I am trying to configure smartd to send mails via s-nail on Fedora 41. I created a .mailrc file (in which I have set the mta variable to directly send via smtps, there is no sendmail installed) in roots home directory and can successfully send mails via: echo "Test" | mail -s Test I also managed to send mails in a bash script started by a custom systemd service. But smartd isn't able to send mails. The following error is shown in the log: Executing test of /usr/libexec/smartmontools/smartdnotify to ... Test of /usr/libexec/smartmontools/smartdnotify to produced unexpected output (163 bytes) to STDOUT/STDERR: s-nail: Cannot start /usr/sbin/sendmail: executable not found (adjust *mta* variable) s-nail: Cannot save to $DEAD: Permission denied s-nail: ... message not sent Selinux is blocking the access to the .mailrc file (therefore s-nail is trying /usr/sbin/sendmail as a default fallback): type=AVC msg=audit(1744370186.375:606): avc: denied { read } for pid=42644 comm="mail" name=".mailrc" dev="nvme0n1p3" ino=140324 scontext=system_u:system_r:smartdwarn_t:s0 tcontext=unconfined_u:object_r:mail_home_t:s0 tclass=file permissive=0 I tried the suggested ausearch -c 'mail' --raw | audit2allow -M my-mail semodule -X 300 -i my-mail.pp systemctl restart smartd.service a couple of times until no new selinux errors appeared. Now I get the following error: Test of /usr/libexec/smartmontools/smartdnotify to produced unexpected output (130 bytes) to STDOUT/STDERR: s-nail: could not initiate TLS connection: error:00000000:lib(0)::reason(0) /root/dead.letter 23/578 s-nail: ... message not sent s-nail now can access the .mailrc file and can connect to the server. But no successfull communication with the server (Error 0 ?). The content of the mail is written to the dead.letter file instead. What could be the reason for this? Is it an improper selinux config? Am I missing an selinux option? Do I have to switch mta client?

I want to run an email server for myself (I'm leaning towards postfix), but I need it to look something like this: [MUA] [Home MTA] [External MTA] The Home MTA sits on a virtualization server at my home, and only accepts mail from the external MTA (through its config but also firewall). The external MTA only accepts email for my domain, has DKIM/SFP, and (initially, anyway) isn't going to be delivering email to anyone except my home MTA. I think if I do this right, that should prevent me from being blacklisted and also from doxxing my home IP as that won't be the MTA listed using my domains’s MX records. I think of this almost like a bastion or jump server, but for mail instead of ssh. 1. is this possible to do? 2. what else should I be thinking about? e.g. I found a page talking about preventing backscatter email, which hadn't occurred to me. 3. I'm assuming the MTA-to-MTA can be encrypted and authenticated. Is that SASL? 4. Is there a term for this type of setup so I can find a basic config and go from there? E.g. is this called smart-relay with a satellite or something? EDIT: I think I would want to disable SMTP in favour of SMTPS, and configure postfix not to accept outbound email from an unauthenticated user. That prevents spammers from using my MTA to get mail out, right? If that’s the case why is it still generally recommended not to run your own mail server due to having to deal with all the spam?

I have been using sendmail to send out mails using internally available mail server. But currently port 25 is blocked for security reasons. I would like to know if there is a way to specify port number in the sendmail utility. I am trying to make use of the secure SMTP-MSA port 587 as an alternative assuming I could get that port opened up. I was not able to find anything in the man pages for sendmail. Is there any alternate utility that could do this?

Sendmail not following smart_host alternative port macro. Using Exchange server, exchange.redacted.com, we are using port 2525 for internal devices and some Linux based machines, allowing them send outgoing email without authentication. The 'receive connector' allows relay for devices on our intranet and external with explicit IP address authorization. Linux based devices MTA is **sendmail**. I edited /etc/mail/sendmail.mc adding: define(SMART_HOST',relay:[exchange.redacted.com]')dnl define(RELAY_MAILER_ARGS', TCP $h 2525')dnl define(ESMTP_MAILER_ARGS', TCP $h 2525')dnl define(SMTP_MAILER_ARGS' , TCP $h 2525')dnl define(SMTP8_MAILER_ARGS', TCP $h 2525')dnl define(DSMTP_MAILER_ARGS', TCP $h 2525')dnl Just in case, I added all the __*_MAILER_ARGS__ trying to specify the port to use with no success. The generated **sendmail.cf**, shows the __SMART_HOST__ selected but not the alternative port. # "Smart" relay host (may be null) DSrelay:[exchange.redacted.com] ... Mrelay, P=[IPC], F=mDFMuXa8, S=EnvFromSMTP/HdrFromSMTP, R=MasqSMTP/MasqRelay, E=\r\n, L=2040, T=DNS/RFC822/SMTP, A=TCP $h modifying the Mrelay, adding 2525 works as intended: Mrelay, P=[IPC], F=mDFMuXa8, S=EnvFromSMTP/HdrFromSMTP, R=MasqSMTP/MasqRelay, E=\r\n, L=2040, T=DNS/RFC822/SMTP, A=TCP $h 2525 How can I specify the port on the **.mc** source without the need of a manual edit the resultant **.cf**?

On my laptop, I have set up a cron that performs a weekly job. If there is something to report, I would like to get an e-mail for it. Right now I am using KMail as MUA which is able to read mail from the Maildir directory at ~/.local/share/local-mail. Thus, I want local mail to be delivered to that folder (with a Maildir structure). I was hoping that a dead simple program already exist that has a sendmail interface (such that echo "$REPORT" | mail -s "$SUBJECT" "$ME" can work with it). Installing exim or forwarding mail to my remote mailserver is considered overkill. The question https://unix.stackexchange.com/q/5771/8250 seems to target remote forwarding which does not suit my needs (I need local delivery). This old Gentoo thread ended up in crafting an old script in Perl. Surely there must exist a well-thought, dead-simple program? Any recommendations? I am using Arch Linux.

How can I configure Postfix to silently drop/discard/reject mails that were sent from specific users to an external address? It should work only for specific users E.g: 1. abc@localdomain can send mail to only @localdomain users. 2. abc@localdomain can only receive from @localdomain users.

Take postfix for example. It's an MTA, but is it also an SMTP server? Are all MTAs SMTP servers? Are any? Is there a difference between a mail transport agent and a mail transfer agent? Can an MTA be configured to be an SMTP server, or vice versa? What's the difference between an MTA and an SMTP server?

I am using /etc/aliases to direct group emails to multiple accounts, for example: postmaster: john abuse: john www: paul team: john, paul, ringo, george One of the lines is getting quite long due to the number of users it forwards to. Is it possible to wrap the line? I am using Exim as the MTA.

I am trying to send all local mail to a single external e-mail, using Google's SMTP servers. Using Google's SMTP servers work, as I am able to send mail to external addresses work using mail from mailutils. The redirection of local mail to the external e-mail does not work. When sending a mail to root from my account test using: echo "Body" | mail -s "Test Postfix To Root" root * I never recieve the message * Nothing new turns up in /var/log/mail.err * This turns up in /var/log/mail.log

Sep  4 18:48:06 desktop1204test postfix/pickup: C9326EE26: uid=1000 from=
Sep  4 18:48:06 desktop1204test postfix/cleanup: C9326EE26: message-id=
Sep  4 18:48:06 desktop1204test postfix/qmgr: C9326EE26: from=, size=401, nrcpt=1 (queue active)
Sep  4 18:48:06 desktop1204test postfix/local: C9326EE26: to=, orig_to=, relay=local, delay=0.03, delays=0.02/0/0/0, dsn=2.0.0, status=sent (delivered to mailbox)
Sep  4 18:48:06 desktop1204test postfix/qmgr: C9326EE26: removed

My /etc/postfix/main.cf:

inet_interfaces = loopback-only
mynetworks = loopback-only
alias_maps = hash:/etc/aliases
alias_database = hash:/etc/aliases
virtual_alias_domains = localhost.localdomain
virtual_alias_maps = hash:/etc/postfix/virtual
relayhost = [smtp.gmail.com]:587
smtp_sasl_auth_enable = yes
smtp_sasl_password_maps = hash:/etc/postfix/sasl_passwd
smtp_sasl_security_options = noanonymous
smtp_tls_CAfile = /etc/postfix/cacert.pem
smtp_use_tls = yes

My etc/aliases: # See man 5 aliases for format postmaster: root My /etc/postfix/virtual: @localhost.localdomain someone@somewhere.com How can I get it working? I don't want any local mail to arrive locally, they should all go to someone@somewhere.com. -------------- I tried the solution suggested in a comment below: put *: someone@somewhere.com in /etc/aliases. But did not work: test@desktop1204test:~$ sudo newaliases test@desktop1204test:~$ cat /etc/aliases # See man 5 aliases for format postmaster: root *: someone@somewhere.com test@desktop1204test:~$ sudo newaliases test@desktop1204test:~$ sudo service postfix restart * Stopping Postfix Mail Transport Agent postfix [ OK ] * Starting Postfix Mail Transport Agent postfix [ OK ] test@desktop1204test:~$ echo "Body" | mail -s "Test Postfix To Root $(date)" root test@desktop1204test:~$ tail /var/log/mail.err test@desktop1204test:~$ tail /var/log/mail.log ...... Sep 4 22:46:12 desktop1204test postfix/master: daemon started -- version 2.9.6, configuration /etc/postfix Sep 4 22:46:23 desktop1204test postfix/pickup: 859AFF6A8: uid=1000 from= Sep 4 22:46:23 desktop1204test postfix/cleanup: 859AFF6A8: message-id= Sep 4 22:46:23 desktop1204test postfix/qmgr: 859AFF6A8: from=, size=431, nrcpt=1 (queue active) Sep 4 22:46:23 desktop1204test postfix/local: 859AFF6A8: to=, orig_to=, relay=local, delay=0.02, delays=0.02/0/0/0, dsn=2.0.0, status=sent (delivered to mailbox) Sep 4 22:46:23 desktop1204test postfix/qmgr: 859AFF6A8: removed test@desktop1204test:~$

How does cron know what MTA to use? Where is this info stored please? (Using Ubuntu KK.) Edit: Here's the whole story: I enabled the MAILTO option in crontab, and scheduled some task. When the task got triggered, the system complained that there is no MTA installed, so I installed ssmtp. Interestingly, I never installed sendmail on my system, nor do I think it is installed by default. Hence all I have is ssmtp. Now, I need to make sure that whatever application does the emailing is bound to some specific NIC, hence it needs to be launched prefixed by firejail. The question is now, how do I know what email application is launched by cron, and how can I launch it with the firejail prefix.

Does sendmail support TLS_Clt access matching based on blocks of IPs? So far the official sendmail TLS documentation I have found does not indicate that it does (nor that it doesn't).

Does anybody on this planet know what the equivalent GNU mailutils command to the following looks like? Is this even possible without a local MTA with mailutils? Using s-nail it works as intended. ```bash s-nail -s "$subject" -S $smtp_server -r $from_user $to mail: Cannot open mailer: Input/output error \ > mail: cannot send message: Input/output error this is not a duplicate of [can-i-set-up-system-mail-to-use-an-external-smtp-server](https://unix.stackexchange.com/questions/36982/can-i-set-up-system-mail-to-use-an-external-smtp-server) as they used a local MTA or heirloom-mailx or s-nail or other tools. I am using Debian.

I have already read [Why use Dovecot AND Postfix/Sendmail?](https://serverfault.com/questions/626803/why-use-dovecot-and-postfix-sendmail) , [Why use both Dovecot and Postfix?](https://serverfault.com/questions/716653/why-use-both-dovecot-and-postfix) , [Why use both Dovecot and Postfix?](https://superuser.com/questions/962074/why-use-both-dovecot-and-postfix) which give the general idea, but here I'm looking for more precise details to understand how email servers connect together. Let's say I own the domain example.com with DNS configured like this: example.com MX mail.example.com mail.example.com A 1.2.3.4 and postfix and dovecot are intalled on my Debian server of IP 1.2.3.4. Let's say I send an email: * from my ISP address johndoe@myisp.com, from my local computer Thunderbird * ... to the address johndoe@example.com **What is the big picture of the different connections that will happen during this process?** Is the following correct: 1. My local computer connects to smtp.myisp.com (port 25 or 587) 2. My ISP's SMTP server (postfix?) queries the DNS of example.com and eventually will connect to mail.example.com, i.e. my own server where postfix is listening on port 25 or 587 3. My server's postfix saves the email on disk (where, by default?) 4. or postfix sends the email to dovecot (how do they connect together?) 5. When eventually I want to fetch mail on johndoe@example.com with another Thunderbird account, the computer will connect my server 1.2.3.4 on port 110 or 995 (POP3). ? In this case, postfix is the MTA, dovecot is the POP3/IMAP server, but what is the MDA/LDA (Delivery Agent)?

I recently moved postfix into a docker container, with ports 25 and 587 exposed on the host machine. Since doing this, cron on the host machine is now unable to send emails due to No MTA installed, discarding output. It appears that cron tries to use the sendmail command, which is normally provided by an MTA package. Installing postfix on the host to provide sendmail seems a bit silly, and it will try to listen on port 25 and conflict with the mappings from the dockerised postfix. I tried installing msmtp and adding a /etc/msmtprc:

account default
host localhost
auto_from on
maildomain [domain]

Cron now tries to send mail through postfix, but even with the auto_from and maildomain options it seems to use a bare or invalid from header like From: root or From: root (Cron Daemon), which gets rejected. opendkim can't handle this either. I'm assuming that when postfix was installed on the host, the postfix provided sendmail sorted out all of the addresses and appended the mailname as needed, but that the minimalist msmtp isn't doing this? What is the solution here? How can I get a sensible sendmail command on the host that will use the dockerised postfix to send the mail?

What I try to achieve is to send all my server logs, errors, crontab outputs etc. with something like telegram or signal. I am locking for a MTA with can handle a different backend and is still compatible with the overall Linux MTA concept. So **without** fiddling with scripts and crontabs I would like all output, which would be transfered by mail to be transfered by a custom backend. Is there a way to change the MTA backend? What I found so far is ntfy , but it's just a script to set different notification backends, I think there is a missing puzzle piece between the script and a proper MTA. Thanks for the input

I am a graduate student and a relative Linux novice. This question is a sort of spin-off of my earlier question . My institution has a Ubuntu Linux cluster. I am just a user; I do not have sysadmin permissions, and I certainly do not have the expertise to be a sysadmin! My question is, how do I find my local mail spool? As far as I know, I am not using a mail transfer agent. Answers to my previous question suggested that I look for a file /var/spool/mail/$USER, but unfortunately, I do not see a file corresponding to my user name. In /var/spool/mail/, using ls I only see two files: nobody and www-data, which are both extensionless files. Do you have any other ideas of where I can look for my mail spool (which is probably local, since I do not have a mail transfer agent configured, as far as I know)?

Reading https://ccm.net/contents/116-how-email-works-mta-mda-mua and https://en.wikipedia.org/wiki/Email_agent_(infrastructure) , I wonder why the email sending process is not symmetric with the receiving process? The sending process has no something similar to MDA, and MUA contacts MTA directly for sending emails. (Imagine when mailing a letter, we go to the nearest public mailbox on streets, and mailmen collect mails from them twice a day.) - How does MUA know what emails have been sent? (How does MUA get the emails in the "Sent" folder?) - Are sent emails stored in the same place as received mails in MDA in a server?

I have looked at the definitions of MTA, MUA, MDA, for example, https://ccm.net/contents/116-how-email-works-mta-mda-mua and https://en.wikipedia.org/wiki/Email_agent_(infrastructure) - In Lubuntu, the default email client is Sylpheed, and many also use Thunderbird. If I am correct, both Sylpheed and Thunderbird are MUA. - In https://unix.stackexchange.com/a/479613/674 , Stephen mentioned that cron uses MTA (such as postfix or sendmail) to send the outputs of its jobs as emails. My questions are: - Why does cron use MTA instead of MUA to send emails? If cron can use MUA to send emails, how? - why do we use MUA (Sylpheed or Thunderbird) instead of MTA to send emails? If we can use MTA to send emails, how? - When we install MUA (Sylpheed or Thunderbird) , does Sylpheed or Thunderbird need a MTA installed on the **same** machine to send emails? Thanks.

I setup an SMTP server(Postfix), Then I send an email to my gmail, I get the following email from gmail: This is the mail system at host mail.pahlevanzadeh.net. I'm sorry to have to inform you that your message could not be delivered to one or more recipients. It's attached below. For further assistance, please send mail to postmaster. If you do so, please include this problem report. You can delete your own text from the attached returned message. The mail system : host gmail-smtp-in.l.google.com[142.250.31.27] said: 550-5.7.26 This message does not have authentication information or fails to 550-5.7.26 pass authentication checks. To best protect our users from spam, the 550-5.7.26 message has been blocked. Please visit 550-5.7.26 https://support.google.com/mail/answer/81126#authentication for more 550 5.7.26 information. m13si7030284qvk.215 - gsmtp (in reply to end of DATA command) Reporting-MTA: dns; mail.pahlevanzadeh.net X-Postfix-Queue-ID: 96B203C01C0 X-Postfix-Sender: rfc822; mohsen@pahlevanzadeh.net Arrival-Date: Sat, 6 Jun 2020 11:37:33 -0400 (AST) Final-Recipient: rfc822; m.pahlevanzadeh@gmail.com Original-Recipient: rfc822;m.pahlevanzadeh@gmail.com Action: failed Status: 5.7.26 Remote-MTA: dns; gmail-smtp-in.l.google.com Diagnostic-Code: smtp; 550-5.7.26 This message does not have authentication information or fails to 550-5.7.26 pass authentication checks. To best protect our users from spam, the 550-5.7.26 message has been blocked. Please visit 550-5.7.26 https://support.google.com/mail/answer/81126#authentication for more 550 5.7.26 information. m13si7030284qvk.215 - gsmtp salamkdsdjfi.eml Subject: salamkdsdjfi From: Mohsen Pahlevanzadeh Date: 6/6/20, 8:06 PM To: Pahlevanzadeh dwjkdjkdjvkdlvnfklvnfklvf Where's my problem?

I have some problems choosing and configuring an MTA for my desktop system. Some tools like apt-listchanges or cron rely on local mail delivery and I think having local mail delivery available is a nice thing. On the other hand, there are tools like [caff](http://pgp-tools.alioth.debian.org/) , which work best with an MTA which can deliver to remote domains. Like most people, I don't handle my mail on my personal desktop machine but on a remote mail server and would like to send my outgoing mail through that host, as sending mail from a machine without a fixed IP and hostname doesn't work well for all recipients. Up to now this can be easily done with [dma](https://github.com/corecode/dma) and probably most other MTAs, but it gets more complicated once I let some of my friends use my computer and create an account for each of them. If I would just leave my regular smarthost + local delivery configuration in place, every user on the machine would have to share the same smarthost with the same authentication. Not nice! A much nicer solution would be to let every system user specify their own smarthost and authentication and. msmtp does exactly that. Unfortunately it doesn't do local mail delivery, so messages from cron or apt-listchanges are just dropped. So this is not a nice option either. I guess the heavier MTAs like Exim4 or Postfix can be configured to do that, but I really don't know how.