Set log_size_max size for podman logs

Sometimes the logs can take over 20GB in few weeks.
By setting 1GB as log_size_max should avoid situation that we are
out of the disk few times per week.
The feature has been added into the podman containers.conf file in
podman 2.2.0 release [1], but on Centos 7, version is below 2.2.0.
According to the libpod.conf man [2], that option should be also
available in podman 1.6.4, but it is located in libpod.conf file.
More info [3].

[1] https://github.com/containers/podman/releases/tag/v2.2.0
[2] https://manpages.debian.org/unstable/podman/libpod.conf.5.en.html
[3] unifi-utilities/unifios-utilities#100

Change-Id: Ic6d01e11606c9526d1880583876d76c4415250ac

playbooks/site_rdo.yaml

@@ -9,6 +9,7 @@
   roles:
     - service/node-exporter
     - system/update-host
+    - system/podman
 
 - hosts: backup-rdo
   roles:

playbooks/site_sf.yaml

@@ -11,6 +11,7 @@
     - infra/ssh
     - service/node-exporter
     - system/update-host
+    - system/podman
 
 - hosts: backup-sf
   roles:

roles/system/podman/defaults/main.yaml

@@ -0,0 +1,3 @@
+---
+# set 1000MB in bytes
+log_size_max: 1048576000

roles/system/podman/tasks/main.yaml

@@ -0,0 +1,18 @@
+---
+- name: Set max log limit - new way
+  become: true
+  ini_file:
+    path: /etc/containers/containers.conf
+    section: containers
+    option: log_size_max
+    value: "{{ log_size_max }}"
+    backup: true
+
+- name: Set max log limit - legacy
+  become: true
+  lineinfile:
+    path: /etc/containers/libpod.conf
+    regexp: "^max_log_size"
+    line: "max_log_size={{ log_size_max }}"
+    create: true
+  when: ansible_distribution_major_version|int < 8