Slashdot Mirror
Verizon.net Finally Moving Email To Port 587
The Washington Post's Security Fix blog is reporting that Verizon, long identified as the largest ISP source of spam, is moving to require use of the submission port, 587, in outbound mail — and thus to require authentication. While spammers may still be able to relay spam through zombies in Verizon's network, if the victims let their mail clients remember their authentication credentials, at least the zombies will be easily identifiable. Verizon pledges to clean up their zombie problem quickly. We'll see.
I've been routing my traffic thru their traffic for a few years now, they're not limiting anyone and keep great privacy. what i heard their tunnel service will be open for new customers in a few days again so now is a great time.
Sounds like a great opportunity to charge millions of clueless users $50 to change the setting for them. I see a Vegas vacation on my event horizon.
Nothing worthwhile ever happens before noon
I found out I was a spammer when I investigated a message returned to me. I ended up talking with someone from SORBS. After emailing SORBS a couple of times, I received this message from Michelle Sullivan: "SORBS lists IP addresses that send spam. Often there is real email mixed with the spam, sometimes deliberately, sometimes accidentally. In this case you are using an IP address to send your email that has previously, and is still, sending spam. The IP address is blocked. I'd contact your provider and complain bitterly about it, because it's the provider that is listed, not you specifically."
I send out a newsletter with about 250 subscribers. After talking with SORBS, I contacted Verizon and found out that, even though we signed up for Verizon Business, they limit the amount of email I can send a week to 500 messages. I rarely approach 200 messages and the newsletter is a monthly. Verizon told me I couldnâ(TM)t even send the newsletter in one blast; I had to limit it to 100 subscribers an hour! And in late Fall 2008, some providers, like MS, would reject my mail simply because it had @Verizon.net in the senderâ(TM)s address. I knew I wasn't sending out large amounts of email, let alone spam.
Within those imposed limits, Verizon still could not bring its huge entity to investigate my complaint. In late December, we switch to Constant Contact to email the newsletter. While my boss uses Cox since he works mostly from home, the office is still âoeconnectedâ with Verizon!
Boy, I hate Verizon! Now, maybe they will kill the Zombies from all those dead zones they claim not to have!
=smidge=
Is it just my observation, or is eldavojohn an idiot?
You can set up port 25 SMTP to require authentication for relay purposes, without having to configure end user's machines for another port.
Don't most zombies implement their own SMTP clients? In other words, they wouldn't even use the ISP's mail servers...
I feel a great disturbance in the Force, as if millions of voices cried out in terror and were suddenly silenced...
Can't see how this will prevent sending spam.
Maybe in future zombies have their own built-in sendmail.
Last week I routed an email through PORT 587 and this came out of it:
Hai Adonai Abmozedel, Adonai Garntaturagah, Adonai Hai Prezelbuuub, Adonai Hai Koadze....and so on.
Is their choice really smart ?
This is a good thing, but it's unlikely to improve things in anything other than the short term. They are quite capable of identifying which customers are zombie spam relays already by looking at IP addresses and authentication logs. I did this back in the days of dialup when i did a lot of work on mail systems for another large isp/telco. They are still left with the matter of contacting the customer and explaining the problem and guiding through to a solution. This is expensive to do, and requires hand holding as the customer isn't going to understand what do. It's still cheaper for the ISP to ignore the problem. Zombies will still operate, just now they have to steal authentication details. Big deal.
I say I ain't giving you no tree fiddy you goddamned Loch Ness monster, get yo own goddamned money!
No, the guy posting before you did that ;-)
Comcast has required email to be on port 587 for a while now.
Well your spam made it through, but the response must have been throttled since you didn't get first post. You're a Comcast customer, aren't you?
I went to eat some animal crackers and the box said, "Do not eat if seal is broken." I opened the box and sure enough..
As far as I can tell from this article and a few others that are derived from the same press releases, what VZ is doing here is setting up their own mail servers to use Port 587 submission instead of Port 25. That won't stop zombies or legitimate Linux mail systems from sending mail directly to their recipients' systems, though I'm guessing that they'll get around to blocking Port 25 (sigh) once they've got most of their users migrated to 587.
What this will do is give them authentication, which makes it easier for them to block customers who use VZ's mail servers from spamming, but I'd be surprised if there's much of that happening (though botnets keep evolving their techniques.) It's already possible to reduce that simply by using passwords, or using various hokey port 25 authentication methods like receive-before-send; this cleans up the process a bit.
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
Remove the head, destroy the brain.
Most ISPs already do a fair bit of policing on the users of their mail servers, so this probably won't make a big dent (though botnets keep evolving, and if the scalability works to use ISP mail servers, they'll go back to it.) This basically provides a cleaner, more standardized solution for mail submission and authentication. VZ might block Port 25 later, and getting their users onto 587 makes it easier.
Zombies already do deliver their mail directly using Port 25. They're not generally running Real Sendmail (which is way too big and heavy for what they need) - in general they're running stripped-down mail senders that don't bother checking error messages correctly, which is why greylisting's "Go away and come back in 5 minutes" is enough to discourage lots of them. But lots of ISPs have been jumping on the "Block Port 25" bandwagon (with no apologies to Linux users who run their own sendmail), so maybe the zombies will go back to using ISP mail servers more often.
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
This makes sense for 99.9+% of customers including probably 99.99% of non-business customers. Customers who claim to have a legitimate need for port 25 and who can demonstrate they have the technical and management infrastructure in place to prevent abuse and the liability insurance or proof of financial responsibility should they fail should be allowed to continue using it subject to termination at any time if it is abused. Heck, I might even just settle for proof of financial responsibility, if they had enough insurance to cover damages from the time spamming was discovered until the plug was pulled.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
Yeah, it's possible to do authentication on Port 25, but it's generally hokey and often broke things when people did it, and left passwords in the clear for eavesdroppers - 587 is a cleaner and more standardized solution. I remember having to configure Eudora for receive-before-send when my email provider was trying that approach...
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
Verizon has been an epic sewer network for years, and has ignored their spam problem for years. If they want to clean up now (or make a lame attempt to clean up, as most telco's do), fine. It just means less work for iptables at my end.
For those who are sick of Verizon's bullshit, here's my list (no promises this is complete, but it should have most of em) of Verizon's ip blocks.
206.46.0.0/16
66.12.0.0/14
207.68.0.0/17
71.96.0.0/11
72.64.0.0/11
72.42.0.0/18
71.160.0.0/15
71.162.0.0/16
96.224.0.0/11
98.108.0.0/14
98.112.0.0/13
68.160.0.0/14
162.84.0.0/16
162.83.0.0/16
151.204.0.0/15
138.88.0.0/21
66.171.0.0/16
66.14.128.0/17
151.201.0.0/16
138.89.0.0/16
141.149.0.0/16
141.150.0.0/15
141.152.0.0/14
141.156.0.0/15
141.158.0.0/16
68.160.192.0/18
68.161.192.0/18
66.14.0.0/17
151.196.0.0/14
151.200.0.0/14
151.204.0.0/15
129.44.0.0/16
138.88.0.0/16
64.222.0.0/15
68.236.0.0/14
70.104.0.0/13
70.16.0.0/13
71.96.0.0/11
209.158.0.0/16
209.159.0.0/19
71.160.0.0/11
173.64.0.0/12
70.192.0.0/11
66.174.0.0/16
75.224.0.0/12
75.240.0.0/13
75.192.0.0/10
97.0.0.0/10
Lawyers, MBA's, RIAA? A jedi fears not these things!
I wish that more software would default to 587 instead of 25. For example, Thunderbird doesn't even mention the possibility of 587 as a "default" port, which really needs to be changed.
In any case, it's good to see the change to 587 become more widespread and hopefully it will eventually become the default port for sending messages (along with encryption + authentication), while 25 will be reserved exclusively for server-to-server communication.
I like the suggestion that people are somehow lax in security because their mail client remembers their password. Who are these guys who type the password in every 3 minutes when they check their mail?
Everyone knows that damage is done to the soul by bad motion pictures. -Pope Pius XI
As more and more consumer ISP's block outbound connections on port 25, this will only accelerate the development of newer, smarter zombie bots that know how to read the configuration settings of popular email programs (perhaps even the passwords for popular webmail sites stored in your browser's saved password list) and use those settings to send mail.
This will be even more wonderful because all of that spam will now have your name and email address on it.
Tired of FB/Google censorship? Visit UNCENSORED!
I often seen antecdotal numbers in the "millions" when people talk about zombie infected boxen. Yet the article quotes Spamhaus.org claiming "225,454" machines on all networks are sending spam. Even if one were to assume that only a quarter of all zombie machines are sending spam at any one given time, that's still only a million boxes that are compromised and sending spam.
What's the deal? Are there really millions and millions of compromised Windows boxes out there in zombie networks? Or are the numbers over blown when matched up against activity logs that monitor traffic from compromised boxes?
LOL + (-1) stupidity
But lots of ISPs have been jumping on the "Block Port 25" bandwagon (with no apologies to Linux users who run their own sendmail), so maybe the zombies will go back to using ISP mail servers more often.
Many ISPs will let you use outbound port 25 if you request it. This usually means only responsible users will have the ability.
Also, you can configure sendmail to use port 587 on another server as the relay, so you could still use your own sendmail and relay through the ISP server.
What the fuck are they doing on 587? That's a secondary half-ass port used as a compromise and a low-end workaround for ISPs and network admins who blanket-block port 25. If you're to move away from port 25 (which can easily accept TLS for encrypted authentication or even just encrypted data without authentication), you might as well move to the one that requires both authentication and encryption.
NO responsible network or ISP should use plain-text authorization as the default method. I was astounded when I heard that RCN (et al!) fail to offer HTTPS webmail and POP3S email (if not the vastly superior IMAPS), and that TLS commands get dropped on the floor. This is completely unacceptable.
Verizon and co should not be commended for this trivial step, they should be scolded for not going full-on SSL.
Use my userscript to add story images to Slashdot. There's no going back.
It's already possible to reduce that simply by using passwords, or using various hokey port 25 authentication methods like receive-before-send; this cleans up the process a bit.
There is no requirement for any "hokey" authentication...port 25 for connections from inside an ISP could be routed (netcat, iptables, etc.) straight to where an MTA that allows relaying would be listening. For bonus points, any connection from inside the ISP to port 25 on any machine would end up at the same ISP "internal" MTA.
Meanwhile, connections to port 25 from outside the ISP would be routed to a "normal" MTA that doesn't require authentication and will not relay...it would only accept e-mail for domains local to "isp.com".
You don't even need authentication to make this work...authentication just gives you one more piece of proof where a connection came from.
..."if the victims let their mail clients remember their authentication credentials, at least the zombies will be easily identifiable."...
are they saying they don't keep track of who uses which IP? you gotta be kidding me.
what they're doing is the easy thing, block port 25 and with it the majority of the spambots. they just dont want the hassle of getting the trojans removed from those thousands of machines. too expensive.
Don't suggest that.
Transparent proxies are the work of the devil and a long step towards full-blown internet censorship.
Or do you work for a company that sells Great Firewalls to China?
Is a very good idea for reduce spam
I herd you like emails in your emails, so I put some traffic thru yo traffic.
What hokey port 25 authentication methods? Any authentication methods offered on port 587 can also be offered on port 25. There is nothing magical about "25" that makes strong authentication unpossible. There is nothing magical about "587" that makes it any more secure than "25." You can run a open relay just as easily on port 587 as you can run one on port 25. You can run SMTP-AUTH and TLS on port 25, and permit relaying to authenticated clients that use TLS, while non-authenticated and/or plain-text clients can only send mail destined for your own domains.
Setting aside port 587 for smtp-submit simply makes the firewall rules at the border easier to manage.
Edith Keeler Must Die
In my opinion, the transition to port 587 is nearly pointless. I already use authentication on port 25 to identify customers.
And according to one of the only people I'd trust on SMTP issues, "the SUBMIT specification has several fundamental flaws that make compliance practically impossible. I advise against all use of port 587" -- djb.
- Michael T. Babcock (Yes, I blog)
with no apologies to Linux users who run their own sendmail
None deserved, because probably 80% of them have some line in their sendmail.cf file like
>&;:$=M m QQ!2z ~dnl
that not only turns their server into an open relay, but sends every email 5 times, and on every third tuesday sends the entire contents of your harddrive as an attachment too.
For the love of God, if you don't have a clue as to what you're doing, don't do sendmail. Use exim, the installation script configures 80% of the sites out there in 1-4 questions, and for the remainder, the configuration is in a human-readable syntax that doesn't require learning a whole macro language just to configure the program that creates the configuration file for you.
(PROTIP: if you even think about asking a question about sendmail.cf, you are demonstrating that have no fucking clue what you're doing, and by continuing to use sendmail you deserve to have your computing license revoked until you have memorized the entire m4 documentation.)
It is useful because it allows ISPs to block port 25 for customers who do not run their own mail server (the vast majority of them). This makes it impossible for zombied machines to send mail directly , instead having to go through a relay. Open relays are much easier to filter against / get shutdown for abuse, than a whole swath of zombie computers. Mail going through authenticated relays is also easier to monitor for abuse, plus once the mailhosts relaying the authenticated mail are affected by zombie generated SPAM, they then have an incentive to do something about it.
In short it forces zombie SPAM to be channeled through choke points where it can be more easily identified and shutdown.
As for DJB, IIRC, his complaints against SUBMIT were entirely restricted to the fact that it will be yet another case where everyone implements defacto behavior, rather than following the standard to the letter, because the standard has some flaws in the way it is written. I agree that this is annoying for new implementers, as they have to look beyond the standard to "conventional wisdom" to figure out how to be interoperable. But this is true of every single network protocol in existence to varying degrees. I don't think he had any complaints about the idea of authenticated relays happening on a different port than mailhost-to-mailhost delivery. But, I can't find anything more detailed than what you posted so I can't say for sure.
If you want the ISP's MTAs to relay mail sent from internal computers, then this will break TLS over port 25 as the certificates will (by design) be invalid for the ISP's servers.
grr! Spamhaus is a sock puppet for industry forcing little guys running mail servers off the internet.
-- Programming with boost is like building a house with lego. It's a cool but I wouldn't want to live in it
I just reread your link. In it DJB explicitly advises against running authentication on port 25. In fact, for security reasons, he wrote two separate programs, qmail-smptd and ofmipd, to keep the tasks of relaying authenticated email and accepting mail for local delivery as removed from one another as possible.
He defends the idea of separating these two tasks, not only to separate ports but separate programs, on this thread on the IETF-SUBMIT mailing list.
So, yeah, his complaint against port 587 was simply that if you can't implement the SUBMIT standard correctly (which according to him noone can), you should use a different port then the one specified in that standard. The rest of the world doesn't care, because it sees all the various authentication methods (including SUBMIT) as extensions to SMTP, and not as a different protocol (OFMIP as DJB calls them collectively), and have no qualms running a standard (non-SUBMIT compliant) SMTP server on port 587.
"my boss uses Cox"
Cool. How do you like working for a female boss?
YAY port 587 is a great thing !
but are they going to sign their mail ?
now that would be a good thing so people can not FAKE a @Verizon.net address
google paypal yahoo etc do this
if Verizon did it people would start to respect @Verizon.net
simple if I get a Verizon.net address and it pass's the DKIM then I know it came from their domain
but a big WELL DONE ! someone with a clue got this done !
regards
John Jones
... are intriguing and I wish to subscribe to your newsletter.
music lover since 1969
Changing the port is like modulating your shields against the borg: it just delays the inevitable. There is a way to stop this, though. All that is required is some kind of registration service, say a web-accessible form, for the ISP's users to register their outgoing e-mail addresses with the mail server. It's no longer a problem then: the spambot suddenly needs to be able to spoof one of the user's valid outgoing e-mail addresses - and if it does, the ISP can now trace the offending account, and notify the account holder. If I'm missing something, someone please let me know. A lot of ISP's already do it this way.
verizon obviously has some equipment or customers behind their mailservers that do not support starttls. to avoid total breakage i would imagine they will include port forwarding on a few nets as well. moving the ports is...a bandage at best.
Good people go to bed earlier.
This article makes no sense!
Port 25 is traditionally the one used to send e-mail. ALL e-mail. It is not the one used "by large corporations", it is the one used by everyone. If some ISPs change the port, that's fine, but it still does not change the fact that port 25 is the known port for SMTP transactions.
A spammer can set up a "junk e-mail relay" on any port. The security has has nothing to do with the port.
WTF? Port 587 requires authentication? Port 25 does not?! The port is not some magical concept with special abilities. A port is just the communications channel used within the TCP/IP bandwidth. As an analogy, imagine if your cable company said "We had HBO in channel 20 and everybody was stealing it. We moved it to channel 42, because--as you know--channel 42 requires scrambling."
Switch ports. Wow! Why didn't anybody else think of that. That magical port 587, which is impervious to spam.
The ISP can set to require authentication on their SMTP server on any port. They could do this on port 25, though I'm sure that they would piss off some big clients in their network, if say, their e-mail stopped working one day. It's easier to push the cattle consumer to a different port and require authentication (arbitrarily, by them, not because of the port). After all, who cares if they complain; it's not as if Verizon answers its phones or offers proper customer service...
I'm sure that what Verizon is doing could be a Good Thing, but this article does not explain their reasoning properly. It makes it sound as if Verizon hit on a technological solution to the Spam problem, instead of saying what is really on their minds:
"We don't want to piss off our large clients by forcing everyone to authenticate and go through extra hoops to configure their system. So we'll offer a two tierred service: Large clients will continue using SMTP on port 25 as normal (inviting spam, as normal), and we'll force the rest of the users to use the SMTP ghetto on port 587. We don't even need to make the service on that port work well or reliably, since the big guys will still have the premium servers allocated on port 25."
-dZ.
Carol vs. Ghost
Verizon pledges to clean up their zombie problem quickly.
That's what they said abot Ravenholm, and see what happened!
Maybe they can add SSL to their webmail too
Now that's a response that'll shut 'em up. Right comes not only from correct analysis, but also requires a refusal to live in fear! Nice job.
It's only been about five or six years since I wrote a letter to a Verizon executive about email I was receiving from Verizon zombies. I was frustrated by no way of contacting them online and looked up the executive's postal mailing address. I got no response.
Their track record continues. I looked for a way to find out if they will be blocking TCP/25 connections to other ISPs or just to their mail servers, and there seems to be no way to contact a live human being at that company.
"Who controls the past controls the future. Who controls the present controls the past." -- George Orwell
Is this spam problem being taken advantage of by the new world order as a reason for identifying oneself before sending email?
Big brother is making moves.
Don't worry, modern deep inspection can do almost anything that a transparent proxy can do, and it's generally harder to detect.
Finally! A year of moderation! Ready for 2019?
DKIM is helpful in some cases but not too many.
The real solution to spam is individual sender signatures, because:
* A mail server (ISP or IT or self owned) can never accurately decide what is SPAM for the recipient.
* Signed e-mail allows the recipient to filter accordingly
* Unknown senders can be assigned a trust score based on the network of trust and filtered accordingly
* Keys can be bought form commercial vendors, but they don't have to
* Mail lists can re-sign a message, so no forwarding problems there, just a bit of computation.
Do you sign you e-mail? Start today and make the world better. Once the signature is universal, even the ISP get rid of the 80% + useless SPAM, because it will be not profitable anymore. If the ISPs want to do something about it, give signature keys to your customers or sign the e-mail automatically with the customers key (by default).
Busy helping non technical users of OpenOffice.org - http://plan-b-for-openoffice.org/