4 ms·
Tracking down these systems is easy, so these issues can normally be solved pretty easily. Thats because typically any amplification vector doesn't allow the s
by londons_explore 5y ago
Tracking down these systems is easy, so these issues can normally be solved pretty easily.
Thats because typically any amplification vector doesn't allow the source IP of the amplifier to be spoofed. So as soon as a DDoS attack begins, a sample of the packets can be taken to get a list of the amplifiers used. Those can then be tracked down and patched to no longer act as amplifiers.
- egberts1 5y agounless the amplifier mechanism is widespread.
- DFHippie 5y agoConcerning this particular vector: > Approximately 2,600 of these systems have been incorrectly provisioned so that an unauthenticated system test facility has been inadvertently exposed to the public internet
- sp332 5y agoIt could be easily solved by the operator, but that doesn't mean it's easy for the victims to get the operators to fix their stuff. These amplifiers are already run by people who ignored the software manufacturer's directions. What are the odds they will actually install the new version that's harder to abuse?
- amalcon 5y agoUsually[0] contacting the operator's ISP and informing them of the situation will get said ISP to contact said operator. All that outbound traffic does represent a cost to the ISP, after all. A call from your ISP usually gets a bit more respect than a call from some random person. [0]- In the US; I don't know about anywhere else
- bombcar 5y agoIn the past what usually happens is the ISP disconnects you until you prove you've fixed whatever it was (sometimes they're nice and block just part of the connection, or give you a warning). Surprisingly enough, the ISP often has no real way of contacting anyone; the easiest is to cut the connection and wait for a complaint.
- autoexec 5y agoYep. Sad but true. Nobody bothers to keep their contact info up to date with their ISP it seems. Non-critical stuff sometimes can be mailed to a customer's service address, but often disconnecting someone is all an ISP can do to make them aware they have a problem.
- toast0 5y agoIt really depends on the ISP. After spending some time trying to get phishing sources taken down and not getting anywhere, I wouldn't be hopeful about DDoS (reflection) sources being taken down either. When I was running servers that were getting DDoSed frequently (but thankfully for short intervals and not with tons of bandwidth), trying to get chargen servers or wordpress servers fixed didn't even seem like an option. Just make sure my servers wouldn't fall over, or at least would fall over gracefully.
- LinuxBender 5y agoLet's get started! [1][2] You wanna take the odd numbered IP's and I take the even? Just kidding I am way too lazy for this [1] - https://www.shodan.io/search?query=mitel https://www.shodan.io/search?query=mitel [2] - https://www.shodan.io/search?query=mivoice https://www.shodan.io/search?query=mivoice
- treesknees 5y agoYour comment underestimates the task of remediation. Sure, we can very easily get a list of DDoS source IP addresses. Any decent network operator can get a list of flows matching some DDoS criteria and generate a report of IP addresses. In the case of this TP240 attack, you're talking about ~2600 independent businesses across the world. Assuming you are able to determine the actual source of the traffic and work with a vendor to patch it, you're still tasked with somehow getting 2600 businesses to patch their systems or modify firewall rules. In the case of the memcached amplification attack, Cloudflare saw upwards of 5800 source IPs in the attacks, and Shodan reported nearly 88000 IPs responding on port 11211 [1]. Tracking down the owners of 88k installations across public clouds, businesses, probably some residential networks, is a monumental task. There's nothing easy about it. [1] https://blog.cloudflare.com/memcrashed-major-amplification-attacks-from-port-11211/ https://blog.cloudflare.com/memcrashed-major-amplification-a...
- wtarreau 5y ago> you're talking about ~2600 independent businesses across the world. Assuming you are able to determine the actual source of the traffic and work with a vendor to patch it, you're still tasked with somehow getting 2600 businesses to patch their systems or modify firewall rules. You can be sure that by only null-routing their entire C-class, adjacent customers will loudly complain to the operator who will quickly identify the source and disconnect it. The best way to deploy fixes on the net has always been to first disconnect them. This way you don't have to convince anyone, it's done the other way around. Typically the CEO will instantly throw all the phones to the trash to get the net opened again.
- treesknees 5y agoIn general that's not really an option. Unless you have coordination with the network operators on which those amplifiers are sitting, your null-routing of the amplifier in your own network isn't going to stop it from attacking other targets. If the amplifier is something like a DNS server, then your collateral damage isn't just "adjacent customers", it's potentially thousands of other users and resolvers on your own network. If those amplifiers are on a cloud service provider like AWS, you're going to potentially inflict even more pain onto your own paying customers who will no longer be able to communicate with AWS. You will essentially perform the DoS they were aiming for.
- Nextgrid 5y agoWe need proper liability laws for malicious traffic. You are liable unless you can pass off that liability to someone else. So the ISP would be liable by default, and would have an incentive to filter their customers, or require them to abide by certain rules, pass some audits, provide proof of insurance or post a large deposit. You could have insurers who in exchange of automated security scans will insure you, solving the problem for end-users at a reasonable cost. This will actually encourage internet users (both consumers and businesses) to take security more seriously.
- hombre_fatal 5y agoLitigation seems too heavy handed for these kinds of attacks. A major issue here is how your smart toaster or MiVoice box can be spamming the internet and there's no real way to realize it for most people. Since you pitched a controversial solution, let me make one that's probably even more controversial: maybe bandwidth is too cheap. Maybe the problem would fix itself without legal hell if your C&C'd smart toaster / VoIP box had an impact on your ISP bill instead of being folded into your unlimited bandwidth billing.