ceph
/
ceph-ansible
mirror of https://github.com/ceph/ceph-ansible.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
ceph-ansible/roles/ceph-defaults/handlers/main.yml

361 lines
12 KiB
YAML
Raw Normal View History

Add Ceph Playbook Initial commit. Signed-off-by: Sébastien Han <sebastien.han@enovance.com>
2014-03-04 02:08:51 +08:00
---
Sweep of all sorts of things.
2014-09-05 03:14:11 +08:00
- name: update apt cache
Refactor ceph-common, adjust styling
2015-09-04 00:18:53 +08:00
apt:
update-cache: yes
name includes and set_fact for clarity When Ansible is not run with verbose options it's difficult to see which include and/or set_fact does what. So adding a name for each clarifies. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-15 06:48:53 +08:00
when:
- ansible_os_family == 'Debian'
Add Ceph Playbook Initial commit. Signed-off-by: Sébastien Han <sebastien.han@enovance.com>
2014-03-04 02:08:51 +08:00
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
# We only want to restart on hosts that have called the handler.
# This var is set when he handler is called, and unset after the
# restart to ensure only the correct hosts are restarted.
- name: set _mon_handler_called before restart
set_fact:
_mon_handler_called: True
listen: "restart ceph mons"
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: copy mon restart script
template:
src: restart_mon_daemon.sh.j2
dest: /tmp/restart_mon_daemon.sh
owner: root
group: root
mode: 0750
listen: "restart ceph mons"
when:
- mon_group_name in group_names
- name: restart ceph mon daemon(s) - non container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_mon_daemon.sh
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
listen: "restart ceph mons"
Common: Fix handlers that are not properly triggered. Until now, only the first task were executed. The idea here is to use `listen` statement to be able to notify multiple handler and regroup all of them in `./handlers/main.yml` as notifying an included handler task is not possible. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-04-04 01:55:11 +08:00
when:
ceph-defaults: fix handlers for mds and rgw The way we handle the restart for both mds and rgw is not ideal, it will try to restart the daemon on the host that don't run the daemon, resulting in a service file being created (see bug description). Now we restart each daemon precisely and in a serialized fashion. Note: the current implementation does NOT support multiple mds or rgw on the same node. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1469781 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-08-31 17:22:33 +08:00
# We do not want to run these checks on initial deployment (`socket.rc == 0`)
Common: Fix handlers that are not properly triggered. Until now, only the first task were executed. The idea here is to use `listen` statement to be able to notify multiple handler and regroup all of them in `./handlers/main.yml` as notifying an included handler task is not possible. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-04-04 01:55:11 +08:00
- mon_group_name in group_names
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- not containerized_deployment
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_mon_handler_called'] | default(False)
defaults: only run socket checks on their specific roles Running the socket check on all the hosts will override the default value of docker_exec_cmd, leaving it with the last value (currently rbd-mirror), as a result the subsequent docker_exec_cmd usage for the :x Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 06:55:31 +08:00
- mon_socket_stat.rc == 0
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[mon_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
Common: Fix handlers that are not properly triggered. Until now, only the first task were executed. The idea here is to use `listen` statement to be able to notify multiple handler and regroup all of them in `./handlers/main.yml` as notifying an included handler task is not possible. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-04-04 01:55:11 +08:00
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: restart ceph mon daemon(s) - container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_mon_daemon.sh
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
listen: "restart ceph mons"
when:
# We do not want to run these checks on initial deployment (`socket.rc == 0`)
- mon_group_name in group_names
- containerized_deployment
handlers: restart daemons only if docker is running In case where docker CLI is available but docker is not running, we don't want to trigger the restart of the daemons. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1510555 Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-11-27 21:59:30 +08:00
- ceph_mon_container_stat.get('rc') == 0
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_mon_handler_called'] | default(False)
use get to check stdout_lines During the initial play, the docker command doesn't not exist and then there is no stdout_lines to the command. So get allows us to fix this by declaring an array if the command fails. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-06 22:49:46 +08:00
- ceph_mon_container_stat.get('stdout_lines', [])|length != 0
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[mon_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
- name: set _mon_handler_called after restart
set_fact:
_mon_handler_called: False
listen: "restart ceph mons"
- name: set _osd_handler_called before restart
set_fact:
_osd_handler_called: True
listen: "restart ceph osds"
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
Common: Fix handlers that are not properly triggered. Until now, only the first task were executed. The idea here is to use `listen` statement to be able to notify multiple handler and regroup all of them in `./handlers/main.yml` as notifying an included handler task is not possible. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-04-04 01:55:11 +08:00
# 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
# for restarting them specifically.
Fix restarting OSDs twice during a rolling update. During a rolling update, OSDs are restarted twice currently. Once, by the handler in roles/ceph-defaults/handlers/main.yml and a second time by tasks in the rolling_update playbook. This change turns off restarts by the handler. Further, the restart initiated by the rolling_update playbook is more efficient as it restarts all the OSDs on a host as one operation and waits for them to rejoin the cluster. The restart task in the handler restarts one OSD at a time and waits for it to join the cluster.
2018-03-17 01:10:14 +08:00
# This does not need to run during a rolling update as the playbook will
# restart all OSDs using the tasks "start ceph osd" or
# "restart containerized ceph osd"
Add handlers for containerized deployment Until now, there is no handlers for containerized deployments. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-07-29 07:00:06 +08:00
- name: copy osd restart script
template:
src: restart_osd_daemon.sh.j2
dest: /tmp/restart_osd_daemon.sh
owner: root
group: root
mode: 0750
listen: "restart ceph osds"
when:
- osd_group_name in group_names
Fix restarting OSDs twice during a rolling update. During a rolling update, OSDs are restarted twice currently. Once, by the handler in roles/ceph-defaults/handlers/main.yml and a second time by tasks in the rolling_update playbook. This change turns off restarts by the handler. Further, the restart initiated by the rolling_update playbook is more efficient as it restarts all the OSDs on a host as one operation and waits for them to rejoin the cluster. The restart task in the handler restarts one OSD at a time and waits for it to join the cluster.
2018-03-17 01:10:14 +08:00
- not rolling_update
Common: Fix handlers that are not properly triggered. Until now, only the first task were executed. The idea here is to use `listen` statement to be able to notify multiple handler and regroup all of them in `./handlers/main.yml` as notifying an included handler task is not possible. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-04-04 01:55:11 +08:00
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: restart ceph osds daemon(s) - non container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_osd_daemon.sh
Add handlers for containerized deployment Until now, there is no handlers for containerized deployments. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-07-29 07:00:06 +08:00
listen: "restart ceph osds"
when:
defaults: only run socket checks on their specific roles Running the socket check on all the hosts will override the default value of docker_exec_cmd, leaving it with the last value (currently rbd-mirror), as a result the subsequent docker_exec_cmd usage for the :x Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 06:55:31 +08:00
- osd_group_name in group_names
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- not containerized_deployment
Fix restarting OSDs twice during a rolling update. During a rolling update, OSDs are restarted twice currently. Once, by the handler in roles/ceph-defaults/handlers/main.yml and a second time by tasks in the rolling_update playbook. This change turns off restarts by the handler. Further, the restart initiated by the rolling_update playbook is more efficient as it restarts all the OSDs on a host as one operation and waits for them to rejoin the cluster. The restart task in the handler restarts one OSD at a time and waits for it to join the cluster.
2018-03-17 01:10:14 +08:00
- not rolling_update
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
# We do not want to run these checks on initial deployment (`socket_osd_container.results[n].rc == 0`)
# except when a crush location is specified. ceph-disk will start the osds before the osd crush location is specified
osd: remove old crush_location implementation This was causing a lot of pain with the handlers. Also the implementation was not ideal since we were assembling files. Everything can now be done with the ceph_crush module so let's remove that. Signed-off-by: Sébastien Han <seb@redhat.com>
2018-02-21 22:56:32 +08:00
- osd_socket_stat.rc == 0
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- ceph_current_fsid.rc == 0
Add handlers for containerized deployment Until now, there is no handlers for containerized deployments. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-07-29 07:00:06 +08:00
- handler_health_osd_check
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_osd_handler_called'] | default(False)
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[osd_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
Common: Restore check_socket Restore the check_socket that was removed by `5bec62b`. This commit also improves the logging in `restart_*_daemon.sh` scripts Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-04-18 18:40:43 +08:00
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: restart ceph osds daemon(s) - container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_osd_daemon.sh
Add handlers for containerized deployment Until now, there is no handlers for containerized deployments. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-07-29 07:00:06 +08:00
listen: "restart ceph osds"
Common: Fix handlers that are not properly triggered. Until now, only the first task were executed. The idea here is to use `listen` statement to be able to notify multiple handler and regroup all of them in `./handlers/main.yml` as notifying an included handler task is not possible. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-04-04 01:55:11 +08:00
when:
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
# We do not want to run these checks on initial deployment (`socket_osd_container_stat.results[n].rc == 0`)
ceph-defaults: fix handlers for mds and rgw The way we handle the restart for both mds and rgw is not ideal, it will try to restart the daemon on the host that don't run the daemon, resulting in a service file being created (see bug description). Now we restart each daemon precisely and in a serialized fashion. Note: the current implementation does NOT support multiple mds or rgw on the same node. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1469781 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-08-31 17:22:33 +08:00
# except when a crush location is specified. ceph-disk will start the osds before the osd crush location is specified
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- osd_group_name in group_names
- containerized_deployment
Fix restarting OSDs twice during a rolling update. During a rolling update, OSDs are restarted twice currently. Once, by the handler in roles/ceph-defaults/handlers/main.yml and a second time by tasks in the rolling_update playbook. This change turns off restarts by the handler. Further, the restart initiated by the rolling_update playbook is more efficient as it restarts all the OSDs on a host as one operation and waits for them to rejoin the cluster. The restart task in the handler restarts one OSD at a time and waits for it to join the cluster.
2018-03-17 01:10:14 +08:00
- not rolling_update
handlers: restart daemons only if docker is running In case where docker CLI is available but docker is not running, we don't want to trigger the restart of the daemons. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1510555 Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-11-27 21:59:30 +08:00
- ceph_osd_container_stat.get('rc') == 0
osd: remove old crush_location implementation This was causing a lot of pain with the handlers. Also the implementation was not ideal since we were assembling files. Everything can now be done with the ceph_crush module so let's remove that. Signed-off-by: Sébastien Han <seb@redhat.com>
2018-02-21 22:56:32 +08:00
- inventory_hostname == groups.get(osd_group_name) | last
- ceph_osd_container_stat.get('stdout_lines', [])|length != 0
Add handlers for containerized deployment Until now, there is no handlers for containerized deployments. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-07-29 07:00:06 +08:00
- handler_health_osd_check
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_osd_handler_called'] | default(False)
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[osd_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
- name: set _osd_handler_called after restart
set_fact:
_osd_handler_called: False
listen: "restart ceph osds"
- name: set _mds_handler_called before restart
set_fact:
_mds_handler_called: True
listen: "restart ceph mdss"
Improve handlers Signed-off-by: leseb <seb@redhat.com>
2015-07-01 00:07:03 +08:00
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: copy mds restart script
template:
src: restart_mds_daemon.sh.j2
dest: /tmp/restart_mds_daemon.sh
owner: root
group: root
mode: 0750
listen: "restart ceph mdss"
when:
- mds_group_name in group_names
- name: restart ceph mds daemon(s) - non container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_mds_daemon.sh
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
listen: "restart ceph mdss"
ceph-defaults: fix handlers for mds and rgw The way we handle the restart for both mds and rgw is not ideal, it will try to restart the daemon on the host that don't run the daemon, resulting in a service file being created (see bug description). Now we restart each daemon precisely and in a serialized fashion. Note: the current implementation does NOT support multiple mds or rgw on the same node. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1469781 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-08-31 17:22:33 +08:00
when:
# We do not want to run these checks on initial deployment (`socket.rc == 0`)
- mds_group_name in group_names
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- not containerized_deployment
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_mds_handler_called'] | default(False)
defaults: only run socket checks on their specific roles Running the socket check on all the hosts will override the default value of docker_exec_cmd, leaving it with the last value (currently rbd-mirror), as a result the subsequent docker_exec_cmd usage for the :x Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 06:55:31 +08:00
- mds_socket_stat.rc == 0
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[mds_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
ceph-defaults: fix handlers for mds and rgw The way we handle the restart for both mds and rgw is not ideal, it will try to restart the daemon on the host that don't run the daemon, resulting in a service file being created (see bug description). Now we restart each daemon precisely and in a serialized fashion. Note: the current implementation does NOT support multiple mds or rgw on the same node. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1469781 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-08-31 17:22:33 +08:00
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: restart ceph mds daemon(s) - container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_mds_daemon.sh
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
listen: "restart ceph mdss"
when:
# We do not want to run these checks on initial deployment (`socket.rc == 0`)
- mds_group_name in group_names
- containerized_deployment
handlers: restart daemons only if docker is running In case where docker CLI is available but docker is not running, we don't want to trigger the restart of the daemons. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1510555 Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-11-27 21:59:30 +08:00
- ceph_mds_container_stat.get('rc') == 0
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_mds_handler_called'] | default(False)
use get to check stdout_lines During the initial play, the docker command doesn't not exist and then there is no stdout_lines to the command. So get allows us to fix this by declaring an array if the command fails. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-06 22:49:46 +08:00
- ceph_mds_container_stat.get('stdout_lines', [])|length != 0
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[mds_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
- name: set _mds_handler_called after restart
set_fact:
_mds_handler_called: False
listen: "restart ceph mdss"
- name: set _rgw_handler_called before restart
set_fact:
_rgw_handler_called: True
listen: "restart ceph rgws"
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
ceph-defaults: fix handlers for mds and rgw The way we handle the restart for both mds and rgw is not ideal, it will try to restart the daemon on the host that don't run the daemon, resulting in a service file being created (see bug description). Now we restart each daemon precisely and in a serialized fashion. Note: the current implementation does NOT support multiple mds or rgw on the same node. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1469781 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-08-31 17:22:33 +08:00
- name: copy rgw restart script
template:
src: restart_rgw_daemon.sh.j2
dest: /tmp/restart_rgw_daemon.sh
owner: root
group: root
mode: 0750
listen: "restart ceph rgws"
Common: Fix handlers that are not properly triggered. Until now, only the first task were executed. The idea here is to use `listen` statement to be able to notify multiple handler and regroup all of them in `./handlers/main.yml` as notifying an included handler task is not possible. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-04-04 01:55:11 +08:00
when:
ceph-defaults: fix handlers for mds and rgw The way we handle the restart for both mds and rgw is not ideal, it will try to restart the daemon on the host that don't run the daemon, resulting in a service file being created (see bug description). Now we restart each daemon precisely and in a serialized fashion. Note: the current implementation does NOT support multiple mds or rgw on the same node. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1469781 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-08-31 17:22:33 +08:00
- rgw_group_name in group_names
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: restart ceph rgw daemon(s) - non container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_rgw_daemon.sh
ceph-defaults: fix handlers for mds and rgw The way we handle the restart for both mds and rgw is not ideal, it will try to restart the daemon on the host that don't run the daemon, resulting in a service file being created (see bug description). Now we restart each daemon precisely and in a serialized fashion. Note: the current implementation does NOT support multiple mds or rgw on the same node. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1469781 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-08-31 17:22:33 +08:00
listen: "restart ceph rgws"
when:
# We do not want to run these checks on initial deployment (`socket.rc == 0`)
Common: Fix handlers that are not properly triggered. Until now, only the first task were executed. The idea here is to use `listen` statement to be able to notify multiple handler and regroup all of them in `./handlers/main.yml` as notifying an included handler task is not possible. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-04-04 01:55:11 +08:00
- rgw_group_name in group_names
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- not containerized_deployment
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_rgw_handler_called'] | default(False)
defaults: only run socket checks on their specific roles Running the socket check on all the hosts will override the default value of docker_exec_cmd, leaving it with the last value (currently rbd-mirror), as a result the subsequent docker_exec_cmd usage for the :x Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 06:55:31 +08:00
- rgw_socket_stat.rc == 0
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[rgw_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
Remove rgw installation from the ceph-rgw role The installation of rgw is now handled by the ceph-common role. Fixes: #307 Signed-off-by: Sébastien Han <seb@redhat.com>
2015-08-03 23:39:57 +08:00
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: restart ceph rgw daemon(s) - container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_rgw_daemon.sh
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
listen: "restart ceph rgws"
when:
# We do not want to run these checks on initial deployment (`socket.rc == 0`)
- rgw_group_name in group_names
- containerized_deployment
handlers: restart daemons only if docker is running In case where docker CLI is available but docker is not running, we don't want to trigger the restart of the daemons. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1510555 Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-11-27 21:59:30 +08:00
- ceph_rgw_container_stat.get('rc') == 0
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_rgw_handler_called'] | default(False)
use get to check stdout_lines During the initial play, the docker command doesn't not exist and then there is no stdout_lines to the command. So get allows us to fix this by declaring an array if the command fails. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-06 22:49:46 +08:00
- ceph_rgw_container_stat.get('stdout_lines', [])|length != 0
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[rgw_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
- name: set _rgw_handler_called after restart
set_fact:
_rgw_handler_called: False
listen: "restart ceph rgws"
- name: set _nfs_handler_called before restart
set_fact:
_nfs_handler_called: True
listen: "restart ceph nfss"
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
defaults: only run socket checks on their specific roles Running the socket check on all the hosts will override the default value of docker_exec_cmd, leaving it with the last value (currently rbd-mirror), as a result the subsequent docker_exec_cmd usage for the :x Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 06:55:31 +08:00
- name: copy nfs restart script
template:
src: restart_nfs_daemon.sh.j2
dest: /tmp/restart_nfs_daemon.sh
owner: root
group: root
mode: 0750
listen: "restart ceph nfss"
when:
- nfs_group_name in group_names
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: restart ceph nfs daemon(s) - non container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_nfs_daemon.sh
defaults: only run socket checks on their specific roles Running the socket check on all the hosts will override the default value of docker_exec_cmd, leaving it with the last value (currently rbd-mirror), as a result the subsequent docker_exec_cmd usage for the :x Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 06:55:31 +08:00
listen: "restart ceph nfss"
Add support for Ceph NFS Gateway Ceph has the ability to export it's filesystem via NFS using Ganesha. Add a ceph-nfs role that will start Ganesha and export the Ceph filesystems. Note that, although support is going in to export RGW via NFS, this is not working yet. Signed-off-by: Daniel Gryniewicz <dang@redhat.com>
2016-05-06 02:20:03 +08:00
when:
defaults: only run socket checks on their specific roles Running the socket check on all the hosts will override the default value of docker_exec_cmd, leaving it with the last value (currently rbd-mirror), as a result the subsequent docker_exec_cmd usage for the :x Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 06:55:31 +08:00
# We do not want to run these checks on initial deployment (`socket.rc == 0`)
Add support for Ceph NFS Gateway Ceph has the ability to export it's filesystem via NFS using Ganesha. Add a ceph-nfs role that will start Ganesha and export the Ceph filesystems. Note that, although support is going in to export RGW via NFS, this is not working yet. Signed-off-by: Daniel Gryniewicz <dang@redhat.com>
2016-05-06 02:20:03 +08:00
- nfs_group_name in group_names
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- not containerized_deployment
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_nfs_handler_called'] | default(False)
defaults: only run socket checks on their specific roles Running the socket check on all the hosts will override the default value of docker_exec_cmd, leaving it with the last value (currently rbd-mirror), as a result the subsequent docker_exec_cmd usage for the :x Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 06:55:31 +08:00
- nfs_socket_stat.rc == 0
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[nfs_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
defaults: add missing handlers for rbd mirorr and mgr Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 08:08:40 +08:00
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: restart ceph nfs daemon(s) - container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_nfs_daemon.sh
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
listen: "restart ceph nfss"
when:
# We do not want to run these checks on initial deployment (`socket.rc == 0`)
- nfs_group_name in group_names
- containerized_deployment
handlers: restart daemons only if docker is running In case where docker CLI is available but docker is not running, we don't want to trigger the restart of the daemons. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1510555 Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-11-27 21:59:30 +08:00
- ceph_nfs_container_stat.get('rc') == 0
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_nfs_handler_called'] | default(False)
use get to check stdout_lines During the initial play, the docker command doesn't not exist and then there is no stdout_lines to the command. So get allows us to fix this by declaring an array if the command fails. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-06 22:49:46 +08:00
- ceph_nfs_container_stat.get('stdout_lines', [])|length != 0
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[nfs_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
- name: set _nfs_handler_called after restart
set_fact:
_nfs_handler_called: False
listen: "restart ceph nfss"
- name: set _rbdmirror_handler_called before restart
set_fact:
_rbdmirror_handler_called: True
listen: "restart ceph rbdmirrors"
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
defaults: add missing handlers for rbd mirorr and mgr Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 08:08:40 +08:00
- name: copy rbd mirror restart script
template:
src: restart_rbd_mirror_daemon.sh.j2
dest: /tmp/restart_rbd_mirror_daemon.sh
owner: root
group: root
mode: 0750
listen: "restart ceph rbdmirrors"
when:
- rbdmirror_group_name in group_names
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: restart ceph rbd mirror daemon(s) - non container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_rbd_mirror_daemon.sh
defaults: add missing handlers for rbd mirorr and mgr Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 08:08:40 +08:00
listen: "restart ceph rbdmirrors"
when:
# We do not want to run these checks on initial deployment (`socket.rc == 0`)
- rbdmirror_group_name in group_names
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- not containerized_deployment
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_rbdmirror_handler_called'] | default(False)
defaults: add missing handlers for rbd mirorr and mgr Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 08:08:40 +08:00
- rbd_mirror_socket_stat.rc == 0
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[rbdmirror_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
defaults: add missing handlers for rbd mirorr and mgr Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 08:08:40 +08:00
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: restart ceph rbd mirror daemon(s) - container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_rbd_mirror_daemon.sh
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
listen: "restart ceph rbdmirrors"
when:
# We do not want to run these checks on initial deployment (`socket.rc == 0`)
- rbdmirror_group_name in group_names
- containerized_deployment
handlers: restart daemons only if docker is running In case where docker CLI is available but docker is not running, we don't want to trigger the restart of the daemons. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1510555 Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-11-27 21:59:30 +08:00
- ceph_rbd_mirror_container_stat.get('rc') == 0
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_rbdmirror_handler_called'] | default(False)
use get to check stdout_lines During the initial play, the docker command doesn't not exist and then there is no stdout_lines to the command. So get allows us to fix this by declaring an array if the command fails. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-06 22:49:46 +08:00
- ceph_rbd_mirror_container_stat.get('stdout_lines', [])|length != 0
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[rbdmirror_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
- name: set _rbdmirror_handler_called after restart
set_fact:
_rbdmirror_handler_called: False
listen: "restart ceph rbdmirrors"
- name: set _mgr_handler_called before restart
set_fact:
_mgr_handler_called: True
listen: "restart ceph mgrs"
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
defaults: add missing handlers for rbd mirorr and mgr Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 08:08:40 +08:00
- name: copy mgr restart script
template:
src: restart_mgr_daemon.sh.j2
dest: /tmp/restart_mgr_daemon.sh
owner: root
group: root
mode: 0750
listen: "restart ceph mgrs"
when:
- mgr_group_name in group_names
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: restart ceph mgr daemon(s) - non container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_mgr_daemon.sh
defaults: add missing handlers for rbd mirorr and mgr Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 08:08:40 +08:00
listen: "restart ceph mgrs"
when:
# We do not want to run these checks on initial deployment (`socket.rc == 0`)
- mgr_group_name in group_names
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- not containerized_deployment
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_mgr_handler_called'] | default(False)
defaults: add missing handlers for rbd mirorr and mgr Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-27 08:08:40 +08:00
- mgr_socket_stat.rc == 0
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[mgr_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
- name: restart ceph mgr daemon(s) - container
FIX: run restart scripts in `noexec` /tmp - One can not run scripts directly in place, that mounted with `noexec` option. But one can run scripts as arguments for `bash/sh`. Signed-off-by: Arano-kai <captcha.is.evil@gmail.com>
2017-11-06 22:02:47 +08:00
command: /usr/bin/env bash /tmp/restart_mgr_daemon.sh
ci: add collocation scenario Signed-off-by: Sébastien Han <seb@redhat.com>
2017-09-28 00:22:15 +08:00
listen: "restart ceph mgrs"
when:
# We do not want to run these checks on initial deployment (`socket.rc == 0`)
- mgr_group_name in group_names
- containerized_deployment
handlers: restart daemons only if docker is running In case where docker CLI is available but docker is not running, we don't want to trigger the restart of the daemons. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=1510555 Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
2017-11-27 21:59:30 +08:00
- ceph_mgr_container_stat.get('rc') == 0
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
- hostvars[item]['_mgr_handler_called'] | default(False)
use get to check stdout_lines During the initial play, the docker command doesn't not exist and then there is no stdout_lines to the command. So get allows us to fix this by declaring an array if the command fails. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-06 22:49:46 +08:00
- ceph_mgr_container_stat.get('stdout_lines', [])|length != 0
defaults: fix handlers for collocation When doing collocation the condition "inventory_hostname in play_hosts" is breaking the restart workflow. Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-17 21:54:17 +08:00
with_items: "{{ groups[mgr_group_name] }}"
handler: add serial restart back We now restart daemons on each machine in a serialized fashion. Closes: https://github.com/ceph/ceph-ansible/issues/1989 Signed-off-by: Sébastien Han <seb@redhat.com>
2017-10-07 07:23:30 +08:00
delegate_to: "{{ item }}"
Restart services if handler called This patch fixes an issue where if hosts have different service lists, it will prevent restarting changes on services that run later on. For example, hostA in the mons and rgws group would initiate a config change and restart of services on all mons and rgws hosts, even though a separate hostB (which is only in the rgws group) has not had its configuration changed yet. Additionally, when the second host has its coniguration changed as part of the ceph-rgw role, it will not initiate a restart since its inventory name != the first hosts. To fix this we should run the restart once (using run_once: True) as long as the host has called the handler. This will ensure that even if only 1 host has called the handler it will initiate a restart on all hosts that have called the handler. Additionally, we add a var that is set when the handler runs, this will ensure that only hosts that have called the handler get restarted. Includes minor fix to remove unrequired "inventory_hostname in play_hosts" when: clause. This is no longer required since the handlers were changed. The host calling the handler will be in play_hosts already.
2017-12-20 11:49:16 +08:00
run_once: True
- name: set _mgr_handler_called after restart
set_fact:
_mgr_handler_called: False
listen: "restart ceph mgrs"