Postfix or Dovecot stops after subscription IP address is changed: Failed to start Postfix Mail Transport Agent

Symptoms

When the IP address for a subscription with a large number of domains is changed, Postfix or Dovecot service stops:

# systemctl status postfix.service
● postfix.service – Postfix Mail Transport Agent
Loaded: loaded (/usr/lib/systemd/system/postfix.service; enabled; vendor preset: disabled)
Active: failed (Result: start-limit) since Tue 2018-12-04 16:47:01 CET; 6s ago
Process: 95048 ExecStop=/usr/sbin/postfix stop (code=exited, status=0/SUCCESS)
Process: 94905 ExecStart=/usr/sbin/postfix start (code=exited, status=0/SUCCESS)
Process: 94902 ExecStartPre=/usr/libexec/postfix/chroot-update (code=exited, status=0/SUCCESS)
Process: 94900 ExecStartPre=/usr/libexec/postfix/aliasesdb (code=exited, status=0/SUCCESS)
Main PID: 94977 (code=killed, signal=TERM)
plesktest-plesk001 systemd[1]: postfix.service failed.
plesktest-plesk001 systemd[1]: start request repeated too quickly for postfix.service
plesktest-plesk001 systemd[1]: Failed to start Postfix Mail Transport Agent.

# systemctl status dovecot.service
…
failed to start dovecot imap/pop3 mail server
…

Cause

This is a Plesk bug with ID #PPPM-9733, which is planned to be fixed in one of the future product updates.

Resolution

As a workaround, perform the following steps:

1. Connect to the server via SSH;

2. Create the following directory:

   For Postfix:

   # mkdir /etc/systemd/system/postfix.service.d/

   For Dovecot:

   # mkdir /etc/systemd/system/dovecot.service.d/

3. Create the configuration file in the directory from step 1:

   # touch 10-start_limit_interval.conf

4. Edit the configuration file 10-start_limit_interval.conf adding the following strings:

   [Service]
   StartLimitInterval=0

5. Reload unit files and restart Postfix/Dovecot:

   # systemctl daemon-reload
   # systemctl restart postfix.service
   # systemctl restart dovecot.service