Zero Errors? Spamhaus Flubs Causing Domain Deletions 170
Since 2005 I've been running a proxy mailing list where users sign up to receive new proxy sites by email. (Proxy sites are sites for getting around Internet blocking software; most proxy sites that you can find through Google are already blocked by major blocking programs, which is why you would sign up to receive new ones by email, to use them until they get blocked as well.) In all that time, we've followed what are considered best practices for email newsletters: every new subscriber is sent a confirmation message by email, and they have to reply to that message, confirming that they really want to subscribe to the emails, before being added to the list. This practice, known as "verified-opt-in," is considered the gold standard for responsible emailing, since it ensures that everyone on your list actually wants to get your emails. (It also ensures that if you accuse an email publisher of spamming because you received their unwanted emails, they can't say, "Oh, one of your friends must have added you" — since if they're using verified-opt-in like they're supposed to, your friends can't add you.) I'm front-loading a lot of information here, although if you saw the words "Spamhaus errors" in the title, you may recognize the technique of literary foreshadowing being employed.
Despite conforming to verified-opt-in standards, the proxy emails have at times been blocked by spam filters used by Hotmail, Gmail, Yahoo Mail, AOL Mail, and various other systems. However, last month was the first time that an incorrect blacklisting caused the domains themselves to be disabled, so that the sites disappeared from the Internet entirely.
On September 17th I registered 10 new .info domains through NameCheap, set up new proxy sites at each of those domains, and mailed each site to 1/10th of our proxy mailing list. (Sending new sites only to a subset of the list makes it harder for blocking software companies to join the list and find all new sites as soon as they're released.) All seemed to be going well until October 2, when subscribers started telling me that they were getting "host not found" errors when trying to reach the sites. I tried the sites myself, found that they were indeed inaccessible, and spent about an hour testing for various problems with DNS servers and domain record settings, before logging in to NameCheap and seeing a message next to each of the new domains saying "domain locked due to illegal activity; please email legal@enom.com." (NameCheap being a reseller for the domain registrar eNom.)
So I sent eNom an email and followed up with a phone call to see if they could speed things up, since complaints kept pouring in from users that the sites were unreachable. eNom said that the domains had actually been suspended by Afilias, the company that handles all .info domain registrations no matter who you buy the domain from, and eNom was in the process of talking with Afilias. So I called Afilias myself to ask about getting the domains unlocked, but they refused to talk to me and said that they could only respond to inquiries from eNom. This, of course, is ridiculous — if someone notifies you that you or your company has made a error, you can investigate the issue no matter who brings it to your attention — and especially in cases where you're literally accusing someone of unspecified "illegal activity," you should bend over backwards to respond to any indication that you might have made a mistake. But they refused to do anything, so I waited for a response back from eNom.
A day and a half ticked by, with emails continuing to come in from our users wondering why the domains had disappeared, until finally eNom forwarded me a response from Afilias saying that two of my ten domains ("drybook.info" and "rootface.info") had been blacklisted by the UK-based organization Spamhaus on their Domain Block List. Spamhaus operates several different alleged "spam" blacklists, and claims that the DBL is a list of domains found in spam messages. The DBL FAQ says that it is "built predominantly using automated spamtraps and email flow monitoring" and "has many checks to prevent legitimate domains being listed," even going so far as to call it a "zero false-positive" list.
Even though only two of the ten domains that I had registered that day had been blacklisted by Spamhaus, Afilias had responded by disabling the entire group of ten domains that I had bought at the same time.
Now here's where I caught a bit of a break: It turns out I was able to get the domains instantly removed from the DBL by entering them in a form on the Spamhaus site and clicking a button, which took me to a page saying:
DBL removal successful
The domain was successfully removed from the DBL. Please allow 30 minutes for servers around the world to update their data. Please note that the domain will be re-listed if malicious activity is detected in the future.
Although, even this easy part of the process didn't inspire much confidence. Not that I wanted Spamhaus to make it harder for me to de-list by domain names, of course, but if you really think your blacklist is 100% accurate, why would you let anyone get any domain removed at any time just by submitting it in a form? In fact, this would seem to give an advantage to spammers over regular website owners — because a spammer, who knows about blacklists and would find it worthwhile to game the system in his favor, would be more likely to know about the Spamhaus DBL and the form for getting their domains de-listed. Whereas for a regular non-spamming website owner, it would take far more time to find out that their domains had been de-activated, that the de-activation had occurred because of an incorrect Spamhaus listing, etc.
Once the listing had been removed, I emailed eNom, who emailed Afilias, who eventually re-activated the domains after a few more hours. But the traffic never returned to the levels that it had been at before the domains were deleted, as most of our users had apparently concluded that the sites had been blocked or taken offline.
Spamhaus did not respond to requests for comment on this story. In fact, Spamhaus does not give you a way to contact them if you have been wrongly blacklisted — their "contacts" page redirects you to the "Blocklist Removal Center" if your domain is blocked, but that only leads you to the automated removal tools, not a way to contact the organization. I did email their "Press Office" email address, on the grounds that I was writing an article for Slashdot in addition to being a wrongly blacklisted domain owner, but didn't get an answer.
So I have no idea what will happen with the next group of domains that I send out to our proxy list. If Spamhaus signed up one of their "spamtrap" email addresses to our mailing list, then presumably any domain mentioned in a message sent to that email, will get automatically blacklisted (even though of course since they signed up the email address to our mailing list, that means it's not spam). If that happens, the entire next batch of domains might get disabled by Afilias as well.
Meanwhile, Spamhaus continues to claim that the DBL is a "zero false-positive" list. I don't know how many other false positives are on the list or how many domains have been abruptly disabled as a result, but if it's this easy to get incorrectly blacklisted, my money is not on "zero."
registries (Score:5, Informative)
If you have the time, find better contacts at Afilias and get them to clarify their policy. If you have the money, call a lawyer. If you are really bored and love
Re:registries (Score:5, Informative)
Yes, the answer to the poster's problem is to not use .info domains with this highly restrictive policy: http://info.info/information/anti-abuse-policy [info.info]
What is interesting about all of this is Afilias (the registry operator for .info) appears to be using the Spamhaus DBL in an automated fashion to add "serverHold" status to listed domains. ("serverHold" effectively removes the domain from the TLD root servers and can only be modified by the Registry. "clientHold" does the same thing, but can be modified by the Registrar, in this case eNom.)
This is the official ICANN agreement and related documents that allows .info to function: https://www.icann.org/en/about/agreements/registries/info [icann.org]
This is the Registry-Registrar Agreement (RRA) containing section 3.6.5 referred do by the .info anti-abuse-policy: https://www.icann.org/en/about/agreements/registries/info/appendix-08-08dec06-en.htm [icann.org]
In all of those documents, I see no mention of the registry operator (Afilias) being able to invoke their rights of RRA section 3.6.5 in an automated (API-used) fashion. You could email Afilias about it, but doubt they would respond. If we want to get to the bottom of how they are auto-serverHold-listing domains, it seems a lawsuit is the only way. Perhaps someone really did email abuse@afilias.info, and a human checked the SBL and looked at the batch of domains created near the same time from the same registrar.
Thanks, Bennett Haselton, for posting this article and telling us about these shady practices from Afilias.
If you wish to continue using .info, and eNom (namecheap), then it appears you should create separate accounts, and register 1-2 domains in each account, so at least they are not blocked as a group. Additionally, using multiple sets of nameservers will make the domains look "different" from each other.
Spamhaus DBL IS network abuse (Score:5, Insightful)
Spamhaus DBL is poorly run and full of spite listings and other garbage. Zero false positives? They mean zero legit entries. Spamhaus has become what it set out to oppose, and it's time they were exposed for what they are today. A disgrace to the anti-spam, anti-abuse community.
Re: (Score:2)
I assume you are talking about pink contracts.
Which is just another case where it pays to be corrupt.
no sympathy (Score:3, Informative)
You should consider this a wake-up call. It's time to switch from mass-email to a web page with RSS.
If people really want your newsletter, they'll come to you.
Re:no sympathy (Score:5, Informative)
Until the services their customers are trying to get around block his web page. Email works a bit better for this as it's not easily blocked (unless the people doing the blocking are going to block hotmail and gmail).
Re:no sympathy (Score:4, Funny)
Until the services their customers are trying to get around block his web page. Email works a bit better for this as it's not easily blocked (unless the people doing the blocking are going to block hotmail and gmail).
Well if those people would use a proxy they could get around that block.
Oh, wait...
Re: (Score:2)
For the most part there will be a lot less trying to block access to legitimate page. Vs. Blocking bulk emails.
You have your customers check an RSS Feed. They subscribe. And there is a little traffic all day.
They get emails. the server gets 100 emails. the email is then copied hundreds of times. So you are adding 100x the storage for each mass email. Plus you cannout opt out easily. Unlike an RSS feed you just turn it off
Re:no sympathy (Score:5, Informative)
That's great, but his list is a list of proxy servers. The purpose of those proxy servers is 'proxy avoidance'. My content filtering automatically filters pages in the category of 'proxy avoidance'.
Therefore, if someone wanted to use his proxy servers (which he's constantly adding new domains to to get around my attempts to keep my employees from avoiding my filters) he needs a way to get them those proxy servers and they need a way to find him. I'm not allowed to block email services, but I am allowed to block sites related to getting around my filters.
This is why email works better. They can sign up at home or on some page before I find and block it, confirm via email, then get updates even if I'm blocking the place where they signed up in the first place.
There is a problem with emails being blocked as well, but that is spam filtering not my active attempt to keep them from getting around my filters. Overall this is the fundamental problem with getting around content blocking/filtering. You have to be able to find the site that tells you how to get around the filtering before the people doing the filtering filter that site.
Re: (Score:2)
What is the purpose of those filters?
Unless it is to block full Internet access at places with public access or similar it is a waste of time. If someone has adequate time in front of the machine with filters, they can be bypassed - often fairly easily. The only thing that works is to block traffic in a separate firewall, proxy or similar, i.e. to move the filter off the machine in question. If you block all traffic except to that proxy/firewall, there's nothing the user can do to bypass. And you don't have
Re: (Score:2)
We do run everything to a single proxy/firewall. You still are trying to keep up with the blocking. Sure I can tell my firewall to block all proxy avoidance sites and it will, until there is a new one and then I need a update that includes that site (or I have to manually add it). Sure if I only allowed a set lists of URLs then I wouldn't have an arms race, but I'm trying to block content that management feels is not appropriate, not block the whole internet and just allow the select sites. There are alway
Re: (Score:2)
Hold on a sec. Let me summarize the exchange I just heard.
A: Bennet's mass e-mailing is getting blocked.
B: He should just put it on the web.
A: No, the web page will be blocked, while mail isn't easily blocked.
Are you serious? [knowyourmeme.com]
Re:no sympathy (Score:5, Insightful)
You should consider this a wake-up call. It's time to switch from mass-email to a web page with RSS.
If people really want your newsletter, they'll come to you.
...it's a proxy list.
how long do you think those sites would stay off chinas webfilters ?
a proxy list you can't get to is rather useless.
Re: (Score:2)
"a proxy list you can't get to is rather useless."
the very best kind
well, that's why he was sending it to 10% of subscribers.. so you can't just sub to the list and ban all proxies on the list. of course you could get around that by flooding the list I suppose - and the proxies themselves ending up on spam filter lists I would think of as granted(if they do pure tcp anyways).
Re: (Score:1)
1. He wants to better control the release of email, by distributing it in pieces
2. Email is a bit easier to re-route in case of censorship
Re:no sympathy (Score:4, Informative)
Re: (Score:1)
Sounds like (Score:4, Informative)
an Afilias issue, not a Spamhaus issue.
Secondly, how sure are you somebody didn't forward your email to their own not-so-double-opt-in list which got reported as spam.
Re: (Score:3)
Indeed.
1: Create a new e-mail address at a free service
2: Subscribe to various opt-in services run by people you don't like
3: Forward all e-mails to this address to reportphishing@antiphishing.org and linford@spamhaus.org
4: Schadenfreude
If your e-mail list can't deal with this, you may want to fix the last part of #2, or use a different method of propagation, like RSS.
Re:Sounds like (Score:4, Insightful)
Secondly, how sure are you somebody didn't forward your email to their own not-so-double-opt-in list which got reported as spam.
2/10 domains were blacklisted by Spamhaus, which means 2/10ths of his e-mail list might be contaminated.
It shouldn't be too much of a hassle to subdivide those users and flush out the one(s) which are causing the problem,
Ideally, you'd notify Afilias ahead of time so that they don't blacklist your honeypot domain(s).
Re:Sounds like (Score:5, Insightful)
If that's what happened, it sounds like a DOS attack waiting to happen.How long do you suppose it will be before someone sets up an operation to spam your competition's websites to get them plonked.
Re: (Score:2)
an Afilias issue, not a Spamhaus issue.
Agreed!
Secondly, how sure are you somebody didn't forward your email to their own not-so-double-opt-in list which got reported as spam.
This would not add the proxy servers listed in the email to the DBL. Blocklists are created by logging the source of the spam, not by searching through the text of the spam for possible domains then listing those domains as spammers (although such content filters are useful for identifying messages as spam)
Re: (Score:2)
This is wrong. Spamhaus maintain four blocklists (and an aggregate blocklist) and the DBL is exactly as the grandparent described. It blocks the domains that are found in the content of spam messages. Not the IP addresses of the domains... the actual domains.
I stand corrected. Not for the first, or last time...
But... that is a horrible concept for a block list! I can see using such information to identify potential spam messages by content filtering, but since they are not the originators of the message I don't see how walling them off will in any way reduce the amount of crap in my inbox. I guess it is an application of the "starve them out" method. If you cut off business to those paying the spammers, they will stop paying the spammers, and then spammers w
The Internet is badly regulated (Score:1)
The real problem here is the lack of real procedures and rules. This is just like the FBI seizing domains that were declared legal in their jurisdictions: stupid problems that harm everybody. If there was a nice and clear set of rules, and a single international authority, none of these things would happen.
Re: (Score:2)
You want a .info domain you get to deal with the (silly) policies that TLD registrar enforces.
Don't like it? Don't register a .info.
Re: (Score:2)
Doesn't seem like there was a lack of procedures or rules at all. Read the terms of service agreement that will likely say they have a right to take down the domain if they feel like it. Additionally, you can't use it for illegal/abusive purposes of which they are the sole determiner if what you are doing is illegal or abusive.
Why don't you just register a single .com domain and run your stuff from there. Sounds like a large number of people think what you are doing is spam or aiding spammers and you don
Re: (Score:3)
>Why don't you just register a single .com domain and run your stuff from there
Because that would make it easy for China, Iran, and other regimes to block users from using his services. That's what he *does* -- he enables people stuck behind oppressive (often government-run) firewalls to get to blocked sites & surf without frustration and/or fear.
Re: (Score:2)
Doesn't seem like there was a lack of procedures or rules at all. Read the terms of service agreement that will likely say they have a right to take down the domain if they feel like it.
In other words, there is only one rule and its says that there are no procedures or rules.
Re: (Score:2)
No illegal activity? (Score:5, Insightful)
He claims that no illegal activity was taking place, but if he's running proxies that are open to the public via a mailing list, doesn't it seem entirely likely that a spammer may be making use of his mailing list to get more proxies that can be used for their operations? And, if so, isn't it entirely likely that that's exactly what got him blacklisted in the first place?
What evidence is there that his proxies weren't being used by others for illegal activities? Seems like he conveniently skirted that point in his entire write up.
Re: (Score:2)
It doesn't matter.
The registrar is the one who revoked the domains, so it's the registrar's decision on how or even if that decision can be appealed.
Only if they actually give a crap what spamhaus has to say later would spamhaus's word even matter.
It's called the law of "My box, my rules. Don't like it, take a hike"
Sounds likely, that his proxies WERE used for spam (Score:1)
Re:No illegal activity? (Score:5, Informative)
You must be new here. Bennett is quite well-known in anti-spam, and anti-censorware world. While you were in diapers, he testified in Congress against COPA. He runs peacefire.org - dedicated to free speech for those who are under 18. Accusing him of supporting spam in some way is ridiculous.
http://en.wikipedia.org/wiki/Bennett_Haselton
http://en.wikipedia.org/wiki/Peacefire
Re: (Score:2, Insightful)
That he runs peacefire isnt necessarily a mark in his favor. The idea that people have a right to circumvent filtering on computers they do not own is about as equally shady as whats being discussed here.
Theres "fighting for an ideal", and theres "going over the edge".
Re: (Score:2)
Hi, I read the wikipedia articles linked by the GP. Your assertion doesn't seem to match what was presented there. Do you know something the GP/wikipedia does not?
Re: (Score:2)
If you read the "about us" on peacefire, he makes it pretty clear that his goal is to help minors circumvent filtering. What filtering? Why, filtering at libraries, schools, home computers, etc. But theres a big problem here: those computers dont belong to the minor, and thus the minor has absolutely no right to circumvent it-- it is likely, in fact, that doing so is a violation of whatever acceptable use policy is in place at said location.
Think about it: if these individuals owned / had the right to
Re: (Score:2)
I thought his goal was more about circumventing, highlighting and opposing filtering that doesn't have ethical human oversight and due process. I remember several occasions over the years where filtering software has made the news for blocking sites for "pornography", "gambling" or the like that were actually about something completely different - e.g. just happening to have opposing political or religious views to the producers of the software, or just plain laziness on the part of the filter makers. I als
Re: (Score:2)
His focus is on getting minors around filtering software. That is quite certainly different than fighting national internet filters; the one is a government censorship issue, the other is a "go buy your own internet kiosk" issue.
Re: (Score:2)
1) Child protection. They are minors, and we have a duty of care to protect them, while in our care, from the dangers of the unfiltered internet. Don't tell me it's not necessary; The first time one of them finds rotten.com or sends a link to Goatse, it's game over publicity wise.
2. Network protection. We don't want kids downloading the latest and greatest screensavers to our (Yes, OUR, paid for by your taxes) PCs, meaning we
Re: (Score:2)
Sorry, I fully support a parent's right to use filtering for their child, on a computer that the parent owns.
The parents may have a right to use filtering but the child still has an even more fundamental right to seek and receive information. It is part of the fundamental human rights afforded to everyone regardless of age.
If the child doesn't care and can live with the filters, fine. But if the child hits them and find that its access to desired information is blocked, and starts looking at peacefire for help, it's no longer fine. It's a human rights violation.
I have personally disabled filtering on many public l
Re: (Score:2)
If you are disabling software on Computers you don't own, you can be charged with a felony under the CFAA.
You are not authorized to make those changes, and by exceeding the authorization of what you can do, you have violated the law.
IANAL, but a prosecutor could have a field day with you.
Re: (Score:2)
The parents may have a right to use filtering but the child still has an even more fundamental right to seek and receive information.
Problem: Its not the child's computer. There is no fundamental right to use / own a computer.
I also disagree more generally with your notion of fundamental rights, but we can just stick with the whole "you have no right to anything on someone elses computer".
Re: (Score:3)
That he runs peacefire isnt necessarily a mark in his favor. The idea that people have a right to circumvent filtering on computers they do not own is about as equally shady as whats being discussed here.
Theres "fighting for an ideal", and theres "going over the edge".
Actually Peacefire IS a mark in his favor - free speech (which includes the right to seek and obtain information) is so fundamental that it trumps the right on those who happen to 'own' the media through which this happens. I fully support this. Children has a right to obtain exactly the same information as an adult. We can discuss whether it should be 'offered' to the children, but if they decide to seek it, they have a right to obtain it. The job of the adults are to guide and advice the children in using
Re: (Score:2)
Actually Peacefire IS a mark in his favor - free speech (which includes the right to seek and obtain information) is so fundamental that it trumps the right on those who happen to 'own' the media through which this happens.
Lets say I have filtering software on my computer. If a friend and his child come over, and I permit the child on my computer, are you saying he has a right to start mucking around with the OS to remove the filtering on MY computer?
Because lets be clear, the school computers do NOT belong to the child. If he has a problem with that, he can get his own computer.
Even assuming for the moment that there is some "fundamental right to information" (odd how every day a new fundamental right is created), why does
Re: (Score:2)
No one said he violated any laws. The assertion was that his servers were being used for illegal activities. I then suggested that the illegal activity was being engaged in by spammers who were making use of his service (whether with his knowledge or not is something I don't particularly care to address, though I did note that it was obviously absent as a consideration in his write up).
Re: (Score:2)
I'm not accusing him of supporting spam, merely of providing a service which would incidentally be useful to spammers in addition to his target audience, which I suggested may have been the actual problem here.
Also, as it's been pointed out, those initiatives don't exactly indicate that he takes an anti-spam approach. If anything, he's skirting a fine legal line in some of what he does.
And as a quick aside, I don't recall wearing diapers while in high school, but since you insist I was at the time he was te
False Positives... (Score:2)
Neither of those prove these are false positives, infact given the background as public proxies is suggestive they have been abused.
Crooks are just as likely to abuse his rules as they are RFCs.
Its not Spamhaus, its your registrar (Score:5, Insightful)
Your registrar sucks, its nothing to do with Spamhaus.
Very true - really depends on the registrar (Score:5, Informative)
From NameSilo regarding an alleged malware domain:
From Spamhaus:
So in short - the registrar asked for evidence that the domain was violating their terms of service and spamhaus simply replies they are cybercriminals... trust us! After seeing other abuse reports from them, I can tell you that spamhaus has a very snub attitude and expects to be listened to. Once when Namesilo did not listen to them enough to their liking, they added namesilo.com to their RBL - they had me modify their MTA to route email around the block, but still - I think you can see the problem here - someone has to keep spamhaus in check.
Nobody keeps Spamhaus in check (Score:5, Informative)
Unfortunately nobody keeps Spamhaus in check, that's why they've become a degenerate network abuse source. Their DBL shows them for what they are now, something the rest of the Internet needs to shun.
Spamhaus is better than you think (Score:3, Informative)
Spamhaus always send an automatic notification to abuse@YourDomain.info, if they add you to the blacklist. I suspect you may not have configured an MX to receive mail on these domains. If you had, you would have received a notification.
Re: (Score:3)
I tried this when I ran an Anon server, but the e-mail from Spamhaus kept getting filtered into the spam folder. Kind of hard to find the e-mail among the 1000's of other spam e-mails received to a publicly listed abuse address.
Re: (Score:2)
No, they don't send any such thing.
Re: (Score:2, Interesting)
Ah.....no they don't.
On two ocassions Spamhaus blacklisted one of my corporate sub domains. No notice to any of my contact E-mails {abuse, info, technical, root, admin, webadmin, emailadmin, help, etc}. Just suddenly blacklisted it....I have no explanation why because they certainly would not have received any E-mail from it....those domains don't send E-mail....the domains just receive from a very specific set of customers.
The reason I found out both times was a customer who used spamhaus was having troubl
Re: (Score:2)
Welcome to cyberspace (Score:4, Funny)
Where administrators are gods of their personal fiefdomes and you have no say in anything unless you happen to own the wires or boxes yourself or are in the good graces of whoever does.
Suck it up.
Terms of Service (Score:2)
Look on OR-TALK (TOR mailing list) for all the problems those folks have with VPS providers and the like
Re: (Score:2)
Ferengi rules of acquisition
#253. A contract without fine print is a fool's document
blame games, politics and spammers (Score:1)
"If Spamhaus signed up one of their "spamtrap" email addresses to our mailing list" ...
If you are that freakin paranoid, then you KNOW you are doing something that agitates those of us
that have to deal with the end result of your "work" on a daily basis.
Only you have the power to clean your lists. Go forth, my son. Empower yourself and waste not
another breath in the realm of wizards.
Spamhaus and RBL = evil (Score:5, Insightful)
It was a good idea in the beginning; Getting network and system administrators to share their stories of problems on the frontier. And for awhile, it was good. But as these services developed, they decided to start automating the process. And that's when the problems started. As an example, let's say all spammers use open relays. The logic here then is to test for open relays and block any that are found. Spam problem solved! Except it doesn't look at the reverse case: Namely, that not all open relays are used by spammers. In fact, it could be the case that the vast majority of open relays are perfectly harmless and have a legitimate reason for existing.
Now I'm not trying to discuss open relays from a technical standpoint, or the arguments for or against them -- what I'm trying to show is the logic problem in assuming that just because when 'A' is often found next to 'B', that means that 'B' is often found next to 'A'. That's the crux of the problem with the RBL and Spamhaus -- it's a logic fail of epic proportions.
Automation is attractive because it can catch things faster and with greater accuracy than humans can. But humans are better at making judgement calls, looking at the evidence, and problem resolution with other humans. Spamhaus and the RBL fail here because they implimented the automation and then because of their perceived success, they decided Automation Was God and made appealing the decision of its robot overlords increasingly difficult if not impossible. And that's when Spamhaus and the RBL became evil: The process stopped being overseen by humans, started to assume everyone was an evil spammer, and that the solution in every case was to follow the De Facto Anti-Spammer Laws as laid down by its robotic overlords. "Fix your open relay!" became the reply, instead of checking to see whether said open relay had actually sent any spam, or whether there was a good reason for its existance (again: No debates about open relays please! It's just the example!).
Of course, spammers got smarter and started coming up with more sophisticated methods of injecting their crap... which led to more complex robots, and as each new counter-measure was rolled out, the reply to hapless admins caught in the motorized wheels o spammy justice was "It's your problem, not ours!" My advice to system and network admins these days is to not use spamhaus or the RBL, or if you must, make sure your mailboxes and such are setup similar to how gmail and many exchange servers are: Have a separate spam folder, and give the user the option to whitelist anything your filters catch. Ultimately, you're providing a service to them... you have no duty or obligation to anyone else. Make sure they can use what you've given them.
Re: (Score:2, Insightful)
"My advice to system and network admins these days is to not use spamhaus or the RBL, or if you must, make sure your mailboxes and such are setup similar to how gmail and many exchange servers are: Have a separate spam folder, and give the user the option to whitelist anything your filters catch. "
you, sir, must have unlimited network resources. With spam taking up +90% [1] of internet traffic, you just rolled over and admitted that you
weren't as skilled as the opposition and let them sap your resources. I
Re: (Score:1)
With spam taking up +90% [1] of internet traffic
Err... and peer-to-peer presumably accounts for the other 90% of internet traffic?
Hint: you misread your source in spectacular fashion.
Re: (Score:2)
you, sir, must have unlimited network resources. With spam taking up +90% [1] of internet traffic, you just rolled over and admitted that you weren't as skilled as the opposition and let them sap your resources. I was hoping for better advice.
Yeah, let's have a look here at my current google spam folder... okay, about 64 messages. Each message is at best about 4KB in size. 4 * 64 = 256KB of spam per month. But let's quadruple that, because maybe my mailbox, which has been around since 2003 and subscribed to approximately a hundred lists, is lower than average. Comcast states that the average user uses 1-2GB per month; Ludicriously low, but for the sake of debate let's say the average user only uses 1GB of bandwidth per month. That means that spa
Re:Spamhaus and RBL = evil (Score:4, Informative)
Re: (Score:2)
You're proceeding from a faulty premise. You're assuming that you are seeing all of the traffic being sent to you.
My "premise" is that e-mail makes up a very small minority of internet traffic. I'm arguing against the size of the problem as automatically justifying extreme and extraordinary measures to control because of its severity.
Even if 99.999% of all e-mail is spam, the author's original assertion is busted: E-mail makes up a very small amount of total internet traffic. The idea that filtering is mandatory is silly -- even if 100% of that spam went through, it would be a drop in the bucket compared to the amount
Re: (Score:2)
Re: (Score:2)
greylisting stops a lot more spam than blacklisting, and has a zero false positive rate as long as the originating server follows the rfc detailing how smtp is supposed to work.
Re: (Score:3)
greylisting stops a lot more spam than blacklisting, and has a zero false positive rate as long as the originating server follows the rfc detailing how smtp is supposed to work.
Back in the real world, greylisting is anything but a panacea and has its own set of impacts on email.
Different greylisting implementations remember send attempts differently.
Senders relaying through outbound services (yes, we run our outbound through antivirus/malware/spam filtering) with outbound farms or ranges of addresses may never be allowed through as their email presents from a different IP each time and greylisters don't all follow greylisting best practices by whitelisting the large outbound 'farm
Re: (Score:2)
Namely, that not all open relays are used by spammers. In fact, it could be the case that the vast majority of open relays are perfectly harmless and have a legitimate reason for existing.
I'm trying to think of one...
Re: (Score:2)
"The process stopped being overseen by humans, started to assume everyone was an evil spammer..."
How is SORBS these days anyhow?
Re: (Score:2)
"The" RBL...
"that's when Spamhaus and the RBL became evil" ... "is to not use spamhaus or the RBL".
What is the RBL? It generally sounds like you're a native speaker of English, so I'm wondering why you're using the definite article ("the") for this. Because there is no single RBL, neither as a (non-Spamhaus) blacklist with the name "RBL", nor as a product of Spamhaus's called "RBL", nor as a distinguished technology/method called "the RBL". There are RBL's.
Not All A's Are A Certain Kind Of A ...
No, but i
People forget, or stop caring ... (Score:5, Insightful)
Sooner or later people forget they signed up, stop giving a damn, or otherwise get tired of what you're sending.
If they can't figure out how to get out of it (because, really, who is going to respond to something they think is spam to make it stop), they'll flag you as spam.
Or, something automated comes along and decides that whatever you're sending is spam.
As long as it stops coming when people get tired of it ... they really don't give a crap about what happens to you.
Maybe the domains were alread on the list. (Score:1)
Be a rebel (Score:2)
Post direct IP adresses to your proxies.
OK, so I read the rant... (Score:4, Interesting)
In summary:
1. You run a mailing list
2 You *claim* that it's opt-in
3 Somehow Spamhaus gets your list in its honeypots
4. Spamhaus lists you
5. Afilias nukes you, all 10 of your domains.
6. You easily get your domains off Spamhaus by filling out a form
7. Somehow this is Spamhaus' fault and not Afilias for giving you the run-around
Spamhaus has servers that collect spam from the internet by just being on the internet. Spammers blindly send mail to addresses and the Spamhaus servers read the headers to see where they came from. Headers can be forged, but a good algorithm can do the same thing that a human does when reading a header - follow the chain of Received: until it hits the inevitably forged nonexistent or non-sequitur domain. The one before that gets listed at Spamhaus.
Spamhaus has no users on its honeypots that subscribe to lists. They are just "there" on the net silently collecting spam and they give no 5xx or 4xx errors (because, you know, why bother?). The only way for the honeypot to get messages from you is if your list actually contains the addresses of the honeypots.
Spamhaus has a good reputation. They are probably the most reliable blacklisting service out there and this maddens spammers to no end. There are others that shouldn't be used, but Spamhaus is used by nearly everyone who uses a blacklist because of its accuracy.
>If Spamhaus signed up one of their "spamtrap" email addresses to our mailing list
It doesn't work that way. Clean up your list.
--
BMO
Re: (Score:1)
>If Spamhaus signed up one of their "spamtrap" email addresses to our mailing list
It doesn't work that way. Clean up your list.
--
BMO
More precisely, spam traps are real, valid e-mail addresses that fall in disuse. When hotmail or gmail or whoever finds a mailbox that hasn't been opened in a long time, it has two choices: close the account, or set it up as a spam trap.
If you follow proper list higiene, you shouldn't reach this stage ever. Make sure you unsuscribe addresses that bounce more than once, sign up in ESP's feedback programs, and make sure your list keeps current. The very simplest thing would be to send a reconfirmation mail fr
Re: (Score:1)
That's not how I read it, although I think he is being obtuse on purpose.
1. He runs open proxies
2. He runs a mailing list advocating said proxies (and others?)
3. He registers 10 new domains to run proxies on
4. proxies running on 2 domains are used to send spam which ends up on a honeypot
continue at your 4.
As far as I can tell, Spamhaus does exactly as advertised, Afilias does exactly as advertised, and he is pissed that the world doesn't bend to his will.
Don't kneejerk react, readers (Score:5, Interesting)
Don't talk to him like a noob, people. Bennett has been around a very, very long time. He has had a beef with DNS distributed blocklists for most of that time. Others publishing their opinions gets in his craw when it interferes with his operations. He comes in here periodically with his latest incident to rally the "freedom to do whatever I want" crowd into a frenzy. He also posts lots of other stuff worth reading. *grin*
If one considers the DBL a list of domains who have appeared in emails to spamtraps, then I would contend that it very possible that the "zero false positive" claim holds up because it very well might have happened. If it claims that all listed entities are domains owned by spam operators, then he might have an argument.
Haselton's fundamental gripe is that he should be free to communicate until a real person decides he shouldn't. The fact that automated systems now make the blocking decision, requiring human intervention to override them, is an inverted model compared to the "old internet." (The necessity came from the raw volume of spam) The death of the "old internet" began with Canter and Siegel [http]. Some of our long-term, asylum residents just haven't accepted that fact.
Re: (Score:1)
Yeah, I know who he is, and to be forthright, he was not being honest in his article. OK, you could make an argument that he's just a principled wearer of a tinfoil hat, but I believe he was actively seeking to deceive. He knows perfectly well that spammers abuse proxies like his all the time and that they see far more use by spammers than by people actually evading censorship. He also knows, or should know, perfectly well that Spamhaus did not put those domains on a blocklist because they were on his mail
Re: (Score:2)
If one considers the DBL a list of domains who have appeared in emails to spamtraps
Fail. That is not how the lists are generated. The domain would have to be seen as the source of the email in order to be added to the blocklist. Simply having your domain name appear in the text of an email which has been flagged as spam is not going to add your domain to a blocklist.
Re: (Score:2)
Simply having your domain name appear in the text of an email which has been flagged as spam is not going to add your domain to a blocklist.
This is wrong. Some of the best (least bad, if you prefer) blocklists are exactly that: lists of domains which have appeared in spam emails for victims to click on.
Flaw in basic concept (Score:2)
Not significantly. Sure, they have to join with multiple recipient email addresses, but that's not that much of a burden. There really is no way you can use email lists or similar direct-distribution methods to get information to anonymous strangers who you want to have the information and simultaneously keep it out of the hands of people you d
Blacklist owners are never contactable (Score:5, Interesting)
I do not believe it is possible to be contactable and run a blacklist. It would require an army of support people, and most of the blacklists just do not get the kind of income necessary to pay for that.
Blacklists are a pain to deal with in general. Some simply hold you for ransom. Yet it is also a pain to run a mailserver without blacklists, so... Spamhaus has fewer false positives than most, in my experience, but it is stupid of them to claim that any list has zero of them.
we've had a blacklist problem... (Score:2)
Spamhaus DBL is not the problem (Score:2)
The misuse and abuse of the spamhaus DBL is the problem.
It was never intended as a tool for registrars to use in vetting customers.
It does not (as the OP suggested) add entries based on their inclusion in a list contained within an email message.
It does collect, and collate, information from email providers, users, ISPs regarding domains from which spam has been sent. If the OPs mailing list were the problem, the domain from which the list is sent would be the one marked as a spammer if that were the case
Re: (Score:2)
Check me on this.
Haselton has long been an advocate of open mail servers. For the longest time he claimed to have been running one and that he had his own system to control the spam through it. I admit I never really cared what his system of control was. He continued to run one at the same time the industry was quickly realizing that open mail servers were a bigger nuisance than they were worth, so were locking them down to send outgoing mail only from their internal netblocks and terminating the spammers o
Legit .info user? You must be the first. (Score:5, Interesting)
I've yet to receive any piece of e-mail from a .info domain that wasn't spam. Simply matching on .info is the most reliable filter I've found for identifying e-mail from scumbags who deserve death.
Anyone else notice this?
Re: (Score:2)
Sigh .. this type of stupid attitude reminds me of the bad old days when I would listen to blowhard admins arguing (I'm not kidding, I still had arguments on /. about this) "I've yet to see legitimate email from China, so I've blocked all emails from China automatically". Apparently nowadays morons block entire top-level domains .. that is a massive WTF.
Re: (Score:3)
I dont have any users on my server from russia and most if not all of the -stan countries...
Blacklisting the whole ip-ranges in my firewall cut down the brute-force login spam from tens of thousands to the occasional one.
Hardly a solution for a proper service, but for a private server used mostly by friends it is a simple solution to an annoying problem :p
Re: (Score:2)
What is stupid about looking at my own data set of tens of thousands of data points with 100% corelation? Are you trying to tell me I can't do the math? My situation may be an anomoly, and it may be different from yours. But for the mail that hits *my* domains, I have *never* gotten a legit mail from a info domain. That's my data. Its *not* attitude.
Re: (Score:2)
Nah, I'd see it. Eventually. All .info mail goes into it's own, special little spam folder. Unless spamassassin scored it 10 or higher, I'll see it eventually. I clean the folder out once a week or so. And I've never once gotten a piece of mail in there that I wanted. YMMV, but that's my data set. And it is mine. It may be an anomaly, but that doesn't mean I can't do the math.
Your password was (Score:1)
I have to challenge this (Score:5, Insightful)
Like the subject says, I have to challenge the claim that Spamhaus is wrong (full disclosure: I've been professionally involved in email and web security for more than a decade, but am not, and have never been, affiliated with Spamhaus. I do, however, hold them in high regard).
First of all, when I went to those domains, what was the first thing that caught my eye? "Get a green card" ads for usagc.org. I'm not specifically accusing usagc.org of spamming, but these sorts of businesses are most typically advertised by spam. I'm sure you've seen some.
Next, those sites are open proxies (by design). Anyone can create a URL like this: http://rootface.info/ojgnl.php?ZlQc9TMpAmsr3onaDWV0g=t1wn6QmM0TaAEo7rD%2F%2Bm%2Fy%2B365U2AwdnE4VH60DF8%2BU%3D [rootface.info] (nothing dangerous, it goes to cnn.com, but of course, you shouldn't trust me) and send it out in spam advertizing whatever they want.
Finally, you do not appear to state anywhere in your article that Spamhaus said your proxy mailing list was the source of the spam complaints (although they would not tell you if it was), and I doubt that it was. The most likely scenario is that someone abused your proxies to send spam, and since running an open proxy (regardless of noble motive) makes you complicit in that abuse, Spamhaus listed those domains.
Whether the registry's actions were justified or correct is a separate consideration. Maybe they were, maybe they weren't, but you are claiming without evidence that Spamhaus made a mistake. I'm pretty confident they didn't, for the reasons outlined above.
Re: (Score:3)
The two domains in question probably really were used in spam
Used TO spam or used IN spam? That's what's been confusing me about the whole exchange. You could proxy to gmail.com and send spam but it'd still be a gmail.com address.
Re: (Score:2)
domains found in URL's in spam emails are blacklisted too, in separate blacklists. Those blacklists are extremely effective in my experience, and have very few false positives.
Wait now... (Score:2)
Re: (Score:2)
The TOS might not be fair but it reflects that someone other than you owns the wires and boxes, and that you are beholden to them in all things so long as you wish to do business with them.
Businesses are not obligated to be fair to you unless the government says so.
Re: (Score:2)
I ran an anon-proxy for years and the above is not true.
Not all providers prohibit proxies in their TOS!
If you buy the T1/DS3/high speed connection you can set up a server and run your own proxy. I had my own T1 at the time and the only issue I had was my upstream cutting me off from time to time and I would fight them to get the connection back on. That lasted until I had my lawyer send them a letter.
Proxies are not just used for spam, mine was used by millions of people in places like Saudi Arabia, Lebano
Re:not suprising (Score:5, Informative)
GIven that his article was about him setting up 10 new proxies and emailing them out, it would seem that, at least for the domains relevant to this discussion, the OP was the owner/administrator and most definitely intended them to be used in that manner. Also, from context, it appears that he was running webproxies, not email proxies. They're generally used as anonymizers, or to circumvent geo-IP techniques, not to spam people.
Re: (Score:2)
I agree, this is a bad analysis from the Bennett Haselton. He is looking this from the wrong angle.
As his mailling list is about open-proxies, some spammers are in his list and after getting the new address for open-proxies, they started to use then (only 2 it seems). After spamhaus got spam from those IPs they blocked the IPs and sent the usual emails to the abuse@ ip owners and dns registers (and possible too dns owners).
Afilias got the abuse@ report, checked the username and blocked all domains of the s