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.

(cherry picked from commit 59a4335a5639c9be12ee8a23805aaa14882b077e)
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1548357

diff --git a/roles/ceph-defaults/handlers/main.yml b/roles/ceph-defaults/handlers/main.yml
index 289343976808700077b0904821f5f04281c27004..b3ea834bacea8463b68ddaaa404074f65a278db8 100644 (file)
--- a/roles/ceph-defaults/handlers/main.yml
+++ b/roles/ceph-defaults/handlers/main.yml
@@ -5,6 +5,14 @@
   when:
     - ansible_os_family == 'Debian'
 
+# 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"
+
 - name: copy mon restart script
   template:
     src: restart_mon_daemon.sh.j2
@@ -15,7 +23,6 @@
   listen: "restart ceph mons"
   when:
     - mon_group_name in group_names
-    - inventory_hostname in play_hosts
 
 - name: restart ceph mon daemon(s) - non container
   command: /tmp/restart_mon_daemon.sh
@@ -24,11 +31,11 @@
     # We do not want to run these checks on initial deployment (`socket.rc == 0`)
     - mon_group_name in group_names
     - not containerized_deployment
-    - inventory_hostname == groups.get(mon_group_name) | last
-    - inventory_hostname in play_hosts
+    - hostvars[item]['_mon_handler_called'] | default(False)
     - mon_socket_stat.rc == 0
   with_items: "{{ groups[mon_group_name] }}"
   delegate_to: "{{ item }}"
+  run_once: True
 
 - name: restart ceph mon daemon(s) - container
   command: /tmp/restart_mon_daemon.sh
@@ -38,11 +45,21 @@
     - mon_group_name in group_names
     - containerized_deployment
     - ceph_mon_container_stat.get('rc') == 0
-    - inventory_hostname == groups.get(mon_group_name) | last
-    - inventory_hostname in play_hosts
+    - hostvars[item]['_mon_handler_called'] | default(False)
     - ceph_mon_container_stat.get('stdout_lines', [])|length != 0
   with_items: "{{ groups[mon_group_name] }}"
   delegate_to: "{{ item }}"
+  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"
 
 # 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
@@ -57,7 +74,6 @@
   listen: "restart ceph osds"
   when:
     - osd_group_name in group_names
-    - inventory_hostname in play_hosts
 
 - name: restart ceph osds daemon(s) - non container
   command: /tmp/restart_osd_daemon.sh
@@ -65,16 +81,15 @@
   when:
     - osd_group_name in group_names
     - not containerized_deployment
-    - inventory_hostname == groups.get(osd_group_name) | last
     # 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
     - ((crush_location is defined and crush_location) or osd_socket_stat.rc == 0)
     - ceph_current_fsid.rc == 0
     - handler_health_osd_check
-    # See https://github.com/ceph/ceph-ansible/issues/1457 for the condition below
-    - inventory_hostname in play_hosts
+    - hostvars[item]['_osd_handler_called'] | default(False)
   with_items: "{{ groups[osd_group_name] }}"
   delegate_to: "{{ item }}"
+  run_once: True
 
 - name: restart ceph osds daemon(s) - container
   command: /tmp/restart_osd_daemon.sh
@@ -85,13 +100,22 @@
     - osd_group_name in group_names
     - containerized_deployment
     - ceph_osd_container_stat.get('rc') == 0
-    - inventory_hostname == groups.get(osd_group_name) | last
     - ((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
+    - hostvars[item]['_osd_handler_called'] | default(False)
   with_items: "{{ groups[osd_group_name] }}"
   delegate_to: "{{ item }}"
+  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"
 
 - name: copy mds restart script
   template:
@@ -103,7 +127,6 @@
   listen: "restart ceph mdss"
   when:
     - mds_group_name in group_names
-    - inventory_hostname in play_hosts
 
 - name: restart ceph mds daemon(s) - non container
   command: /tmp/restart_mds_daemon.sh
@@ -112,11 +135,11 @@
     # We do not want to run these checks on initial deployment (`socket.rc == 0`)
     - mds_group_name in group_names
     - not containerized_deployment
-    - inventory_hostname == groups.get(mds_group_name) | last
-    - inventory_hostname in play_hosts
+    - hostvars[item]['_mds_handler_called'] | default(False)
     - mds_socket_stat.rc == 0
   with_items: "{{ groups[mds_group_name] }}"
   delegate_to: "{{ item }}"
+  run_once: True
 
 - name: restart ceph mds daemon(s) - container
   command: /tmp/restart_mds_daemon.sh
@@ -126,11 +149,21 @@
     - mds_group_name in group_names
     - containerized_deployment
     - ceph_mds_container_stat.get('rc') == 0
-    - inventory_hostname == groups.get(mds_group_name) | last
-    - inventory_hostname in play_hosts
+    - hostvars[item]['_mds_handler_called'] | default(False)
     - ceph_mds_container_stat.get('stdout_lines', [])|length != 0
   with_items: "{{ groups[mds_group_name] }}"
   delegate_to: "{{ item }}"
+  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"
 
 - name: copy rgw restart script
   template:
@@ -142,7 +175,6 @@
   listen: "restart ceph rgws"
   when:
     - rgw_group_name in group_names
-    - inventory_hostname in play_hosts
 
 - name: restart ceph rgw daemon(s) - non container
   command: /tmp/restart_rgw_daemon.sh
@@ -151,11 +183,11 @@
     # We do not want to run these checks on initial deployment (`socket.rc == 0`)
     - rgw_group_name in group_names
     - not containerized_deployment
-    - inventory_hostname == groups.get(rgw_group_name) | last
-    - inventory_hostname in play_hosts
+    - hostvars[item]['_rgw_handler_called'] | default(False)
     - rgw_socket_stat.rc == 0
   with_items: "{{ groups[rgw_group_name] }}"
   delegate_to: "{{ item }}"
+  run_once: True
 
 - name: restart ceph rgw daemon(s) - container
   command: /tmp/restart_rgw_daemon.sh
@@ -165,11 +197,21 @@
     - rgw_group_name in group_names
     - containerized_deployment
     - ceph_rgw_container_stat.get('rc') == 0
-    - inventory_hostname == groups.get(rgw_group_name) | last
-    - inventory_hostname in play_hosts
+    - hostvars[item]['_rgw_handler_called'] | default(False)
     - ceph_rgw_container_stat.get('stdout_lines', [])|length != 0
   with_items: "{{ groups[rgw_group_name] }}"
   delegate_to: "{{ item }}"
+  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"
 
 - name: copy nfs restart script
   template:
@@ -181,7 +223,6 @@
   listen: "restart ceph nfss"
   when:
     - nfs_group_name in group_names
-    - inventory_hostname in play_hosts
 
 - name: restart ceph nfs daemon(s) - non container
   command: /tmp/restart_nfs_daemon.sh
@@ -190,11 +231,11 @@
     # We do not want to run these checks on initial deployment (`socket.rc == 0`)
     - nfs_group_name in group_names
     - not containerized_deployment
-    - inventory_hostname == groups.get(nfs_group_name) | last
-    - inventory_hostname in play_hosts
+    - hostvars[item]['_nfs_handler_called'] | default(False)
     - nfs_socket_stat.rc == 0
   with_items: "{{ groups[nfs_group_name] }}"
   delegate_to: "{{ item }}"
+  run_once: True
 
 - name: restart ceph nfs daemon(s) - container
   command: /tmp/restart_nfs_daemon.sh
@@ -204,11 +245,21 @@
     - nfs_group_name in group_names
     - containerized_deployment
     - ceph_nfs_container_stat.get('rc') == 0
-    - inventory_hostname == groups.get(nfs_group_name) | last
-    - inventory_hostname in play_hosts
+    - hostvars[item]['_nfs_handler_called'] | default(False)
     - ceph_nfs_container_stat.get('stdout_lines', [])|length != 0
   with_items: "{{ groups[nfs_group_name] }}"
   delegate_to: "{{ item }}"
+  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"
 
 - name: copy rbd mirror restart script
   template:
@@ -220,7 +271,6 @@
   listen: "restart ceph rbdmirrors"
   when:
     - rbdmirror_group_name in group_names
-    - inventory_hostname in play_hosts
 
 - name: restart ceph rbd mirror daemon(s) - non container
   command: /tmp/restart_rbd_mirror_daemon.sh
@@ -229,11 +279,11 @@
     # We do not want to run these checks on initial deployment (`socket.rc == 0`)
     - rbdmirror_group_name in group_names
     - not containerized_deployment
-    - inventory_hostname == groups.get(rbdmirror_group_name) | last
-    - inventory_hostname in play_hosts
+    - hostvars[item]['_rbdmirror_handler_called'] | default(False)
     - rbd_mirror_socket_stat.rc == 0
   with_items: "{{ groups[rbdmirror_group_name] }}"
   delegate_to: "{{ item }}"
+  run_once: True
 
 - name: restart ceph rbd mirror daemon(s) - container
   command: /tmp/restart_rbd_mirror_daemon.sh
@@ -243,11 +293,21 @@
     - rbdmirror_group_name in group_names
     - containerized_deployment
     - ceph_rbd_mirror_container_stat.get('rc') == 0
-    - inventory_hostname == groups.get(rbdmirror_group_name) | last
-    - inventory_hostname in play_hosts
+    - hostvars[item]['_rbdmirror_handler_called'] | default(False)
     - ceph_rbd_mirror_container_stat.get('stdout_lines', [])|length != 0
   with_items: "{{ groups[rbdmirror_group_name] }}"
   delegate_to: "{{ item }}"
+  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"
 
 - name: copy mgr restart script
   template:
@@ -259,7 +319,6 @@
   listen: "restart ceph mgrs"
   when:
     - mgr_group_name in group_names
-    - inventory_hostname in play_hosts
 
 - name: restart ceph mgr daemon(s) - non container
   command: /tmp/restart_mgr_daemon.sh
@@ -268,11 +327,11 @@
     # We do not want to run these checks on initial deployment (`socket.rc == 0`)
     - mgr_group_name in group_names
     - not containerized_deployment
-    - inventory_hostname == groups.get(mgr_group_name) | last
-    - inventory_hostname in play_hosts
+    - hostvars[item]['_mgr_handler_called'] | default(False)
     - mgr_socket_stat.rc == 0
   with_items: "{{ groups[mgr_group_name] }}"
   delegate_to: "{{ item }}"
+  run_once: True
 
 - name: restart ceph mgr daemon(s) - container
   command: /tmp/restart_mgr_daemon.sh
@@ -282,8 +341,13 @@
     - mgr_group_name in group_names
     - containerized_deployment
     - ceph_mgr_container_stat.get('rc') == 0
-    - inventory_hostname == groups.get(mgr_group_name) | last
-    - inventory_hostname in play_hosts
+    - hostvars[item]['_mgr_handler_called'] | default(False)
     - ceph_mgr_container_stat.get('stdout_lines', [])|length != 0
   with_items: "{{ groups[mgr_group_name] }}"
   delegate_to: "{{ item }}"
+  run_once: True
+
+- name: set _mgr_handler_called after restart
+  set_fact:
+     _mgr_handler_called: False
+  listen: "restart ceph mgrs"