4 ms·
Where do you have the information from that "all malicious traffic attempts to use privacy respecting technology."? I just checked an access.log (from a searx i
by bauruine 4y ago
Where do you have the information from that "all malicious traffic attempts to use privacy respecting technology."? I just checked an access.log (from a searx instance) and an auth.log for malicious traffic and it doesn't look at all like that.
ssh bruteforce top 5 offenders:
147 (Tor: 0) TENCENT-NET-AP-CN
133 (Tor: 0) DIGITALOCEAN-ASN
31 (Tor: 0) CHINANET-BACKBONE
17 (Tor: 0) BAIDU
13 (Tor: 0) CHINANET-SH-AP
Overall there where 737 unique IPs from 241 ASNs and 1 was a Tor node.
access log top 5 offenders:
76 (Tor: 0) DIGITALOCEAN-ASN
58 (Tor: 0) CONTABO
54 (Tor: 0) AMAZON-AES
50 (Tor: 0) KAZTELECOM-AS
34 (Tor: 0) CORBINA-AS
Overall there where 1672 unique IPs from 618 ASNs and 4 where Tor nodes.
- brightball 4y agoNot necessarily Tor, but a VPN could be coming from any of those sources. I spent a recent chunk of my career combating fraud on a niche-eBay style site and the people trying to defraud other users, pay with stolen credit cards, login with phished credentials, etc were consistently trying to hide their origin. Until we started using fingerprinting techniques to track them across multiple accounts and IPs, we had no way to spot this. It was a shock to me when I realized there were legitimate uses for fingerprinting technology because I'd always associated it with ad networks and trackers. They're fairly necessary for combating fraud though. When we stopped letting any untrusted users run a credit card if their connection couldn't be trusted, our charge backs virtually stopped. That experience makes me completely understand sites scrutinizing anonymized traffic.
- maccolgan 4y agoMost Tor exits block SSH and SMTP. Look elsewhere.