Delivers upgrade scripts where upgrade steps are disabled

This delivers a /root/tripleo_upgrade_node.sh to those nodes
that have the disable_upgrade_deployment flag set to true.
They will later be upgraded manually by the operator who will
invoke the script delivered here using upgrade-non-controller.sh

We can also deliver any service specific upgrade configuration,
such as configuring nova-compute to use the placement API as this
is required in order for placement to be configured and installed
during the subsequent upgrade steps for controller services.

This removes the compute and swift specific upgrade scripts as
they are now merged into the common
tripleo_upgrade_node.sh - removing any hard coded
reference to a particular role name (compute/objectstorage) and
only relying on the disable_upgrade_deployment is roles_data.yaml

Change-Id: I4531a4038b78087ef4a1a62c35f1328822427817
Co-Authored-By: Mathieu Bultel <mbultel@redhat.com>

8 files changed, 119 insertions, 176 deletions

diff --git a/extraconfig/tasks/major_upgrade_compute.sh b/extraconfig/tasks/major_upgrade_compute.sh
deleted file mode 100644
index 7a3e1073..00000000
--- a/extraconfig/tasks/major_upgrade_compute.sh
+++ /dev/null
@@ -1,35 +0,0 @@
-#!/bin/bash
-#
-# This delivers the compute upgrade script to be invoked as part of the tripleo
-# major upgrade workflow.
-#
-set -eu
-
-UPGRADE_SCRIPT=/root/tripleo_upgrade_node.sh
-
-cat > $UPGRADE_SCRIPT << ENDOFCAT
-### DO NOT MODIFY THIS FILE
-### This file is automatically delivered to the compute nodes as part of the
-### tripleo upgrades workflow
-
-set -eu
-
-# pin nova to kilo (messaging +-1) for the nova-compute service
-
-crudini  --set /etc/nova/nova.conf upgrade_levels compute $upgrade_level_nova_compute
-
-$(declare -f special_case_ovs_upgrade_if_needed)
-special_case_ovs_upgrade_if_needed
-
-yum -y install python-zaqarclient  # needed for os-collect-config
-yum -y update
-
-# Due to bug#1640177 we need to restart compute agent
-echo "Restarting openstack ceilometer agent compute"
-systemctl restart openstack-ceilometer-compute
-
-ENDOFCAT
-
-# ensure the permissions are OK
-chmod 0755 $UPGRADE_SCRIPT
-
diff --git a/extraconfig/tasks/major_upgrade_object_storage.sh b/extraconfig/tasks/major_upgrade_object_storage.sh
deleted file mode 100644
index d9d1b4d5..00000000
--- a/extraconfig/tasks/major_upgrade_object_storage.sh
+++ /dev/null
@@ -1,42 +0,0 @@
-#!/bin/bash
-#
-# This delivers the swift-storage upgrade script to be invoked as part of the tripleo
-# major upgrade workflow.
-#
-set -eu
-
-UPGRADE_SCRIPT=/root/tripleo_upgrade_node.sh
-
-cat > $UPGRADE_SCRIPT << ENDOFCAT
-### DO NOT MODIFY THIS FILE
-### This file is automatically delivered to the swift-storage nodes as part of the
-### tripleo upgrades workflow
-
-set -eu
-
-function systemctl_swift {
-    action=\$1
-    for S in openstack-swift-account-auditor openstack-swift-account-reaper openstack-swift-account-replicator openstack-swift-account \
-             openstack-swift-container-auditor openstack-swift-container-replicator openstack-swift-container-updater openstack-swift-container \
-             openstack-swift-object-auditor openstack-swift-object-replicator openstack-swift-object-updater openstack-swift-object; do
-                systemctl \$action \$S
-    done
-}
-
-$(declare -f special_case_ovs_upgrade_if_needed)
-special_case_ovs_upgrade_if_needed
-
-systemctl_swift stop
-
-yum -y install python-zaqarclient  # needed for os-collect-config
-yum -y update
-
-systemctl_swift start
-
-
-
-ENDOFCAT
-
-# ensure the permissions are OK
-chmod 0755 $UPGRADE_SCRIPT
-
diff --git a/extraconfig/tasks/major_upgrade_pacemaker_init.j2.yaml b/extraconfig/tasks/major_upgrade_pacemaker_init.j2.yaml
deleted file mode 100644
index c308720b..00000000
--- a/extraconfig/tasks/major_upgrade_pacemaker_init.j2.yaml
+++ /dev/null
@@ -1,96 +0,0 @@
-heat_template_version: ocata
-description: 'Upgrade for Pacemaker deployments'
-
-parameters:
-
-  servers:
-    type: json
-  input_values:
-    type: json
-    description: input values for the software deployments
-
-  UpgradeInitCommand:
-    type: string
-    description: |
-      Command or script snippet to run on all overcloud nodes to
-      initialize the upgrade process. E.g. a repository switch.
-    default: ''
-  UpgradeLevelNovaCompute:
-    type: string
-    description: Nova Compute upgrade level
-    default: ''
-
-resources:
-
-  # For the UpgradeInit also rename /etc/resolv.conf.save for +bug/1567004
-
-  UpgradeInitConfig:
-    type: OS::Heat::SoftwareConfig
-    properties:
-      group: script
-      config:
-        list_join:
-        - ''
-        - - "#!/bin/bash\n\n"
-          - "if [[ -f /etc/resolv.conf.save ]] ; then rm /etc/resolv.conf.save; fi\n\n"
-          - get_param: UpgradeInitCommand
-
-  # TODO(jistr): for Mitaka->Newton upgrades and further we can use
-  # map_merge with input_values instead of feeding params into scripts
-  # via str_replace on bash snippets
-
-  # FIXME(shardy) we have hard-coded per-role *ScriptConfig's here
-  # Would be better to have a common config for all roles
-  ComputeDeliverUpgradeScriptConfig:
-    type: OS::Heat::SoftwareConfig
-    properties:
-      group: script
-      config:
-        list_join:
-        - ''
-        - - str_replace:
-              template: |
-                #!/bin/bash
-                upgrade_level_nova_compute='UPGRADE_LEVEL_NOVA_COMPUTE'
-              params:
-                UPGRADE_LEVEL_NOVA_COMPUTE: {get_param: UpgradeLevelNovaCompute}
-          - get_file: pacemaker_common_functions.sh
-          - get_file: major_upgrade_compute.sh
-
-  ObjectStorageDeliverUpgradeScriptConfig:
-    type: OS::Heat::SoftwareConfig
-    properties:
-      group: script
-      config:
-        list_join:
-        - ''
-        - - get_file: pacemaker_common_functions.sh
-          - get_file: major_upgrade_object_storage.sh
-
-  CephStorageDeliverUpgradeScriptConfig:
-    type: OS::Heat::SoftwareConfig
-    properties:
-      group: script
-      config:
-        list_join:
-        - ''
-        - - get_file: pacemaker_common_functions.sh
-          - get_file: major_upgrade_ceph_storage.sh
-
-{% for role in roles %}
-  UpgradeInit{{role.name}}Deployment:
-    type: OS::Heat::SoftwareDeploymentGroup
-    properties:
-      servers:  {get_param: [servers, {{role.name}}]}
-      config: {get_resource: UpgradeInitConfig}
-      input_values: {get_param: input_values}
-
-  {% if not role.name in ['Controller', 'BlockStorage'] %}
-  {{role.name}}DeliverUpgradeScriptDeployment:
-    type: OS::Heat::SoftwareDeploymentGroup
-    properties:
-      servers:  {get_param: [servers, {{role.name}}]}
-      config: {get_resource: {{role.name}}DeliverUpgradeScriptConfig}
-      input_values: {get_param: input_values}
-  {% endif %}
-{% endfor %}
diff --git a/extraconfig/tasks/tripleo_upgrade_node.sh b/extraconfig/tasks/tripleo_upgrade_node.sh
new file mode 100644
index 00000000..7f056021
--- /dev/null
+++ b/extraconfig/tasks/tripleo_upgrade_node.sh
@@ -0,0 +1,51 @@
+#!/bin/bash
+#
+# This delivers the operator driven upgrade script to be invoked as part of
+# the tripleo major upgrade workflow. The utility 'upgrade-non-controller.sh'
+# is used from the undercloud to invoke the /root/tripleo_upgrade_node.sh
+#
+set -eu
+
+UPGRADE_SCRIPT=/root/tripleo_upgrade_node.sh
+
+cat > $UPGRADE_SCRIPT << ENDOFCAT
+### DO NOT MODIFY THIS FILE
+### This file is automatically delivered to those nodes where the
+### disable_upgrade_deployment flag is set in roles_data.yaml.
+
+set -eu
+NOVA_COMPUTE=""
+if systemctl show 'openstack-nova-compute' --property ActiveState | grep '\bactive\b'; then
+   NOVA_COMPUTE="true"
+fi
+
+DEBUG="true"
+SCRIPT_NAME=$(basename $0)
+$(declare -f log_debug)
+$(declare -f manage_systemd_service)
+$(declare -f systemctl_swift)
+
+# pin nova messaging +-1 for the nova-compute service
+if [[ -n \$NOVA_COMPUTE ]]; then
+    crudini  --set /etc/nova/nova.conf upgrade_levels compute auto
+fi
+
+$(declare -f special_case_ovs_upgrade_if_needed)
+special_case_ovs_upgrade_if_needed
+
+yum -y install python-zaqarclient  # needed for os-collect-config
+systemctl_swift stop
+yum -y update
+systemctl_swift start
+
+# Due to bug#1640177 we need to restart compute agent
+if [[ -n \$NOVA_COMPUTE ]]; then
+    echo "Restarting openstack ceilometer agent compute"
+    systemctl restart openstack-ceilometer-compute
+fi
+
+ENDOFCAT
+
+# ensure the permissions are OK
+chmod 0755 $UPGRADE_SCRIPT
+
diff --git a/overcloud.j2.yaml b/overcloud.j2.yaml
index dea748ed..5b2ca4a2 100644
--- a/overcloud.j2.yaml
+++ b/overcloud.j2.yaml
@@ -610,6 +610,7 @@ resources:
 {% for role in roles %}
         {{role.name}}: {get_attr: [{{role.name}}, attributes, nova_server_resource]}
 {% endfor %}
+      EndpointMap: {get_attr: [EndpointMap, endpoint_map]}
       role_data:
 {% for role in roles %}
         {{role.name}}: {get_attr: [{{role.name}}ServiceChain, role_data]}
diff --git a/puppet/major_upgrade_steps.j2.yaml b/puppet/major_upgrade_steps.j2.yaml
index aa19dcb8..e0dac6a1 100644
--- a/puppet/major_upgrade_steps.j2.yaml
+++ b/puppet/major_upgrade_steps.j2.yaml
@@ -1,6 +1,7 @@
 {% set enabled_roles = roles|rejectattr('disable_upgrade_deployment')|list -%}
 {% set batch_upgrade_steps_max = 3 -%}
 {% set upgrade_steps_max = 6 -%}
+{% set deliver_script = {'deliver': False} -%}
 heat_template_version: ocata
 description: 'Upgrade steps for all roles'
 
@@ -17,6 +18,19 @@ parameters:
     description: >
       Setting to a previously unused value during stack-update will trigger
       the Upgrade resources to re-run on all roles.
+  EndpointMap:
+    default: {}
+    description: Mapping of service endpoint -> protocol. Typically set
+                 via parameter_defaults in the resource registry.
+    type: json
+  KeystoneRegion:
+    type: string
+    default: 'regionOne'
+    description: Keystone region for endpoint
+  NovaPassword:
+    description: The password for the nova service and db account, used by nova-api.
+    type: string
+    hidden: true
 
 conditions:
   # Conditions to disable any steps where the task list is empty
@@ -35,6 +49,43 @@ conditions:
 
 resources:
 
+{% for role in roles if role.disable_upgrade_deployment|default(false) %}
+  {{role.name}}DeliverUpgradeScriptConfig:
+    type: OS::Heat::SoftwareConfig
+    properties:
+      group: script
+      config:
+        list_join:
+        - ''
+        - - "#!/bin/bash\n\n"
+          - "set -eu\n\n"
+          - "if hiera -c /etc/puppet/hiera.yaml service_names | grep nova_compute ; then\n\n"
+          - "  crudini --set /etc/nova/nova.conf placement auth_type password\n\n"
+          - "  crudini --set /etc/nova/nova.conf placement username placement\n\n"
+          - "  crudini --set /etc/nova/nova.conf placement project_domain_name Default\n\n"
+          - "  crudini --set /etc/nova/nova.conf placement user_domain_name Default\n\n"
+          - "  crudini --set /etc/nova/nova.conf placement project_name service\n\n"
+          - "  systemctl restart openstack-nova-compute\n\n"
+          - "fi\n\n"
+          - str_replace:
+              template: |
+                crudini --set /etc/nova/nova.conf placement password 'SERVICE_PASSWORD'
+                crudini --set /etc/nova/nova.conf placement region_name 'REGION_NAME'
+                crudini --set /etc/nova/nova.conf placement auth_url 'AUTH_URL'
+              params:
+                SERVICE_PASSWORD: { get_param: NovaPassword }
+                REGION_NAME: { get_param: KeystoneRegion }
+                AUTH_URL: { get_param: [EndpointMap, KeystoneAdmin, uri_no_suffix]}
+          - get_file: ../extraconfig/tasks/pacemaker_common_functions.sh
+          - get_file: ../extraconfig/tasks/tripleo_upgrade_node.sh
+
+  {{role.name}}DeliverUpgradeScriptDeployment:
+    type: OS::Heat::SoftwareDeploymentGroup
+    properties:
+      servers:  {get_param: [servers, {{role.name}}]}
+      config: {get_resource: {{role.name}}DeliverUpgradeScriptConfig}
+{% endfor %}
+
 # Upgrade Steps for all roles, batched updates
 # The UpgradeConfig resources could actually be created without
 # serialization, but the event output is easier to follow if we
@@ -50,8 +101,18 @@ resources:
     {% if role.name in enabled_roles %}
     depends_on:
       - {{role.name}}UpgradeBatch_Step{{step -1}}
+    {%- endif %}
+  {% else %}
+    {% for role in roles if role.disable_upgrade_deployment|default(false) %}
+      {% if deliver_script.update({'deliver': True}) %} {% endif %}
+    {% endfor %}
+    {% if deliver_script.deliver %}
+    depends_on:
     {% endif %}
-  {%- endif %}
+      {% for dep in roles if dep.disable_upgrade_deployment|default(false) %}
+      - {{dep.name}}DeliverUpgradeScriptDeployment
+      {% endfor %}
+  {% endif %}
     properties:
       UpgradeStepConfig: {get_param: [role_data, {{role.name}}, upgrade_batch_tasks]}
       step: {{step}}
diff --git a/puppet/post.j2.yaml b/puppet/post.j2.yaml
index 39155c36..21202775 100644
--- a/puppet/post.j2.yaml
+++ b/puppet/post.j2.yaml
@@ -12,7 +12,11 @@ parameters:
   role_data:
     type: json
     description: Mapping of Role name e.g Controller to the per-role data
-
+  EndpointMap:
+    default: {}
+    description: Mapping of service endpoint -> protocol. Typically set
+                 via parameter_defaults in the resource registry.
+    type: json
   DeployIdentifier:
     default: ''
     type: string
diff --git a/roles_data.yaml b/roles_data.yaml
index 4edefcee..92c5ff19 100644
--- a/roles_data.yaml
+++ b/roles_data.yaml
@@ -184,7 +184,6 @@
     - OS::TripleO::Services::Collectd
 
 - name: CephStorage
-  disable_upgrade_deployment: True
   ServicesDefault:
     - OS::TripleO::Services::CACerts
     - OS::TripleO::Services::CephOSD