I run a personal email server. I can't emphasize enough how geofencing has reduced the useless hacking on my email server. I only leave port 25 open to the world. I use port 587. I maintain a list of hosting companies that I block from using my web server since they are just going to scrape anyway. I also keep that IP space off of my email other than port 25. Firewalls use memory but tend to be very light on the CPU other than when you first start up the firewall. I assume they take the deny list and create a table in RAM to efficiently block IPs. I have found that dynamic IP blocking programs such as sshguard or fail2ban are a CPU burden since that table needs to be refreshed as new IPs are added or removed so I have stopped using them. Not that the programs themselves are CPU intensive, but they cause the firewall to be CPU intensive. I am considering using sshguard again but with a very high threshold to add an IP to the deny list. Regarding attempts to add 2FA by using RoundCube or similar web based email, I think those programs just increase the attack surface. When I used a hosting service I was hacked by an unpatched exploit in RoundCube. ? Original Message ? From: sebastian at sebbe.eu Sent: July 15, 2021 3:55 AM To: dovecot at dovecot.org Reply-to: dovecot at dovecot.org Subject: Sv: function for whitelisting IPs Most such functions would need to be custom. You need to write a custom login script, which also accepts the user's IP as input to a function, which then checks if password is right. And then it returns that password is invalid if IP isn't approved. Then you just need to write some custom functions in roundcube or similiar to have the webmail insert the IP into a database. Or just match it against a GeoIP database and save the latest country the webmail was logged in from, and then SMTP/IMAP is only approved for that country. That reduces the attack surface greatly. -----Ursprungligt meddelande----- Fr?n: dovecot-bounces at dovecot.org <dovecot-bounces at dovecot.org> F?r White, Daniel E. (GSFC-770.0)[NICS] Skickat: den 15 juli 2021 12:21 Till: Dovecot Mailing List <dovecot at dovecot.org> ?mne: function for whitelisting IPs Sebastian, Do you have any examples of such a function and how/where it is used ? ?-----Original Message----- From: dovecot <dovecot-bounces at dovecot.org> on behalf of Sebastian <sebastian at sebbe.eu> Reply-To: Dovecot Mailing List <dovecot at dovecot.org> Date: Thursday, July 15, 2021 at 01:19 To: 'Mailing List' <dovecot at dovecot.org> Subject: [EXTERNAL] Sv: 2FA/MFA with IMAP & postfix/submission ??? Main problem is that not many clients do natively support multifactor. ??? Some clients, do popup a login dialog if the server rejects the password as invalid, which can be used to create a "cheaty variant" of multifactor, but some clients just popup an error dialog and tell the user to just correct password in settings. ??? Some clients even go as long as requiring the user to delete the account with wrong password and set up a new connection. ??? So no, it cannot be relied upon. ??? I have a better idea: ??? Have a function for whitelisting IPs, possible /24's or similiar, where a login to roundcube or other webmail client (with 2FA) will add the IP onto a whitelist for that account. ??? Or perhaps, just "set" the country of the account based on GeoIP. ??? When an account tries to login via IMAP or SMTP, you just check if IP and/or GeoIP country is right, and reject the login as invalid if so not. ??? The only thing a client needs to do to get his IMAP or SMTP client to work again if it stops working, is to login once via the web client. ??? -----Ursprungligt meddelande----- ??? Fr?n: dovecot-bounces at dovecot.org <dovecot-bounces at dovecot.org> F?r Alex ??? Skickat: den 15 juli 2021 02:10 ??? Till: dovecot at dovecot.org ??? ?mne: 2FA/MFA with IMAP & postfix/submission ??? Hi, I have a dovecot-2.3.13 system on fedora34 with a few hundred ??? IMAP4 accounts, as well as postfix users using submission. Clients are ??? using primarily Outlook on Windows and old squirrelmail. ??? Are there multi-factor options available? ??? If it is not available, do you have any recommendations on where I ??? should look to do this? ??? All of the links related to this topic appear to be very old, or ??? limited to Linux PAM users.
> I have found that dynamic IP blocking programs such as sshguard or > fail2ban > are a CPU burden since that table needs to be refreshed as new IPs are > added > or removed so I have stopped using them.Have you seen ipset? https://ipset.netfilter.org/ It is built for dynamically adding/remove IP's from a firewall without changing a table or rules or reloading the firewall. It holds a hashmap in memory of what IP's to block and integrates into the kernel. However you have to build your own mouse trap to use it. I don't know of anything out of the box that would automatically add IP's to it, i wrote my own script that gets fed log lines from rsyslog to do it.
Yeah the idea was to use roundcube or other web service to add kind of
"auth service" or "unlock service" where you can auth with
2FA to move the geofence or permit additional IPs in geofence. For example, if
you are travelling or otherwise need to enable your account for a "outsider
IP".
This could be a simple webpage asking for username and 2FA code, and all it does
it adds the IP to auth list. But could be a full roundcube or other webmail
solution too, to give more usefullness to the web login solution if you
don't have a imap/smtp client for now.
I don't use 587 myself, but instead, I have set so auth is only permitted on
port 25 for authorized IPs (auth_advertise_hosts in exim), thus the server will
refuse to allow outsiders to authenticate.
In combination with some other policies, my server is practically rock solid.
-----Ursprungligt meddelande-----
Fr?n: dovecot-bounces at dovecot.org <dovecot-bounces at dovecot.org> F?r
lists
Skickat: den 15 juli 2021 20:09
Till: 'Mailing List' <dovecot at dovecot.org>
?mne: Re: Sv: function for whitelisting IPs
I run a personal email server. I can't emphasize enough how geofencing has
reduced the useless hacking on my email server. I only leave port 25 open to the
world. I use port 587.
I maintain a list of hosting companies that I block from using my web server
since they are just going to scrape anyway. I also keep that IP space off of my
email other than port 25.
Firewalls use memory but tend to be very light on the CPU other than when you
first start up the firewall. I assume they take the deny list and create a table
in RAM to efficiently block IPs. I have found that dynamic IP blocking programs
such as sshguard or fail2ban are a CPU burden since that table needs to be
refreshed as new IPs are added or removed so I have stopped using them. Not that
the programs themselves are CPU intensive, but they cause the firewall to be CPU
intensive. I am considering using sshguard again but with a very high threshold
to add an IP to the deny list.
Regarding attempts to add 2FA by using RoundCube or similar web based email, I
think those programs just increase the attack surface. When I used a hosting
service I was hacked by an unpatched exploit in RoundCube.
Original Message
From: sebastian at sebbe.eu
Sent: July 15, 2021 3:55 AM
To: dovecot at dovecot.org
Reply-to: dovecot at dovecot.org
Subject: Sv: function for whitelisting IPs
Most such functions would need to be custom.
You need to write a custom login script, which also accepts the user's IP as
input to a function, which then checks if password is right.
And then it returns that password is invalid if IP isn't approved.
Then you just need to write some custom functions in roundcube or similiar to
have the webmail insert the IP into a database.
Or just match it against a GeoIP database and save the latest country the
webmail was logged in from, and then SMTP/IMAP is only approved for that
country.
That reduces the attack surface greatly.
-----Ursprungligt meddelande-----
Fr?n: dovecot-bounces at dovecot.org <dovecot-bounces at dovecot.org> F?r
White, Daniel E. (GSFC-770.0)[NICS]
Skickat: den 15 juli 2021 12:21
Till: Dovecot Mailing List <dovecot at dovecot.org>
?mne: function for whitelisting IPs
Sebastian,
Do you have any examples of such a function and how/where it is used ?
?-----Original Message-----
From: dovecot <dovecot-bounces at dovecot.org> on behalf of Sebastian
<sebastian at sebbe.eu>
Reply-To: Dovecot Mailing List <dovecot at dovecot.org>
Date: Thursday, July 15, 2021 at 01:19
To: 'Mailing List' <dovecot at dovecot.org>
Subject: [EXTERNAL] Sv: 2FA/MFA with IMAP & postfix/submission
Main problem is that not many clients do natively support multifactor.
Some clients, do popup a login dialog if the server rejects the password as
invalid, which can be used to create a "cheaty variant" of
multifactor, but some clients just popup an error dialog and tell the user to
just correct password in settings.
Some clients even go as long as requiring the user to delete the account
with wrong password and set up a new connection.
So no, it cannot be relied upon.
I have a better idea:
Have a function for whitelisting IPs, possible /24's or similiar, where
a login to roundcube or other webmail client (with 2FA) will add the IP onto a
whitelist for that account.
Or perhaps, just "set" the country of the account based on GeoIP.
When an account tries to login via IMAP or SMTP, you just check if IP and/or
GeoIP country is right, and reject the login as invalid if so not.
The only thing a client needs to do to get his IMAP or SMTP client to work
again if it stops working, is to login once via the web client.
-----Ursprungligt meddelande-----
Fr?n: dovecot-bounces at dovecot.org <dovecot-bounces at dovecot.org>
F?r Alex
Skickat: den 15 juli 2021 02:10
Till: dovecot at dovecot.org
?mne: 2FA/MFA with IMAP & postfix/submission
Hi, I have a dovecot-2.3.13 system on fedora34 with a few hundred
IMAP4 accounts, as well as postfix users using submission. Clients are
using primarily Outlook on Windows and old squirrelmail.
Are there multi-factor options available?
If it is not available, do you have any recommendations on where I
should look to do this?
All of the links related to this topic appear to be very old, or
limited to Linux PAM users.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 5715 bytes
Desc: S/MIME Cryptographic Signature
URL:
<https://dovecot.org/pipermail/dovecot/attachments/20210715/b7536388/attachment-0001.p7s>
> I run a personal email server. I can't emphasize enough how geofencing has reduced the useless hacking on my email server. I only leave port 25 open to the world. I use port 587.Unfortunately that's not an option for commercial mailservers. You have to be open to communicate with the world. Geofencing might be inaccurate. Often this data is extracted from ip-net registrations - the country where the company resides that registered that net might not be where the servers are located. There are services like maxmind that are more accurate but are not free.> Firewalls use memory but tend to be very light on the CPU other than when you first start up the firewall. I assume they take the deny list and create a table in RAM to efficiently block IPs. I have found thatThis depends on how your firewall works. A standard linux firewall processes iptables rules one after another. With a lot of rules and high traffic this can cause very high cpu usage. In case you're using ipsets (like a hashmap) that is not the case. There's also a difference if you block single ips or whole subnets.> dynamic IP blocking programs such as sshguard or fail2ban are a CPU burden since that table needs to be refreshed as new IPs are added or removed so I have stopped using them. Not that the programs themselves are CPU intensive, but they cause the firewall to be CPU intensive. I am considering using sshguard again but with a very high threshold to add an IP to the deny list.It's not that cpu intensive when using ipsets. On the other hand fail2ban itself uses quite some cpu and memory (sqlite databases can get large). I haven't been using fail2ban because of that, so I don't know if the situation has improved.> Regarding attempts to add 2FA by using RoundCube or similar web based email, I think those programs just increase the attack surface. When I used a hosting service I was hacked by an unpatched exploit in RoundCube.Programs like fail2ban do not increase the attack surface under normal circumstances. They just scan logs and add firewall rules, which does not cost very much when using ipsets. I'm very interested which roundcube bug that was, using roundcube myself. Can you have a look at the cve list, please: https://www.cvedetails.com/vulnerability-list/vendor_id-8905/Roundcube.html Best regards Gerald