Fix the configuration template and the restart script.

README.md

@@ -28,12 +28,14 @@ collection is empty; lists with one or more existing rules are left untouched.
 ## Verified weekly restart
 
 The optional weekly restart performs separate stop and start operations. It
-waits for both systemd and `mailman status` to confirm that Mailman has stopped
+waits for systemd, `mailman status`, and the configured LMTP listener to confirm
-before starting it, then checks both signals again to confirm startup:
+that Mailman has stopped before starting it. It then requires the master and
+the LMTP port to become available before considering startup successful:
 
 ```yaml
 mailman_enable_weekly_verified_restart: true
 mailman_weekly_verified_restart_on_calendar: 'Sun *-*-* 04:00:00'
+mailman_service_stop_timeout: 30
 mailman_weekly_verified_restart_stop_timeout: 120
 mailman_weekly_verified_restart_start_timeout: 120
 ```
@@ -41,3 +43,11 @@ mailman_weekly_verified_restart_start_timeout: 120
 The timer is deliberately not persistent, so a missed run is not executed
 immediately after a server boot. Failures are recorded by systemd in the
 `mailman-verified-restart.service` journal.
+
+The managed `mailman.service` uses `KillMode=control-group` and an explicit
+`TimeoutStopSec`. If Mailman's own stop command hangs, systemd terminates the
+whole service cgroup and removes a stale master PID file. The weekly job still
+verifies the real stopped state and does not start a second instance until the
+master and LMTP listener are gone.
+The normal role execution also waits for the configured LMTP port and fails if
+the listener does not become available after starting or restarting Mailman.

defaults/main.yml

@@ -40,6 +40,7 @@ mailman_enable_daily_notifications: true
 # default and, when enabled, uses the managed host's local timezone.
 mailman_enable_weekly_verified_restart: false
 mailman_weekly_verified_restart_on_calendar: 'Sun *-*-* 04:00:00'
+mailman_service_stop_timeout: 30
 mailman_weekly_verified_restart_stop_timeout: 120
 mailman_weekly_verified_restart_start_timeout: 120
 

tasks/mailman.yml

@@ -103,6 +103,15 @@
       service: name=mailman state=restarted
       when: mailman_hyperkitty_install is changed
 
+    - name: Wait for the Mailman LMTP listener
+      ansible.builtin.wait_for:
+        host: '{{ mailman_lmtp_host }}'
+        port: '{{ mailman_lmtp_port }}'
+        state: started
+        sleep: 1
+        timeout: '{{ mailman_weekly_verified_restart_start_timeout }}'
+      when: not ansible_check_mode
+
     - name: Install the mailman and mailmansuite logrotate configurations
       template: src={{ item }}-logrotate.j2 dest=/etc/logrotate.d/{{ item }} owner=root mode=0444
       with_items:

templates/mailman-verified-restart.service.systemd.j2

@@ -6,4 +6,4 @@ Wants=network-online.target
 [Service]
 Type=oneshot
 ExecStart=/usr/local/sbin/mailman-verified-restart
-TimeoutStartSec={{ (mailman_weekly_verified_restart_stop_timeout | int) + (mailman_weekly_verified_restart_start_timeout | int) + 30 }}
+TimeoutStartSec={{ (mailman_service_stop_timeout | int) + (mailman_weekly_verified_restart_stop_timeout | int) + (mailman_weekly_verified_restart_start_timeout | int) + 30 }}

templates/mailman-verified-restart.sh.j2

@@ -5,6 +5,8 @@ set -uo pipefail
 readonly SERVICE='mailman.service'
 readonly MAILMAN='{{ mailman_bindir }}/mailman'
 readonly CONFIG='{{ mailman_conf_dir }}/mailman.cfg'
+readonly LMTP_PORT={{ mailman_lmtp_port | int }}
+readonly SYSTEMD_STOP_TIMEOUT={{ mailman_service_stop_timeout | int }}
 readonly STOP_TIMEOUT={{ mailman_weekly_verified_restart_stop_timeout | int }}
 readonly START_TIMEOUT={{ mailman_weekly_verified_restart_start_timeout | int }}
 
@@ -13,16 +15,40 @@ log()
     printf 'mailman-verified-restart: %s\n' "$*"
 }
 
+lmtp_is_listening()
+{
+    local sockets
+    local suffix=":${LMTP_PORT}"
+
+    if ! sockets="$(ss -H -ltn)"; then
+        return 2
+    fi
+
+    awk -v suffix="${suffix}" '
+        substr($4, length($4) - length(suffix) + 1) == suffix {
+            found = 1
+        }
+        END { exit !found }
+    ' <<< "${sockets}"
+}
+
 mailman_is_stopped()
 {
-    ! systemctl is-active --quiet "${SERVICE}" &&
+    local listener_status
+
+    lmtp_is_listening
+    listener_status=$?
+
+    [[ ${listener_status} -eq 1 ]] &&
+        ! systemctl is-active --quiet "${SERVICE}" &&
         ! "${MAILMAN}" -C "${CONFIG}" status >/dev/null 2>&1
 }
 
 mailman_is_started()
 {
     systemctl is-active --quiet "${SERVICE}" &&
-        "${MAILMAN}" -C "${CONFIG}" status >/dev/null 2>&1
+        "${MAILMAN}" -C "${CONFIG}" status >/dev/null 2>&1 &&
+        lmtp_is_listening
 }
 
 wait_for_state()
@@ -41,9 +67,8 @@ wait_for_state()
 }
 
 log 'stopping Mailman'
-if ! systemctl stop "${SERVICE}"; then
+if ! timeout "$((SYSTEMD_STOP_TIMEOUT + 10))" systemctl stop "${SERVICE}"; then
-    log 'ERROR: systemctl could not stop Mailman'
+    log 'WARNING: systemctl stop did not complete successfully; verifying the actual state'
-    exit 1
 fi
 
 if ! wait_for_state mailman_is_stopped "${STOP_TIMEOUT}"; then
@@ -65,4 +90,4 @@ if ! wait_for_state mailman_is_started "${START_TIMEOUT}"; then
     exit 1
 fi
 
-log 'Mailman is active and its master process is running'
+log "Mailman is active, its master is running and LMTP port ${LMTP_PORT} is listening"

templates/mailman.cfg.j2

@@ -242,7 +242,7 @@ header_checks:
 # The chain to jump to if any of the header patterns matches.  This must be
 # the name of an existing chain such as 'discard', 'reject', 'hold', or
 # 'accept', otherwise 'hold' will be used.
-jump_chain: "{{ mailman_antispam_chain_behaviour }}"
+jump_chain: {{ mailman_antispam_chain_behaviour }}
 
 [runner.nntp]
 class: mailman.runners.nntp.NNTPRunner
@@ -253,4 +253,4 @@ start: {{ mailman_start_nntp_runner }}
 class: mailman_hyperkitty.Archiver
 enable: yes
 configuration: {{ mailman_conf_dir }}/mailman-hyperkitty.cfg
-{% endif %}
+{% endif %}

templates/mailman.service.systemd.j2

@@ -10,6 +10,7 @@ Environment=PATH={{ mailman_bindir }}
 ExecStart={{ mailman_bindir }}/mailman -C {{ mailman_conf_dir }}/mailman.cfg start
 ExecReload={{ mailman_bindir }}/mailman -C {{ mailman_conf_dir }}/mailman.cfg restart
 ExecStop={{ mailman_bindir }}/mailman -C {{ mailman_conf_dir }}/mailman.cfg stop
+ExecStopPost=/bin/rm -f {{ mailman_var_dir }}/master.pid
 Type=forking
 PIDFile={{ mailman_var_dir }}/master.pid
 SyslogIdentifier=mailman
@@ -17,6 +18,9 @@ User={{ mailman_user }}
 Group={{ mailman_user }}
 Restart=on-failure
 RestartSec=5s
+TimeoutStopSec={{ mailman_service_stop_timeout }}s
+KillMode=control-group
+SendSIGKILL=yes
 
 [Install]
 WantedBy=multi-user.target