Merge pull request #2009 from ceph/fix-clean-pg

[skip ci] handler: do not test if pgs_num = 0

infrastructure-playbooks/purge-docker-cluster.yml

@@ -270,53 +270,135 @@
 
   tasks:
 
+  - name: get all the running osds
+    shell: |
+      systemctl list-units | grep "loaded active" | grep -oE "ceph-osd@([0-9]{1,2}|[a-z]+).service"
+    register: osd_units
+
   - name: disable ceph osd service
     service:
-      name: "ceph-osd@{{ item | basename }}"
+      name: "ceph-osd@{{ item }}"
       state: stopped
       enabled: no
-    with_items: "{{ devices }}"
-    ignore_errors: true
+    with_items: "{{ osd_units.stdout_lines }}"
 
-  - name: resolve device link
-    command: readlink -f {{ item }}
-    changed_when: false
-    with_items: "{{ devices }}"
-    register: purge_devices_links
-
-  - name: set_fact devices generate device list when osd_auto_discovery
-    set_fact:
-      devices: "{{ devices | default([]) + [ item.stdout ] }}"
-    with_items: "{{ purge_devices_links.results }}"
+  - name: get prepare container
+    command: "docker ps -a -q --filter='name=ceph-osd-prepare'"
+    register: prepare_containers
 
   - name: remove ceph osd prepare container
-    docker:
-      image: "{{ ceph_docker_registry }}/{{ ceph_docker_image }}:{{ ceph_docker_image_tag }}"
-      name: "ceph-osd-prepare-{{ ansible_hostname }}{{ item | regex_replace('/dev/', '') }}"
+    command: "docker rm -f {{ item }}"
+    with_items: "{{ prepare_containers.stdout_lines }}"
+
+  - name: see if ceph-disk-created data partitions are present
+    shell: |
+      ls /dev/disk/by-partlabel | grep -q "ceph.*.data"
+    failed_when: false
+    register: ceph_data_partlabels
+
+  - name: see if ceph-disk-created block partitions are present
+    shell: |
+      ls /dev/disk/by-partlabel | grep -q "ceph.*block$"
+    failed_when: false
+    register: ceph_block_partlabels
+
+  - name: see if ceph-disk-created journal partitions are present
+    shell: |
+      ls /dev/disk/by-partlabel | grep -q "ceph.*.journal"
+    failed_when: false
+    register: ceph_journal_partlabels
+
+  - name: see if ceph-disk-created block db partitions are present
+    shell: |
+      ls /dev/disk/by-partlabel | grep -q "ceph.*.block.db"
+    failed_when: false
+    register: ceph_db_partlabels
+
+  - name: see if ceph-disk-created block wal partitions are present
+    shell: |
+      ls /dev/disk/by-partlabel | grep -q "ceph.*.block.wal"
+    failed_when: false
+    register: ceph_wal_partlabels
+
+  - name: see if ceph-disk-created lockbox partitions are present
+    shell: |
+      ls /dev/disk/by-partlabel | grep -q "ceph.*.lockbox"
+    failed_when: false
+    register: ceph_lockbox_partlabels
+
+  # NOTE(leseb): hope someone will find a more elegant way one day...
+  - name: see if encrypted partitions are present
+    shell: |
+      blkid -t TYPE=crypto_LUKS -s PARTLABEL -s PARTUUID | grep "ceph.*." | grep -o PARTUUID.* | cut -d '"' -f 2
+    register: encrypted_ceph_partuuid
+
+  - name: remove osd mountpoint tree
+    file:
+      path: /var/lib/ceph/osd/
       state: absent
-    with_items: "{{ devices }}"
+    register: remove_osd_mountpoints
     ignore_errors: true
 
-  - name: remove ceph osd container
-    docker:
-      image: "{{ ceph_docker_registry }}/{{ ceph_docker_image }}:{{ ceph_docker_image_tag }}"
-      name: "ceph-osd-{{ ansible_hostname }}-{{ item | regex_replace('/dev/', '') }}"
-      state: absent
-    with_items: "{{ devices }}"
-    ignore_errors: true
+  - name: get ceph data partitions
+    shell: |
+      blkid | awk -F: '/ceph data/ { print $1 }'
+    when: ceph_data_partlabels.rc == 0
+    failed_when: false
+    register: ceph_data_partition_to_erase_path
+
+  - name: get ceph lockbox partitions
+    shell: |
+      blkid | awk '/ceph lockbox/ { sub (":", "", $1); print $1 }'
+    when: ceph_lockbox_partlabels.rc == 0
+    failed_when: false
+    register: ceph_lockbox_partition_to_erase_path
+
+  - name: get ceph block partitions
+    shell: |
+      blkid | awk '/ceph block"/ { sub (":", "", $1); print $1 }'
+    when: ceph_block_partlabels.rc == 0
+    failed_when: false
+    register: ceph_block_partition_to_erase_path
+
+  - name: get ceph journal partitions
+    shell: |
+      blkid | awk '/ceph journal/ { sub (":", "", $1); print $1 }'
+    when: ceph_journal_partlabels.rc == 0
+    failed_when: false
+    register: ceph_journal_partition_to_erase_path
+
+  - name: get ceph db partitions
+    shell: |
+      blkid | awk '/ceph block.db/ { sub (":", "", $1); print $1 }'
+    when: ceph_db_partlabels.rc == 0
+    failed_when: false
+    register: ceph_db_partition_to_erase_path
+
+  - name: get ceph wal partitions
+    shell: |
+      blkid | awk '/ceph block.wal/ { sub (":", "", $1); print $1 }'
+    when: ceph_wal_partlabels.rc == 0
+    failed_when: false
+    register: ceph_wal_partition_to_erase_path
 
   - name: zap ceph osd disks
     shell: |
-      docker run \
+      docker run --rm \
       --privileged=true \
-      --name ceph-osd-zap-{{ ansible_hostname }}-{{ item | regex_replace('/dev/', '') }} \
+      --name ceph-osd-zap-{{ ansible_hostname }}-{{ item[:-1] | regex_replace('/dev/', '') }} \
       -v /dev/:/dev/ \
-      -e OSD_DEVICE={{ item }} \
+      -e OSD_DEVICE={{ item[:-1] }} \
       {{ ceph_docker_registry }}/{{ ceph_docker_image }}:{{ ceph_docker_image_tag }} \
       zap_device
     with_items:
-      - "{{ devices }}"
-      - "{{ dedicated_devices|default([]) }}"
+      - "{{ ceph_data_partition_to_erase_path.stdout_lines | default([]) }}"
+      - "{{ ceph_lockbox_partition_to_erase_path.stdout_lines | default([]) }}"
+      - "{{ ceph_block_partition_to_erase_path.stdout_lines | default([]) }}"
+      - "{{ ceph_journal_partition_to_erase_path.stdout_lines | default([]) }}"
+      - "{{ ceph_db_partition_to_erase_path.stdout_lines | default([]) }}"
+      - "{{ ceph_wal_partition_to_erase_path.stdout_lines | default([]) }}"
+    when:
+      - (ceph_data_partlabels.rc == 0 or ceph_block_partlabels.rc == 0 or ceph_journal_partlabels.rc == 0 or ceph_db_partlabels.rc == 0 or ceph_wal_partlabels.rc == 0)
 
   - name: wait until the zap containers die
     shell: |
@@ -330,11 +412,15 @@
   - name: remove ceph osd zap disk container
     docker:
       image: "{{ ceph_docker_registry }}/{{ ceph_docker_image }}:{{ ceph_docker_image_tag }}"
-      name: "ceph-osd-zap-{{ ansible_hostname }}-{{ item | regex_replace('/dev/', '') }}"
+      name: "ceph-osd-zap-{{ ansible_hostname }}-{{ item[:-1] | regex_replace('/dev/', '') }}"
       state: absent
     with_items:
-      - "{{ devices }}"
-      - "{{ dedicated_devices|default([]) }}"
+      - "{{ ceph_data_partition_to_erase_path.stdout_lines | default([]) }}"
+      - "{{ ceph_lockbox_partition_to_erase_path.stdout_lines | default([]) }}"
+      - "{{ ceph_block_partition_to_erase_path.stdout_lines | default([]) }}"
+      - "{{ ceph_journal_partition_to_erase_path.stdout_lines | default([]) }}"
+      - "{{ ceph_db_partition_to_erase_path.stdout_lines | default([]) }}"
+      - "{{ ceph_wal_partition_to_erase_path.stdout_lines | default([]) }}"
 
   - name: remove ceph osd service
     file:

infrastructure-playbooks/rolling_update.yml

@@ -147,7 +147,21 @@
       when:
         - mon_host_count | int == 1
 
-    - name: waiting for the monitor to join the quorum...
+    - name: get num_pgs - non container
+      command: ceph --cluster "{{ cluster }}" -s --format json
+      register: ceph_pgs
+      delegate_to: "{{ mon_host }}"
+      when:
+        - not containerized_deployment
+
+    - name: get num_pgs - container
+      command: docker exec ceph-mon-{{ hostvars[mon_host]['ansible_hostname'] }} ceph --cluster "{{ cluster }}" -s --format json
+      register: ceph_pgs
+      delegate_to: "{{ mon_host }}"
+      when:
+        - containerized_deployment
+
+    - name: non container | waiting for the monitor to join the quorum...
       command: ceph --cluster "{{ cluster }}" -s --format json
       register: ceph_health_raw
       until: >
@@ -157,8 +171,9 @@
       delegate_to: "{{ mon_host }}"
       when:
         - not containerized_deployment
+        - (ceph_pgs.stdout | from_json).pgmap.num_pgs != "0"
 
-    - name: waiting for the containerized monitor to join the quorum...
+    - name: container | waiting for the containerized monitor to join the quorum...
       command: docker exec ceph-mon-{{ hostvars[mon_host]['ansible_hostname'] }} ceph --cluster "{{ cluster }}" -s --format json
       register: ceph_health_raw
       until: >
@@ -168,6 +183,7 @@
       delegate_to: "{{ mon_host }}"
       when:
         - containerized_deployment
+        - (ceph_pgs.stdout | from_json).pgmap.num_pgs != "0"
 
     - name: set osd flags
       command: ceph osd set {{ item }} --cluster {{ cluster }}

infrastructure-playbooks/shrink-osd.yml

@@ -91,12 +91,15 @@
       with_items: "{{ osd_hosts }}"
       delegate_to: "{{ item }}"
       failed_when: false
+      when:
+        - not containerized_deployment
 
     - name: fail when admin key is not present
       fail:
         msg: "The Ceph admin key is not present on the OSD node, please add it and remove it after the playbook is done."
       with_items: "{{ ceph_admin_key.results }}"
       when:
+        - not containerized_deployment
         - item.stat.exists == false
 
     # NOTE(leseb): using '>' is the only way I could have the command working
@@ -127,8 +130,6 @@
 
     - name: deactivating osd(s)
       command: ceph-disk deactivate --cluster {{ cluster }} --deactivate-by-id {{ item.0 }} --mark-out
-      register: deactivate
-      ignore_errors: yes
       run_once: true
       with_together:
         - "{{ osd_to_kill.split(',') }}"
@@ -137,21 +138,8 @@
       when:
         - not containerized_deployment
 
-    - name: set osd(s) out when ceph-disk deactivating fail
-      command: "{{ docker_exec_cmd }} ceph --cluster {{ cluster }} osd out osd.{{ item.0 }}"
-      delegate_to: "{{ groups[mon_group_name][0] }}"
-      with_together:
-        - "{{ osd_to_kill.split(',') }}"
-        - "{{ deactivate.results }}"
-      when:
-        - not containerized_deployment
-        - not item.1.get("skipped")
-        - item.1.stderr|length > 0
-
     - name: destroying osd(s)
       command: ceph-disk destroy --cluster {{ cluster }} --destroy-by-id {{ item.0 }} --zap
-      register: destroy
-      ignore_errors: yes
       run_once: true
       with_together:
         - "{{ osd_to_kill.split(',') }}"
@@ -161,32 +149,20 @@
         - not containerized_deployment
 
     - name: remove osd(s) from crush_map when ceph-disk destroy fail
-      command: "{{ docker_exec_cmd }} ceph --cluster {{ cluster }} osd crush remove osd.{{ item.0 }}"
+      command: "{{ docker_exec_cmd }} ceph --cluster {{ cluster }} osd crush remove osd.{{ item }}"
       run_once: true
       delegate_to: "{{ groups[mon_group_name][0] }}"
-      with_together:
-        - "{{ osd_to_kill.split(',') }}"
-        - "{{ destroy.results }}"
-      when:
-        - (item.1.get("skipped") or item.1.stderr|length > 0)
+      with_items: "{{ osd_to_kill.split(',') }}"
 
     - name: delete osd(s) auth key when ceph-disk destroy fail
-      command: "{{ docker_exec_cmd }} ceph --cluster {{ cluster }} auth del osd.{{ item.0 }}"
+      command: "{{ docker_exec_cmd }} ceph --cluster {{ cluster }} auth del osd.{{ item }}"
       delegate_to: "{{ groups[mon_group_name][0] }}"
-      with_together:
-        - "{{ osd_to_kill.split(',') }}"
-        - "{{ destroy.results }}"
-      when:
-        - (item.1.get("skipped") or item.1.stderr|length > 0)
+      with_items: "{{ osd_to_kill.split(',') }}"
 
     - name: deallocate osd(s) id when ceph-disk destroy fail
-      command: "{{ docker_exec_cmd }} ceph --cluster {{ cluster }} osd rm {{ item.0 }}"
+      command: "{{ docker_exec_cmd }} ceph --cluster {{ cluster }} osd rm {{ item }}"
       delegate_to: "{{ groups[mon_group_name][0] }}"
-      with_together:
-        - "{{ osd_to_kill.split(',') }}"
-        - "{{ destroy.results }}"
-      when:
-        - (item.1.get("skipped") or item.1.stderr|length > 0)
+      with_items: "{{ osd_to_kill.split(',') }}"
 
     - name: show ceph health
       command: "{{ docker_exec_cmd }} ceph --cluster {{ cluster }} -s"

infrastructure-playbooks/switch-from-non-containerized-to-containerized-ceph-daemons.yml

@@ -280,14 +280,16 @@
     - ceph-osd
 
   post_tasks:
-    - name: waiting for clean pgs...
-      shell: |
-        test "[""$(docker exec ceph-mon-{{ hostvars[groups[mon_group_name][0]]['ansible_hostname'] }} ceph --cluster {{ cluster }} -s -f json | python -c 'import sys, json; print(json.load(sys.stdin)["pgmap"]["num_pgs"])')""]" = "$(docker exec ceph-mon-{{ hostvars[groups[mon_group_name][0]]['ansible_hostname'] }} ceph --cluster {{ cluster }} -s -f json | python -c 'import sys, json; print [ i["count"] for i in json.load(sys.stdin)["pgmap"]["pgs_by_state"] if i["state_name"] == "active+clean"]')"
-      register: result
-      until: result.rc == 0
+    - name: container - waiting for clean pgs...
+      command: "docker exec ceph-mon-{{ hostvars[groups[mon_group_name][0]]['ansible_hostname'] }} ceph --cluster {{ cluster }} -s --format json"
+      register: ceph_health_post
+      until: >
+        ((ceph_health_post.stdout | from_json).pgmap.pgs_by_state | length) == 1
+        and
+        (ceph_health_post.stdout | from_json).pgmap.pgs_by_state.0.state_name == "active+clean"
+      delegate_to: "{{ groups[mon_group_name][0] }}"
       retries: "{{ health_osd_check_retries }}"
       delay: "{{ health_osd_check_delay }}"
-      delegate_to: "{{ groups[mon_group_name][0] }}"
 
 
 - name: switching from non-containerized to containerized ceph mds

roles/ceph-common/tasks/installs/prerequisite_rhcs_iso_install_debian.yml

@@ -45,13 +45,15 @@
 
 - name: install red hat storage repository key for debian systems
   apt_key:
-    file: "{{ ceph_rhcs_repository_path }}/RPM-GPG-KEY-redhat-release"
+    file: "{{ ceph_rhcs_repository_path }}/MON/release.asc"
     state: present
 
 - name: add red hat storage repository for debian systems
-  template:
-    src: "{{ role_path }}/templates/redhat_storage_repo.j2"
-    dest: /etc/apt/sources.list.d/rh_storage.list
-    owner: root
-    group: root
-    mode: 0644
+  apt_repository:
+    repo: "deb file://{{ ceph_rhcs_repository_path }}/{{ item }} {{ ansible_lsb.codename }} main"
+    state: present
+  changed_when: false
+  with_items:
+    - MON
+    - OSD
+    - Tools

roles/ceph-defaults/handlers/main.yml

@@ -25,6 +25,9 @@
     - mon_group_name in group_names
     - not containerized_deployment
     - mon_socket_stat.rc == 0
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"
 
 - name: restart ceph mon daemon(s) - container
   command: /tmp/restart_mon_daemon.sh
@@ -33,7 +36,10 @@
     # We do not want to run these checks on initial deployment (`socket.rc == 0`)
     - mon_group_name in group_names
     - containerized_deployment
-    - ceph_mon_container_stat.stdout_lines|length != 0
+    - ceph_mon_container_stat.get('stdout_lines', [])|length != 0
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"
 
 # This does not just restart OSDs but everything else too. Unfortunately
 # at this time the ansible role does not have an OSD id list to use
@@ -63,6 +69,9 @@
     - handler_health_osd_check
     # See https://github.com/ceph/ceph-ansible/issues/1457 for the condition below
     - inventory_hostname in play_hosts
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"
 
 - name: restart ceph osds daemon(s) - container
   command: /tmp/restart_osd_daemon.sh
@@ -72,10 +81,13 @@
     # except when a crush location is specified. ceph-disk will start the osds before the osd crush location is specified
     - osd_group_name in group_names
     - containerized_deployment
-    - ((crush_location is defined and crush_location) or ceph_osd_container_stat.stdout_lines|length != 0)
+    - ((crush_location is defined and crush_location) or ceph_osd_container_stat.get('stdout_lines', [])|length != 0)
     - handler_health_osd_check
     # See https://github.com/ceph/ceph-ansible/issues/1457 for the condition below
     - inventory_hostname in play_hosts
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"
 
 - name: copy mds restart script
   template:
@@ -97,6 +109,9 @@
     - mds_group_name in group_names
     - not containerized_deployment
     - mds_socket_stat.rc == 0
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"
 
 - name: restart ceph mds daemon(s) - container
   command: /tmp/restart_mds_daemon.sh
@@ -105,7 +120,10 @@
     # We do not want to run these checks on initial deployment (`socket.rc == 0`)
     - mds_group_name in group_names
     - containerized_deployment
-    - ceph_mds_container_stat.stdout_lines|length != 0
+    - ceph_mds_container_stat.get('stdout_lines', [])|length != 0
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"
 
 - name: copy rgw restart script
   template:
@@ -127,6 +145,9 @@
     - rgw_group_name in group_names
     - not containerized_deployment
     - rgw_socket_stat.rc == 0
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"
 
 - name: restart ceph rgw daemon(s) - container
   command: /tmp/restart_rgw_daemon.sh
@@ -135,7 +156,10 @@
     # We do not want to run these checks on initial deployment (`socket.rc == 0`)
     - rgw_group_name in group_names
     - containerized_deployment
-    - ceph_rgw_container_stat.stdout_lines|length != 0
+    - ceph_rgw_container_stat.get('stdout_lines', [])|length != 0
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"
 
 - name: copy nfs restart script
   template:
@@ -157,6 +181,9 @@
     - nfs_group_name in group_names
     - not containerized_deployment
     - nfs_socket_stat.rc == 0
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"
 
 - name: restart ceph nfs daemon(s) - container
   command: /tmp/restart_nfs_daemon.sh
@@ -165,7 +192,10 @@
     # We do not want to run these checks on initial deployment (`socket.rc == 0`)
     - nfs_group_name in group_names
     - containerized_deployment
-    - ceph_nfs_container_stat.stdout_lines|length != 0
+    - ceph_nfs_container_stat.get('stdout_lines', [])|length != 0
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"
 
 - name: copy rbd mirror restart script
   template:
@@ -187,6 +217,9 @@
     - rbdmirror_group_name in group_names
     - not containerized_deployment
     - rbd_mirror_socket_stat.rc == 0
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"
 
 - name: restart ceph rbd mirror daemon(s) - container
   command: /tmp/restart_rbd_mirror_daemon.sh
@@ -195,7 +228,10 @@
     # We do not want to run these checks on initial deployment (`socket.rc == 0`)
     - rbdmirror_group_name in group_names
     - containerized_deployment
-    - ceph_rbd_mirror_container_stat.stdout_lines|length != 0
+    - ceph_rbd_mirror_container_stat.get('stdout_lines', [])|length != 0
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"
 
 - name: copy mgr restart script
   template:
@@ -217,6 +253,9 @@
     - mgr_group_name in group_names
     - not containerized_deployment
     - mgr_socket_stat.rc == 0
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"
 
 - name: restart ceph mgr daemon(s) - container
   command: /tmp/restart_mgr_daemon.sh
@@ -225,4 +264,7 @@
     # We do not want to run these checks on initial deployment (`socket.rc == 0`)
     - mgr_group_name in group_names
     - containerized_deployment
-    - ceph_mgr_container_stat.stdout_lines|length != 0
+    - ceph_mgr_container_stat.get('stdout_lines', [])|length != 0
+  with_items: "{{ ansible_play_batch }}"
+  run_once: true
+  delegate_to: "{{ item }}"

roles/ceph-defaults/templates/restart_osd_daemon.sh.j2

@@ -5,6 +5,10 @@ DELAY="{{ handler_health_osd_check_delay }}"
 CEPH_CLI="--name client.bootstrap-osd --keyring /var/lib/ceph/bootstrap-osd/{{ cluster }}.keyring --cluster {{ cluster }}"
 
 check_pgs() {
+  num_pgs=$($docker_exec ceph $CEPH_CLI -s -f json|python -c 'import sys, json; print(json.load(sys.stdin)["pgmap"]["num_pgs"])')
+  if [[ "$num_pgs" == "0" ]]; then
+    return 0
+  fi
   while [ $RETRIES -ne 0 ]; do
     test "[""$($docker_exec ceph $CEPH_CLI -s -f json | python -c 'import sys, json; print(json.load(sys.stdin)["pgmap"]["num_pgs"])')""]" = "$($docker_exec ceph $CEPH_CLI -s -f json | python -c 'import sys, json; print [ i["count"] for i in json.load(sys.stdin)["pgmap"]["pgs_by_state"] if i["state_name"] == "active+clean"]')"
     RET=$?

roles/ceph-osd/tasks/scenarios/collocated.yml

@@ -5,6 +5,7 @@
 - name: prepare ceph containerized osd disk collocated
   shell: |
     docker run --net=host \
+    --rm \
     --pid=host \
     --privileged=true \
     --name=ceph-osd-prepare-{{ ansible_hostname }}-{{ item.1 | regex_replace('/dev/', '') }} \
@@ -30,6 +31,7 @@
 - name: automatic prepare ceph containerized osd disk collocated
   shell: |
     docker run --net=host \
+    --rm \
     --pid=host \
     --privileged=true \
     --name=ceph-osd-prepare-{{ ansible_hostname }}-{{ item.key }} \

roles/ceph-osd/tasks/scenarios/non-collocated.yml

@@ -5,6 +5,7 @@
 - name: prepare ceph "{{ osd_objectstore }}" containerized osd disk(s) non-collocated
   shell: |
     docker run --net=host \
+    --rm \
     --pid=host \
     --privileged=true \
     --name=ceph-osd-prepare-{{ ansible_hostname }}-{{ item.1 | regex_replace('/dev/', '') }} \
@@ -32,6 +33,7 @@
 - name: prepare ceph "{{ osd_objectstore }}" containerized osd disk(s) non-collocated with a dedicated device for db and wal
   shell: |
     docker run --net=host \
+    --rm \
     --pid=host \
     --privileged=true \
     --name=ceph-osd-prepare-{{ ansible_hostname }}-{{ item.1 | regex_replace('/dev/', '') }} \

tests/functional/centos/7/bluestore/group_vars/all

@@ -10,7 +10,7 @@ cluster_network: "192.168.36.0/24"
 journal_size: 100
 osd_objectstore: "bluestore"
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1.0'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - '/dev/sdb'
 dedicated_devices:
   - '/dev/sdc'

tests/functional/centos/7/bs-crypt-ded-jrn/group_vars/all

@@ -11,7 +11,7 @@ osd_scenario: non-collocated
 dmcrypt: true
 osd_objectstore: "bluestore"
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1.0'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - '/dev/sdb'
 dedicated_devices:
   - '/dev/sdc'

tests/functional/centos/7/bs-crypt-jrn-col/group_vars/all

@@ -11,7 +11,7 @@ osd_scenario: collocated
 dmcrypt: true
 osd_objectstore: bluestore
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1.0'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - '/dev/sdb'
 os_tuning_params:
   - { name: kernel.pid_max, value: 4194303 }

tests/functional/centos/7/bs-dock-crypt-jrn-col/group_vars/all

@@ -11,7 +11,7 @@ osd_scenario: collocated
 dmcrypt: true
 osd_objectstore: bluestore
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1.0'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - '/dev/sdb'
 os_tuning_params:
   - { name: kernel.pid_max, value: 4194303 }

tests/functional/centos/7/bs-dock-ded-jrn/group_vars/all

@@ -16,7 +16,7 @@ ceph_rgw_civetweb_port: 8080
 osd_scenario: non-collocated
 osd_objectstore: bluestore
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1.0'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - /dev/sdb
 dedicated_devices:
   - /dev/sdc

tests/functional/centos/7/bs-jrn-col/group_vars/all

@@ -9,7 +9,7 @@ public_network: "192.168.3.0/24"
 cluster_network: "192.168.4.0/24"
 journal_size: 100
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1.0'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - '/dev/sdb'
 osd_scenario: collocated
 osd_objectstore: "bluestore"

tests/functional/centos/7/cluster/group_vars/all

@@ -9,7 +9,7 @@ journal_size: 100
 radosgw_interface: eth1
 osd_objectstore: filestore
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1.0'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - '/dev/sdb'
 dedicated_devices:
   - '/dev/sdc'

tests/functional/centos/7/crypt-ded-jrn/group_vars/all

@@ -11,7 +11,7 @@ osd_scenario: non-collocated
 dmcrypt: true
 osd_objectstore: filestore
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1.0'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - '/dev/sdb'
 dedicated_devices:
   - '/dev/sdc'

tests/functional/centos/7/crypt-jrn-col/group_vars/all

@@ -11,7 +11,7 @@ osd_scenario: collocated
 osd_objectstore: filestore
 dmcrypt: true
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1.0'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - '/dev/sdb'
 os_tuning_params:
   - { name: kernel.pid_max, value: 4194303 }

tests/functional/centos/7/docker-crypt-jrn-col/group_vars/all

@@ -17,6 +17,6 @@ osd_objectstore: filestore
 dmcrypt: true
 ceph_rgw_civetweb_port: 8080
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1.0'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - /dev/sdb
 ceph_osd_docker_prepare_env: -e OSD_JOURNAL_SIZE={{ journal_size }} -e OSD_FORCE_ZAP=1

tests/functional/centos/7/docker-ded-jrn/group_vars/all

@@ -16,7 +16,7 @@ ceph_rgw_civetweb_port: 8080
 osd_objectstore: filestore
 osd_scenario: non-collocated
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1.0'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - /dev/sdb
 dedicated_devices:
   - /dev/sdc

tests/functional/centos/7/docker/group_vars/all

@@ -17,7 +17,7 @@ ceph_rgw_civetweb_port: 8080
 osd_objectstore: filestore
 ceph_osd_docker_prepare_env: -e OSD_JOURNAL_SIZE={{ journal_size }} -e OSD_FORCE_ZAP=1
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1.0'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - /dev/sdb
 ceph_osd_docker_run_script_path: /var/tmp
 rgw_override_bucket_index_max_shards: 16

tests/functional/centos/7/jrn-col/group_vars/all

@@ -10,7 +10,7 @@ radosgw_interface: eth1
 journal_size: 100
 osd_objectstore: filestore
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1.0'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - '/dev/sdb'
 osd_scenario: collocated
 os_tuning_params:

tests/functional/ubuntu/16.04/cluster/group_vars/all

@@ -9,7 +9,7 @@ monitor_interface: eth1
 radosgw_interface: eth1
 journal_size: 100
 devices:
-  - '/dev/disk/by-path/pci-0000:00:01.1-ata-1'
+  - '/dev/disk/by-id/ata-QEMU_HARDDISK_QM00001'
   - '/dev/sdb'
 dedicated_devices:
   - '/dev/sdc'