lint: commands should not change things

Fix ansible lint 301 error:

[301] Commands should not change things if nothing needs doing

Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>

infrastructure-playbooks/lv-teardown.yml

@@ -37,14 +37,18 @@
       set -o pipefail;
       grep /var/lib/ceph/osd /proc/mounts | awk '{print $2}'
     register: old_osd_filesystems
+    changed_when: false
 
-  - name: tear down any existing osd filesystems
-    command: "umount -v {{ item }}"
+  - name: tear down any existing osd filesystem
+    mount:
+      path: "{{ item }}"
+      state: unmounted
     with_items: "{{ old_osd_filesystems.stdout_lines }}"
 
   - name: kill all lvm commands that may have been hung
     command: "killall -q lvcreate pvcreate vgcreate lvconvert || echo -n"
     failed_when: false
+    changed_when: false
 
   ## Logcal Vols
   - name: tear down existing lv for bucket index
@@ -96,7 +100,9 @@
   ## Physical Vols
   - name: tear down pv for nvme device
     command: "pvremove --force --yes {{ nvme_device }}"
+    changed_when: false
 
   - name: tear down pv for each hdd device
     command: "pvremove --force --yes {{ item }}"
+    changed_when: false
     with_items: "{{ hdd_devices }}"

infrastructure-playbooks/purge-cluster.yml

@@ -87,6 +87,7 @@
 
     - name: ensure cephfs mountpoint(s) are unmounted
       command: umount -a -t ceph
+      changed_when: false
 
     - name: find mapped rbd ids
       find:
@@ -96,6 +97,7 @@
 
     - name: use sysfs to unmap rbd devices
       shell: "echo {{ item.path | basename }} > /sys/bus/rbd/remove_single_major"
+      changed_when: false
       with_items: "{{ rbd_mapped_ids.files }}"
 
     - name: unload ceph kernel modules
@@ -415,6 +417,7 @@
       blkid -t TYPE=crypto_LUKS -s PARTLABEL -s PARTUUID | grep "ceph.*." | grep -o PARTUUID.* | cut -d '"' -f 2
     register: encrypted_ceph_partuuid
     failed_when: false
+    changed_when: false
 
   - name: get osd data and lockbox mount points
     shell: |
@@ -425,9 +428,12 @@
 
   - name: drop all cache
     shell: "sync && sleep 1 && echo 3 > /proc/sys/vm/drop_caches"
+    changed_when: false
 
   - name: umount osd data partition
-    shell: umount {{ item }}
+    mount:
+      path: "{{ item }}"
+      state: unmounted
     with_items: "{{ mounted_osd.stdout_lines }}"
 
   - name: remove osd mountpoint tree
@@ -470,12 +476,14 @@
 
   - name: get physical sector size
     command: blockdev --getpbsz /dev/disk/by-partuuid/{{ item }}
+    changed_when: false
     with_items: "{{ encrypted_ceph_partuuid.stdout_lines }}"
     when: encrypted_ceph_partuuid.stdout_lines | length > 0
     register: phys_sector_size
 
   - name: wipe dmcrypt device
     command: dd if=/dev/zero of=/dev/disk/by-partuuid/{{ item.0 }} bs={{ item.1.stdout }} count={{ item.2.stdout }} oflag=direct
+    changed_when: false
     with_together:
       - "{{ encrypted_ceph_partuuid.stdout_lines }}"
       - "{{ payload_offset.results }}"
@@ -484,17 +492,20 @@
   - name: get ceph data partitions
     shell: |
       blkid -o device -t PARTLABEL="ceph data"
+    changed_when: false
     failed_when: false
     register: ceph_data_partition_to_erase_path
 
   - name: get ceph lockbox partitions
     shell: |
       blkid -o device -t PARTLABEL="ceph lockbox"
+    changed_when: false
     failed_when: false
     register: ceph_lockbox_partition_to_erase_path
 
   - name: see if ceph-volume is installed
     command: command -v ceph-volume
+    changed_when: false
     failed_when: false
     register: ceph_volume_present
 
@@ -530,24 +541,28 @@
   - name: get ceph block partitions
     shell: |
       blkid -o device -t PARTLABEL="ceph block"
+    changed_when: false
     failed_when: false
     register: ceph_block_partition_to_erase_path
 
   - name: get ceph journal partitions
     shell: |
       blkid -o device -t PARTLABEL="ceph journal"
+    changed_when: false
     failed_when: false
     register: ceph_journal_partition_to_erase_path
 
   - name: get ceph db partitions
     shell: |
       blkid -o device -t PARTLABEL="ceph block.db"
+    changed_when: false
     failed_when: false
     register: ceph_db_partition_to_erase_path
 
   - name: get ceph wal partitions
     shell: |
       blkid -o device -t PARTLABEL="ceph block.wal"
+    changed_when: false
     failed_when: false
     register: ceph_wal_partition_to_erase_path
 
@@ -563,6 +578,7 @@
   - name: resolve parent device
     command: lsblk --nodeps -no pkname "{{ item }}"
     register: tmp_resolved_parent_device
+    changed_when: false
     with_items: "{{ combined_devices_list }}"
 
   - name: set_fact resolved_parent_device
@@ -574,6 +590,7 @@
       set -o pipefail;
       wipefs --all "{{ item }}"
       dd if=/dev/zero of="{{ item }}" bs=1 count=4096
+    changed_when: false
     with_items: "{{ combined_devices_list }}"
 
   - name: zap ceph journal/block db/block wal partitions
@@ -592,6 +609,7 @@
       partprobe /dev/"{{ item }}"
       udevadm settle --timeout=600
     with_items: "{{ resolved_parent_device }}"
+    changed_when: false
 
 - name: purge ceph mon cluster
 
@@ -809,7 +827,7 @@
   - name: request data removal
     local_action:
       module: command
-        echo requesting data removal
+        echo requesting data removal  # noqa 301
     become: false
     notify: remove data
 
@@ -839,6 +857,7 @@
   - name: check for anything running ceph
     command: "ps -u ceph -U ceph"
     register: check_for_running_ceph
+    changed_when: false
     failed_when: check_for_running_ceph.rc == 0
 
   - name: find ceph systemd unit files to remove

infrastructure-playbooks/purge-container-cluster.yml

@@ -71,6 +71,7 @@
 
     - name: ensure cephfs mountpoint are unmounted
       command: umount -a -t ceph
+      changed_when: false
 
     - name: find mapped rbd ids
       find:
@@ -80,6 +81,7 @@
 
     - name: use sysfs to unmap rbd devices
       shell: "echo {{ item.path | basename }} > /sys/bus/rbd/remove_single_major"
+      changed_when: false
       with_items: "{{ rbd_mapped_ids.files }}"
 
     - name: unload ceph kernel modules
@@ -268,6 +270,7 @@
     shell: |
       systemctl list-units --all | grep -oE "ceph-osd@([0-9]+).service"
     register: osd_units
+    changed_when: false
     ignore_errors: true
 
   - name: disable ceph osd service
@@ -576,6 +579,7 @@
 
     - name: remove ceph container image
       command: "{{ container_binary }} rmi {{ ceph_docker_registry }}/{{ ceph_docker_image }}:{{ ceph_docker_image_tag }}"
+      changed_when: false
       tags:
         - remove_img
 
@@ -668,6 +672,7 @@
 
   - name: remove ceph data
     shell: rm -rf /var/lib/ceph/*
+    changed_when: false
 
   # (todo): remove this when we are able to manage docker
   # service on atomic host.

infrastructure-playbooks/rolling_update.yml

@@ -345,6 +345,7 @@
 
     - name: set osd flags
       command: "{{ container_exec_cmd | default('') }} ceph --cluster {{ cluster }} osd set {{ item }}"
+      changed_when: false
       with_items:
         - noout
         - nodeep-scrub
@@ -415,6 +416,7 @@
 
     - name: get num_pgs - non container
       command: "{{ container_exec_cmd_update_osd|default('') }} ceph --cluster {{ cluster }} pg stat --format json"
+      changed_when: false
       register: ceph_pgs
       delegate_to: "{{ groups[mon_group_name][0] }}"
 
@@ -448,6 +450,7 @@
 
     - name: unset osd flags
       command: "{{ container_exec_cmd_update_osd | default('') }} ceph osd unset {{ item }} --cluster {{ cluster }}"
+      changed_when: false
       with_items:
         - noout
         - nodeep-scrub
@@ -979,10 +982,12 @@
 
     - name: show ceph status
       command: "{{ container_exec_cmd_status|default('') }} ceph --cluster {{ cluster }} -s"
+      changed_when: false
       run_once: True
       delegate_to: "{{ groups[mon_group_name][0] }}"
 
     - name: show all daemons version
       command: "{{ container_exec_cmd_status|default('') }} ceph --cluster {{ cluster }} versions"
       run_once: True
       delegate_to: "{{ groups[mon_group_name][0] }}"
+      changed_when: false

infrastructure-playbooks/shrink-mds.yml

@@ -66,6 +66,7 @@
 
     - name: exit playbook, if can not connect to the cluster
       command: "{{ container_exec_cmd | default('') }} timeout 5 ceph --cluster {{ cluster }} health"
+      changed_when: false
       register: ceph_health
       until: ceph_health is succeeded
       retries: 5
@@ -80,11 +81,13 @@
     # removes the MDS from the FS map.
     - name: exit mds when containerized deployment
       command: "{{ container_exec_cmd | default('') }} ceph tell mds.{{ mds_to_kill_hostname }} exit"
+      changed_when: false
       when: containerized_deployment | bool
 
     - name: get ceph status
       command: "{{ container_exec_cmd | default('') }} ceph --cluster {{ cluster }} -s -f json"
       register: ceph_status
+      changed_when: false
 
     - name: set_fact current_max_mds
       set_fact:
@@ -162,3 +165,4 @@
   post_tasks:
     - name: show ceph health
       command: "{{ container_exec_cmd | default('') }} ceph --cluster {{ cluster }} -s"
+      changed_when: false

infrastructure-playbooks/shrink-mgr.yml

@@ -44,6 +44,7 @@
     - name: exit playbook, if can not connect to the cluster
       command: "{{ container_exec_cmd | default('') }} timeout 5 ceph --cluster {{ cluster }} health"
       register: ceph_health
+      changed_when: false
       until: ceph_health is succeeded
       retries: 5
       delay: 2
@@ -115,6 +116,7 @@
     - name: fail if the mgr is reported in ceph mgr dump
       command: "{{ container_exec_cmd | default('') }} ceph --cluster {{ cluster }} mgr dump -f json"
       register: mgr_dump
+      changed_when: false
       failed_when: mgr_to_kill_hostname in (([mgr_dump.stdout | from_json] | map(attribute='active_name') | list) + (mgr_dump.stdout | from_json)['standbys'] | map(attribute='name') | list)
       until: mgr_to_kill_hostname not in (([mgr_dump.stdout | from_json] | map(attribute='active_name') | list) + (mgr_dump.stdout | from_json)['standbys'] | map(attribute='name') | list)
       retries: 12
@@ -129,3 +131,4 @@
   post_tasks:
     - name: show ceph health
       command: "{{ container_exec_cmd | default('') }} ceph --cluster {{ cluster }} -s"
+      changed_when: false

infrastructure-playbooks/shrink-mon.yml

@@ -82,6 +82,7 @@
     - name: exit playbook, if can not connect to the cluster
       command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health"
       register: ceph_health
+      changed_when: false
       until: ceph_health.stdout.find("HEALTH") > -1
       delegate_to: "{{ mon_host }}"
       retries: 5
@@ -107,13 +108,15 @@
 
     - name: remove monitor from the quorum
       command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} mon remove {{ mon_to_kill_hostname }}"
+      changed_when: false
       failed_when: false
       delegate_to: "{{ mon_host }}"
 
   post_tasks:
     - name: verify the monitor is out of the cluster
       command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} quorum_status -f json"
       delegate_to: "{{ mon_host }}"
+      changed_when: false
       failed_when: false
       register: result
       until: mon_to_kill_hostname not in (result.stdout | from_json)['quorum_names']
@@ -136,7 +139,9 @@
     - name: show ceph health
       command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s"
       delegate_to: "{{ mon_host }}"
+      changed_when: false
 
     - name: show ceph mon status
       command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} mon stat"
       delegate_to: "{{ mon_host }}"
+      changed_when: false

infrastructure-playbooks/shrink-osd.yml

@@ -75,12 +75,14 @@
     - name: exit playbook, if can not connect to the cluster
       command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health"
       register: ceph_health
+      changed_when: false
       until: ceph_health.stdout.find("HEALTH") > -1
       retries: 5
       delay: 2
 
     - name: find the host(s) where the osd(s) is/are running on
       command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}"
+      changed_when: false
       with_items: "{{ osd_to_kill.split(',') }}"
       register: find_osd_hosts
 
@@ -99,6 +101,7 @@
 
     - name: get ceph-volume lvm list data
       command: "{{ container_run_cmd }} lvm list --format json"
+      changed_when: false
       register: _lvm_list_data
       delegate_to: "{{ item.0 }}"
       loop: "{{ _osd_hosts }}"
@@ -135,6 +138,7 @@
 
     - name: mark osd(s) out of the cluster
       command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd out {{ osd_to_kill.replace(',', ' ') }}"
+      changed_when: false
       run_once: true
 
     - name: stop osd(s) service
@@ -220,11 +224,13 @@
 
     - name: ensure osds are marked down
       command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd down {{ osd_to_kill.replace(',', ' ') }}"
+      changed_when: false
       run_once: true
       delegate_to: "{{ groups[mon_group_name][0] }}"
 
     - name: purge osd(s) from the cluster
       command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd purge {{ item }} --yes-i-really-mean-it"
+      changed_when: false
       run_once: true
       with_items: "{{ osd_to_kill.split(',') }}"
 
@@ -237,6 +243,8 @@
 
     - name: show ceph health
       command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s"
+      changed_when: false
 
     - name: show ceph osd tree
       command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree"
+      changed_when: false