4:50 PM, Tim Burke <tim@mid.net> wrote:
The problem I see is that an article like this is intended for an IT/security professional audience.
These TV piracy boxes are often used by uneducated folks that would not read such an article. They just want their sports and $cableNewsChannel, and if you tell them it’s illegal or full of malware, they will just tell you you’re wrong, keep using it, and let it cause their 1Gbps circuit to get saturated by botnet traffic, all in the name of “free television”.
I have joined a few social media groups about these devices out of sheer curiosity, and have seen a number of threads from folks that ask why an ISPs security offering (typically Comcast’s “XFi Security” or AT&T’s “Active Armor”) would be complaining about traffic coming from the device… the common trend is to tell people to disable the security services, as “Infinity [SIC] is just trying to force you to buy their cable”.
Hooray for Stockholm syndrome.
On Jan 16, 2026, at 20:10, Mel Beckman via NANOG <nanog@lists.nanog.org> wrote:
Roland,
The Krebs article you cite is even better than the one I linked, because it shows pictures of the many consumer devices that can be infiltrated. People are likely to immediately recognize any they own, which will drive home the point that this is their problem.
-mel
On Jan 16, 2026, at 5:43 PM, Dobbins, Roland via NANOG <nanog@lists.nanog.org> wrote:
On Jan 16, 2026, at 22:16, Benjamin Hatton via NANOG <nanog@lists.nanog.org> wrote:
As a smaller ISP, I think the biggest thing that would help us would be a 'mainstream' media outlet covering some of it so we have something to show customers who call in about their internet being bad, us telling them it is their android streaming box that is taking up their entire connection moving TBs of data a day, and them responding with "but I bought it from Walmart/Amazon" or "you are just trying to get me to sign up for your cable" and refusing to do anything about it because 'free TV'.
<https://krebsonsecurity.com/2026/01/the-kimwolf-botnet-is-stalking-your-local-network/> The Kimwolf Botnet is Stalking Your Local Network<https://krebsonsecurity.com/2026/01/the-kimwolf-botnet-is-stalking-your-local-network/> krebsonsecurity.com<https://krebsonsecurity.com/2026/01/the-kimwolf-botnet-is-stalking-your-local-network/> [favicon.ico]<https://krebsonsecurity.com/2026/01/the-kimwolf-botnet-is-stalking-your-local-network/>
_______________________________________________ NANOG mailing list https://lists.nanog.org/archives/list/nanog@lists.nanog.org/message/GC4T5N6X...
NANOG mailing list https://lists.nanog.org/archives/list/nanog@lists.nanog.org/message/3LYEDZZ6...
NANOG mailing list https://lists.nanog.org/archives/list/nanog@lists.nanog.org/message/HZIDJSNE...