retire Graylog behind a flag, fix Caddy reloads, reap awsddns zombies

Graylog was the worst cost/benefit tenant on this 4-core box: two JVMs plus
MongoDB holding ~1.6 GB resident and ~3% CPU around the clock to store ~3k
messages a day -- about 28 MB across its four live indices. journald already
retains ~25 days of the same logs at its 500M cap, so this costs searchability,
not the logs.

The switch is `graylog_enabled` in inventory rather than a role default,
because three roles read it (common, podman, graylog-config). The disabled
path is an active teardown, not a skipped create: the containers already on
the host keep running and their systemd user units keep restarting them at
boot unless something stops and removes them. fluent-bit follows the same
flag -- with the GELF sink down it would spin retrying a dead 127.0.0.1:12202
and fill the journal it exists to drain -- but only the service state follows,
so re-enabling is a restart rather than a reinstall.

Caddy reloads were silently no-ops. The handler read /etc/caddy/Caddyfile,
which is a single-file bind mount, and podman binds those by inode; the
template module writes a temp file and renames it into place, so every deploy
gave the host file a new inode while the container kept seeing the one it was
created with. Config changes only ever landed when something recreated the
container. {{ caddy_path }}/config is also mounted, as a *directory*, and
directory mounts resolve names at open() time -- so /config/Caddyfile is
always the file Ansible just wrote.

awsddns and its four siblings had accumulated 12 zombies over 30 days of
uptime. The image's PID 1 is busybox crond, which only waitpid()s the job PIDs
it tracks and does no generic orphan reaping, so whenever the run-parts/sh
layer exited before the script it left a permanent <defunct>. init: true puts
catatonit at PID 1 to reap them, and the recreation clears the existing ones.

Also bumps fulfillr and greg-time-bot images.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

ansible/roles/podman/tasks/containers/base/awsddns.yml

@@ -1,4 +1,14 @@
 ---
+# The image's PID 1 is busybox crond (ENTRYPOINT crond -f -d 8), which runs the
+# /etc/periodic/15min/awsddns script every quarter hour. busybox crond only
+# waitpid()s the job PIDs it is tracking -- it does no generic orphan reaping --
+# so whenever the run-parts/sh layer between crond and the script exits first,
+# the script is reparented to PID 1 and stays <defunct> forever. That had left
+# 12 zombies across these five containers after 30 days of uptime.
+#
+# init: true makes podman inject catatonit as PID 1 with crond as its child, and
+# catatonit reaps every orphan it inherits. Changing this recreates the
+# containers, which also clears the zombies already accumulated.
 - import_tasks: podman/podman-check.yml
   vars:
     container_name: awsddns
@@ -13,6 +23,7 @@
     image: "{{ image }}"
     restart_policy: on-failure:3
     log_driver: journald
+    init: true
     env:
       AWS_ZONE_TTL: 60
       AWS_ZONE_ID: "{{ aws_zone_id }}"
@@ -40,6 +51,7 @@
     image: "{{ image }}"
     restart_policy: on-failure:3
     log_driver: journald
+    init: true
     env:
       AWS_ZONE_TTL: 60
       AWS_ZONE_ID: "{{ aws_skudak_zone_id }}"
@@ -67,6 +79,7 @@
     image: "{{ image }}"
     restart_policy: on-failure:3
     log_driver: journald
+    init: true
     env:
       AWS_ZONE_TTL: 60
       AWS_ZONE_ID: "{{ fulfillr_zone_id }}"
@@ -96,6 +109,7 @@
     image: "{{ image }}"
     restart_policy: on-failure:3
     log_driver: journald
+    init: true
     env:
       AWS_ZONE_TTL: 60
       AWS_ZONE_ID: "{{ fulfillr_zone_id }}"
@@ -123,6 +137,7 @@
     image: "{{ image }}"
     restart_policy: on-failure:3
     log_driver: journald
+    init: true
     env:
       AWS_ZONE_TTL: 60
       AWS_ZONE_ID: "Z07501202A6AYMHCVP50A"

ansible/roles/podman/tasks/containers/debyltech/graylog-teardown.yml

@@ -0,0 +1,59 @@
+---
+# Runs in place of graylog.yml when graylog_enabled is false.
+#
+# A bare `when:` on the import would only stop Ansible from *creating* the
+# stack; containers already on the host would keep running and their systemd
+# user units would keep starting them at boot. So the disabled path has to be an
+# active teardown: disable the units, then remove the containers.
+#
+# Volumes under {{ graylog_path }} are deliberately untouched -- indices, the
+# Mongo database holding streams/pipelines/dashboards, and the node-id file all
+# survive, so flipping graylog_enabled back to true resumes where this left off.
+
+# graylog is stopped first: its unit `requires` the other two, and tearing down
+# a dependency out from under it makes the shutdown noisy for no reason.
+- name: stop and disable graylog stack systemd units
+  become: true
+  become_user: "{{ podman_user }}"
+  ansible.builtin.systemd:
+    name: "{{ item }}.service"
+    enabled: false
+    state: stopped
+    daemon_reload: true
+    scope: user
+  loop:
+    - graylog
+    - graylog-opensearch
+    - graylog-mongo
+    # Left over from the Elasticsearch-to-OpenSearch migration; the container is
+    # long gone but the enabled unit still tries to start it every boot.
+    - graylog-elastic
+  register: graylog_units
+  failed_when: false
+  tags: graylog
+
+- name: remove graylog stack containers
+  become: true
+  become_user: "{{ podman_user }}"
+  containers.podman.podman_container:
+    name: "{{ item }}"
+    state: absent
+  loop:
+    - graylog
+    - graylog-opensearch
+    - graylog-mongo
+  tags: graylog
+
+- name: remove graylog stack systemd unit files
+  become: true
+  become_user: "{{ podman_user }}"
+  ansible.builtin.file:
+    path: "{{ podman_home }}/.config/systemd/user/{{ item }}.service"
+    state: absent
+  loop:
+    - graylog
+    - graylog-opensearch
+    - graylog-mongo
+    - graylog-elastic
+  notify: reload podman systemd
+  tags: graylog