At one time quite recently, Cloudflare was on Spamhaus's "Don't Route Or Peer" list. Imagine where the world would be if anyone cared about that list.
But if an entire ISP is blocking you based on UCEPROTECT, which is designed for email spam filters - are you sure? An ISP that blocked all of DO would get so many complaints and be open to so much legal liability. Usually these are only used for email filtering.
>At one time quite recently, Cloudflare was on Spamhaus's "Don't Route Or Peer" list.
it was only 2 /23's (only 1024 IPs, ASN14061 is 3,140,680 IPs), and was resolved without a monthly subscription in ~24 hours after being posted to the NANOG mailing list
the scale of the issue is significantly different and the resolution is significantly different (both in time and in money). that's important when comparing the situations.
highlighting those differences does not mean i think either situation is "good". one of them is "less bad" though.
Spamhaus does not offer the ability to pay for removal. UCEPROTECT is uniquely evil in insisting on payment. It's a good thing that their lists are not widely used, so, in practice, it doesn't matter if your IPs are listed by them. Their revenue comes from unwitting victims who don't know any better.
Their lists are definitely used, which is the only way I knew about them - one email provider was not receiving email from my DO hosted site at all. I knew enough of what I was doing to know this was practically impossible, spent over a week running every kind of configuration scan and test I could think of, and finally narrowed it down to this organization in particular blacklisting all of DO IP space.
I disagree. Normally you can reach out to request delisting, and it's free.
As I explained, Orange's security filter seems to automatically block requests at the DNS level if your website's domain resolves to an IP listed on UCEPROTECT Level 3. In this case, that appears to be the entire DigitalOcean ASN.
"This time, however, we have a premiere: A Woman. Actually, it reads more like a brat than a woman [...] If we were as emotional as you are and report you to our authorities, the handcuffs might click the next time you move your ass off the British Isles. [...] This means that they have heared that bullshit you spammed to them at times when you still slipped across the floor with your shit diaper. :-) [...]
Grow up and also try to grow a brain before you make a fool out of yourself, again next time.
The uceprotect Level 3 page clearly says using the list will cause collateral damage and should only use as an indicator as part of a spam score, not to act purely on it.
So, yes, if a service provider is known for having a lot of spam coming from it, it makes sense to be on there. They're not saying every IP is bad, they're literally stating the opposite (which seems fair).
So, if Orange are blocking purely based on it, not only are they using a spam filter as a DNS query/web filter (given uceprotect seems to flag based on mail spam traps based detections etc.) and ignoring their documentation (which says it shouldn't be used to block on it's own).
DigitalOcean has been on their list for a while, they say something like “if you don’t want to be flagged as spam, dont host on a site that hosts spammers” which is pretty ridiculous given every cloud platform does
Hosting providers should be good citizens of the network and immediately terminate spammer accounts. This used to be standard practice until about 2015. When suddenly management decided it’s more profitable host spammers than not.
I think this is a great policy decision by uceprotect.
Why DigitalOcean in particular though? Why not Cloudflare? A significant percentage of malicious and spam traffic is hosted on or routed through cloudflare services. Why not AWS? Same thing. It's a personal grudge, and they aren't quite big enough or have powerful enough lawyers to cause a widespread problem if put on a list, so, the big players are left alone.
All that aside, the website and their communication around this over the years is extraordinarily unprofessional and it's astounding to me they wield such power.
I hope all the people who think that the Internet is hostile now understand…
the Internet used to be almost entirely composed of this guy.
Like in 1996 if you wanted to do something on the Internet you had to deal with some version of him. I worked for the top car electronics installer in Houston in the 90s, who ran a few websites as a hobby and he was 100% like this.
Literally sat and typed out Monty Python dialogue so he could just have it hosted on his page, and would make .WAVs of annoying customers to put on the site
This is not new, "UCEPROTECT" has been a garbage RBL doing this for decades (even listing blocks preemptively, so that when they're assigned they're unusable), nobody worth emailing actually uses it.
Do you have any direct evidence that the Orange block is due to the listing in UCEPROTECT?
Digital Ocean is not a great neighborhood.
After seeing periodic bursts of login attempts from there, I now block all inbound ssh from AS14061 (among several such bad neighborhoods) as no authorized clients are hosted there and there is a constant level of attempted attacks from there.
Sounds like this should go via Orange they have means to slap UCEPROTECT on the wrist. It also looks like it would be in Orange business to have proper spam lists not ones like that.
I personally found Meteor in Ireland (previously owned by Orange) to be far more liberal with blocks on an unregistered mobile internet connection than other providers.
Also, Orange UK have a much stricter block list to enforce than, for example, Orange FR.
Playing the devil’s advocate, my domains get constantly scanned from DO ips. I know it’s not feasible but some basic kyc on DO’s side would probably help.
I use my least reputable domains to send emails that are most important for the recipient and least important for me. If they don't receive it now it's their problem not mine.
That's the thing a lot of people don't understand about email. Delivery is not guaranteed. And neither is notification about failed delivery. Bits can just fall on the floor somewhere and you'll never know.
That is technically true for any communication method. In practice, email is extremely reliably except for deliberate decisions to drop mail. You can literally shut down your mail server for days and some time after you boot it back up you'll get all the mail people have been sending during that time. What other protocol gives you that reliability without complex failover infrastructure.
I know of one regulator in USA (WSLCB) that sends critical and time sensitive notifications via email only. Has had delivery problems for at least five years and cannot see the flaw in their system.
I think what's happening is obviously not fair, and should be corrected, but I also what to say that I've had pretty bad experiences when working with DO. I think maybe what happens is that they take too long to weed out problematic accounts. Not going to mention any other provider so it's not assumed that shilling for someone else, but there are quite a few that are no in the AWS, GCP, or Azure tier and work just fine, so, shouldnt be too hard to find a replacement.
Apparently in this uceprotect thing my home ip (business connection, fixed IP) has a "dns problem" because there's a PTR record at my ISP but not an A record.
For a long time, the Internet has been replete with blocklists and reputation scores cultivated by admins who had the means to track such things. And ad-blocking software/DNS often gives users the chance to tap into those bad reputations and protect themselves.
However, there is nothing preventing function/scope creep of these blocklists into things they should not be. Political bias, censorship, morality policing will trickle into blocklists. Furthermore these false positives are quite onerous for legit businesses and customers who sincerely want to connect. I've connected again to an ad-blocking DNS service, and many people may subscribe to filtering services, or simply be involuntarily subscribed, in the hopes that their Internet would stay usable, and scam-free.
I suppose this is the price to pay in low-trust society (wild and wooly Internet). I wonder if the Great Firewall of China obviates the need for their citizens to throw up such protective measures.
Forgive the snark, but I find it incredibly surprising that in your 20-year career you haven't arrived at the conclusion that self-hosting your own outbound email is not worth it, especially if you don't own and/or manage the reputation of your IPv4 block.
I was reviewing uceprotect policies a couple of hours ago and what you say is really strange.
Uce level 3 should never be used alone to block general tcp traffic.
Are you saying that Orange is using uce level 3 to block email sent from your DO vm? Are you really sure that Orange blocks for this list only?
Is it possible that your ip is on another RBL?
If Orange is really blocking smtp based only on uce level 3 they are doing wrong: their system may be mis-configured or they have not read correctly hiw level 3 is populated.
Or - probably - they did on pourpose!
Today I checked september smtp traffic: almost 99% of traffic coming from uce level 1 (single ips) was spam. Traffic hitting level 2 but not present in level 1 (subnets, one I checked was /15!!) was also 99% spam.
Of the 5000k messages received, very few arrived into the mailboxes. A lot of messages were blocked by following spam checks and the few remaining were flagged as spam.
I feel your pain. If you host on DO and you send email, you should look for a email provider that will funnel your emails.
Ps: i repeat, email shoild not be bloocked only by uce level 3
All of the spam blocklists are like this.
At one time quite recently, Cloudflare was on Spamhaus's "Don't Route Or Peer" list. Imagine where the world would be if anyone cared about that list.
But if an entire ISP is blocking you based on UCEPROTECT, which is designed for email spam filters - are you sure? An ISP that blocked all of DO would get so many complaints and be open to so much legal liability. Usually these are only used for email filtering.
>At one time quite recently, Cloudflare was on Spamhaus's "Don't Route Or Peer" list.
it was only 2 /23's (only 1024 IPs, ASN14061 is 3,140,680 IPs), and was resolved without a monthly subscription in ~24 hours after being posted to the NANOG mailing list
https://lists.nanog.org/archives/list/nanog@lists.nanog.org/...
Why does that make it good?
i'm adding context to your comment.
the scale of the issue is significantly different and the resolution is significantly different (both in time and in money). that's important when comparing the situations.
highlighting those differences does not mean i think either situation is "good". one of them is "less bad" though.
Spamhaus does not offer the ability to pay for removal. UCEPROTECT is uniquely evil in insisting on payment. It's a good thing that their lists are not widely used, so, in practice, it doesn't matter if your IPs are listed by them. Their revenue comes from unwitting victims who don't know any better.
Their lists are definitely used, which is the only way I knew about them - one email provider was not receiving email from my DO hosted site at all. I knew enough of what I was doing to know this was practically impossible, spent over a week running every kind of configuration scan and test I could think of, and finally narrowed it down to this organization in particular blacklisting all of DO IP space.
I disagree. Normally you can reach out to request delisting, and it's free.
As I explained, Orange's security filter seems to automatically block requests at the DNS level if your website's domain resolves to an IP listed on UCEPROTECT Level 3. In this case, that appears to be the entire DigitalOcean ASN.
You probably have standing to sue Orange for damages if they're blocking your customers from you based on bullshit criteria.
I am not a lawyer, so consult one.
the more ive looked at this uceprotect site, the more i see how unhinged they are.
for example: "People with a brain would simply fix their systems after getting listed for abuse. Stupid losers are different." http://www.uceprotect.org/cart00neys/index.html
or
"This time, however, we have a premiere: A Woman. Actually, it reads more like a brat than a woman [...] If we were as emotional as you are and report you to our authorities, the handcuffs might click the next time you move your ass off the British Isles. [...] This means that they have heared that bullshit you spammed to them at times when you still slipped across the floor with your shit diaper. :-) [...]
Grow up and also try to grow a brain before you make a fool out of yourself, again next time.
Claus von Wolfhausen
Technical Director" http://www.uceprotect.org/cart00neys/2021-001.html
their main page has an iamverybadass quote, too: "WARNING: Do not play around here. You have no idea who we really are, and what will happen to you!"
The uceprotect Level 3 page clearly says using the list will cause collateral damage and should only use as an indicator as part of a spam score, not to act purely on it.
So, yes, if a service provider is known for having a lot of spam coming from it, it makes sense to be on there. They're not saying every IP is bad, they're literally stating the opposite (which seems fair).
So, if Orange are blocking purely based on it, not only are they using a spam filter as a DNS query/web filter (given uceprotect seems to flag based on mail spam traps based detections etc.) and ignoring their documentation (which says it shouldn't be used to block on it's own).
Well, if it's just an indicator, why not delist automatically on request, as other blocklists do? It seems inappropriate to charge for this.
And not just charge. You need to get a subscription even!
Theoretically speaking, a person could get themselves blocked, request an unblocking, and keep on spamming, hoping they won't get blocked again.
right now, they can apparently just spend $108/year and keep on spamming.
This is what capitalist “regulation” looks like in practice
DigitalOcean has been on their list for a while, they say something like “if you don’t want to be flagged as spam, dont host on a site that hosts spammers” which is pretty ridiculous given every cloud platform does
It’s not ridiculous.
Hosting providers should be good citizens of the network and immediately terminate spammer accounts. This used to be standard practice until about 2015. When suddenly management decided it’s more profitable host spammers than not.
I think this is a great policy decision by uceprotect.
Why DigitalOcean in particular though? Why not Cloudflare? A significant percentage of malicious and spam traffic is hosted on or routed through cloudflare services. Why not AWS? Same thing. It's a personal grudge, and they aren't quite big enough or have powerful enough lawyers to cause a widespread problem if put on a list, so, the big players are left alone.
All that aside, the website and their communication around this over the years is extraordinarily unprofessional and it's astounding to me they wield such power.
Following your logic, we'd need to block the entire US, since that's where DigitalOcean's ASN is registered.
https://uceprotect.wtf
Check this...
Thanks! I was worried that the word 'extorting' was a bit too strong, but it seems I wasn't the first one to apply for the UCEPROTECT service.
I hope all the people who think that the Internet is hostile now understand…
the Internet used to be almost entirely composed of this guy.
Like in 1996 if you wanted to do something on the Internet you had to deal with some version of him. I worked for the top car electronics installer in Houston in the 90s, who ran a few websites as a hobby and he was 100% like this.
Literally sat and typed out Monty Python dialogue so he could just have it hosted on his page, and would make .WAVs of annoying customers to put on the site
> Literally sat and typed out Monty Python dialogue
Dear god.
Right, you know exactly the person I’m describing
Lol. Found this URL on the page you linked https://www.uceprotect.org/cart00neys/index.html which actually says a lot about Uceprotect.
UCEPROTECT Extortion Service: All Your Mails Are Belong To Us! (2009)
https://www.aaroncake.net/misc/showthought.asp?thought=57
This is not new, "UCEPROTECT" has been a garbage RBL doing this for decades (even listing blocks preemptively, so that when they're assigned they're unusable), nobody worth emailing actually uses it.
Do you have any direct evidence that the Orange block is due to the listing in UCEPROTECT?
Digital Ocean is not a great neighborhood.
After seeing periodic bursts of login attempts from there, I now block all inbound ssh from AS14061 (among several such bad neighborhoods) as no authorized clients are hosted there and there is a constant level of attempted attacks from there.
Sounds like this should go via Orange they have means to slap UCEPROTECT on the wrist. It also looks like it would be in Orange business to have proper spam lists not ones like that.
While I absolutely agree that Orange is acting irresponsibly here, I also understand that they're not the only ones who rely on UCEPROTECT.
Which Orange are you blocked on?
I personally found Meteor in Ireland (previously owned by Orange) to be far more liberal with blocks on an unregistered mobile internet connection than other providers.
Also, Orange UK have a much stricter block list to enforce than, for example, Orange FR.
Orange FR.
Playing the devil’s advocate, my domains get constantly scanned from DO ips. I know it’s not feasible but some basic kyc on DO’s side would probably help.
UCEPROTECT is why I stopped self hosting my own email :(
I use my least reputable domains to send emails that are most important for the recipient and least important for me. If they don't receive it now it's their problem not mine.
That's the thing a lot of people don't understand about email. Delivery is not guaranteed. And neither is notification about failed delivery. Bits can just fall on the floor somewhere and you'll never know.
That is technically true for any communication method. In practice, email is extremely reliably except for deliberate decisions to drop mail. You can literally shut down your mail server for days and some time after you boot it back up you'll get all the mail people have been sending during that time. What other protocol gives you that reliability without complex failover infrastructure.
I know of one regulator in USA (WSLCB) that sends critical and time sensitive notifications via email only. Has had delivery problems for at least five years and cannot see the flaw in their system.
Is the flaw on their side or your side?
I think what's happening is obviously not fair, and should be corrected, but I also what to say that I've had pretty bad experiences when working with DO. I think maybe what happens is that they take too long to weed out problematic accounts. Not going to mention any other provider so it's not assumed that shilling for someone else, but there are quite a few that are no in the AWS, GCP, or Azure tier and work just fine, so, shouldnt be too hard to find a replacement.
Apparently in this uceprotect thing my home ip (business connection, fixed IP) has a "dns problem" because there's a PTR record at my ISP but not an A record.
Seriously?
The rest of the list is as credible as that?
Why do you think Orange is using Uceprotect? It seems unlikely that big corp would use some weird unknown list.
Many ISPs across EU use Cisco's Talos Intelligence and similar solutions that automatically feed themselves of random blacklists like Uceprotect.
For a long time, the Internet has been replete with blocklists and reputation scores cultivated by admins who had the means to track such things. And ad-blocking software/DNS often gives users the chance to tap into those bad reputations and protect themselves.
However, there is nothing preventing function/scope creep of these blocklists into things they should not be. Political bias, censorship, morality policing will trickle into blocklists. Furthermore these false positives are quite onerous for legit businesses and customers who sincerely want to connect. I've connected again to an ad-blocking DNS service, and many people may subscribe to filtering services, or simply be involuntarily subscribed, in the hopes that their Internet would stay usable, and scam-free.
I suppose this is the price to pay in low-trust society (wild and wooly Internet). I wonder if the Great Firewall of China obviates the need for their citizens to throw up such protective measures.
its like we are in the grifted age
Forgive the snark, but I find it incredibly surprising that in your 20-year career you haven't arrived at the conclusion that self-hosting your own outbound email is not worth it, especially if you don't own and/or manage the reputation of your IPv4 block.
> _website_ was being blocked
I was reviewing uceprotect policies a couple of hours ago and what you say is really strange.
Uce level 3 should never be used alone to block general tcp traffic.
Are you saying that Orange is using uce level 3 to block email sent from your DO vm? Are you really sure that Orange blocks for this list only? Is it possible that your ip is on another RBL?
If Orange is really blocking smtp based only on uce level 3 they are doing wrong: their system may be mis-configured or they have not read correctly hiw level 3 is populated.
Or - probably - they did on pourpose!
Today I checked september smtp traffic: almost 99% of traffic coming from uce level 1 (single ips) was spam. Traffic hitting level 2 but not present in level 1 (subnets, one I checked was /15!!) was also 99% spam.
Of the 5000k messages received, very few arrived into the mailboxes. A lot of messages were blocked by following spam checks and the few remaining were flagged as spam.
I feel your pain. If you host on DO and you send email, you should look for a email provider that will funnel your emails.
Ps: i repeat, email shoild not be bloocked only by uce level 3