diff options
| -rwxr-xr-x | environments/neutron-sriov.yaml | 3 | ||||
| -rw-r--r-- | extraconfig/tasks/major_upgrade_block_storage.sh | 11 | ||||
| -rwxr-xr-x | extraconfig/tasks/major_upgrade_ceph_mon.sh | 82 | ||||
| -rw-r--r-- | extraconfig/tasks/major_upgrade_ceph_storage.sh | 106 | ||||
| -rw-r--r-- | extraconfig/tasks/major_upgrade_compute.sh | 35 | ||||
| -rw-r--r-- | extraconfig/tasks/major_upgrade_object_storage.sh | 42 | ||||
| -rw-r--r-- | extraconfig/tasks/major_upgrade_pacemaker.yaml | 48 | ||||
| -rw-r--r-- | extraconfig/tasks/major_upgrade_pacemaker_init.j2.yaml | 96 | ||||
| -rw-r--r-- | extraconfig/tasks/tripleo_upgrade_node.sh | 54 | ||||
| -rw-r--r-- | overcloud.j2.yaml | 1 | ||||
| -rw-r--r-- | puppet/major_upgrade_steps.j2.yaml | 135 | ||||
| -rw-r--r-- | puppet/post.j2.yaml | 6 | ||||
| -rw-r--r-- | puppet/services/aodh-api.yaml | 3 | ||||
| -rw-r--r-- | puppet/services/gnocchi-api.yaml | 3 | ||||
| -rw-r--r-- | puppet/services/neutron-plugin-ml2.yaml | 7 | ||||
| -rw-r--r-- | puppet/services/pacemaker.yaml | 18 | ||||
| -rw-r--r-- | roles_data.yaml | 1 |
17 files changed, 187 insertions, 464 deletions
diff --git a/environments/neutron-sriov.yaml b/environments/neutron-sriov.yaml index 9b7e51f9..5e9e15e3 100755 --- a/environments/neutron-sriov.yaml +++ b/environments/neutron-sriov.yaml @@ -9,9 +9,6 @@ parameter_defaults: #NovaSchedulerDefaultFilters: ['RetryFilter','AvailabilityZoneFilter','RamFilter','ComputeFilter','ComputeCapabilitiesFilter','ImagePropertiesFilter','ServerGroupAntiAffinityFilter','ServerGroupAffinityFilter','PciPassthroughFilter'] #NovaSchedulerAvailableFilters: ["nova.scheduler.filters.all_filters","nova.scheduler.filters.pci_passthrough_filter.PciPassthroughFilter"] - # Provide the vendorid:productid of the VFs - #NeutronSupportedPCIVendorDevs: ['8086:154c','8086:10ca','8086:1520'] - #NeutronPhysicalDevMappings: "datacentre:ens20f2" # Number of VFs that needs to be configured for a physical interface diff --git a/extraconfig/tasks/major_upgrade_block_storage.sh b/extraconfig/tasks/major_upgrade_block_storage.sh deleted file mode 100644 index 64c4457e..00000000 --- a/extraconfig/tasks/major_upgrade_block_storage.sh +++ /dev/null @@ -1,11 +0,0 @@ -#!/bin/bash -# -# This runs an upgrade of Cinder Block Storage nodes. -# -set -eu - -# Special-case OVS for https://bugs.launchpad.net/tripleo/+bug/1635205 -special_case_ovs_upgrade_if_needed - -yum -y install python-zaqarclient # needed for os-collect-config -yum -y -q update diff --git a/extraconfig/tasks/major_upgrade_ceph_mon.sh b/extraconfig/tasks/major_upgrade_ceph_mon.sh deleted file mode 100755 index e0d160f1..00000000 --- a/extraconfig/tasks/major_upgrade_ceph_mon.sh +++ /dev/null @@ -1,82 +0,0 @@ -#!/bin/bash -set -eu -set -o pipefail - -echo INFO: starting $(basename "$0") - -# Exit if not running -if ! pidof ceph-mon &> /dev/null; then - echo INFO: ceph-mon is not running, skipping - exit 0 -fi - -# Exit if not Hammer -INSTALLED_VERSION=$(ceph --version | awk '{print $3}') -if ! [[ "$INSTALLED_VERSION" =~ ^0\.94.* ]]; then - echo INFO: version of Ceph installed is not 0.94, skipping - exit 0 -fi - -CEPH_STATUS=$(ceph health | awk '{print $1}') -if [ ${CEPH_STATUS} = HEALTH_ERR ]; then - echo ERROR: Ceph cluster status is HEALTH_ERR, cannot be upgraded - exit 1 -fi - -# Useful when upgrading with OSDs num < replica size -if [[ ${ignore_ceph_upgrade_warnings:-False} != [Tt]rue ]]; then - timeout 300 bash -c "while [ ${CEPH_STATUS} != HEALTH_OK ]; do - echo WARNING: Waiting for Ceph cluster status to go HEALTH_OK; - sleep 30; - CEPH_STATUS=$(ceph health | awk '{print $1}') - done" -fi - -MON_PID=$(pidof ceph-mon) -MON_ID=$(hostname -s) - -# Stop daemon using Hammer sysvinit script -service ceph stop mon.${MON_ID} - -# Ensure it's stopped -timeout 60 bash -c "while kill -0 ${MON_PID} 2> /dev/null; do - sleep 2; -done" - -# Update to Jewel -yum -y -q update ceph-mon ceph - -# Restart/Exit if not on Jewel, only in that case we need the changes -UPDATED_VERSION=$(ceph --version | awk '{print $3}') -if [[ "$UPDATED_VERSION" =~ ^0\.94.* ]]; then - echo WARNING: Ceph was not upgraded, restarting daemons - service ceph start mon.${MON_ID} -elif [[ "$UPDATED_VERSION" =~ ^10\.2.* ]]; then - # RPM could own some of these but we can't take risks on the pre-existing files - for d in /var/lib/ceph/mon /var/log/ceph /var/run/ceph /etc/ceph; do - chown -L -R ceph:ceph $d || echo WARNING: chown of $d failed - done - - # Replay udev events with newer rules - udevadm trigger - - # Enable systemd unit - systemctl enable ceph-mon.target - systemctl enable ceph-mon@${MON_ID} - systemctl start ceph-mon@${MON_ID} - - # Wait for daemon to be back in the quorum - timeout 300 bash -c "until (ceph quorum_status | jq .quorum_names | grep -sq ${MON_ID}); do - echo WARNING: Waiting for mon.${MON_ID} to re-join quorum; - sleep 10; - done" - - # if tunables become legacy, cluster status will be HEALTH_WARN causing - # upgrade to fail on following node - ceph osd crush tunables default - - echo INFO: Ceph was upgraded to Jewel -else - echo ERROR: Ceph was upgraded to an unknown release, daemon is stopped, need manual intervention - exit 1 -fi diff --git a/extraconfig/tasks/major_upgrade_ceph_storage.sh b/extraconfig/tasks/major_upgrade_ceph_storage.sh deleted file mode 100644 index a745e723..00000000 --- a/extraconfig/tasks/major_upgrade_ceph_storage.sh +++ /dev/null @@ -1,106 +0,0 @@ -#!/bin/bash -# -# This delivers the ceph-storage upgrade script to be invoked as part of the tripleo -# major upgrade workflow. -# -set -eu -set -o pipefail - -UPGRADE_SCRIPT=/root/tripleo_upgrade_node.sh - -declare -f special_case_ovs_upgrade_if_needed > $UPGRADE_SCRIPT -# use >> here so we don't lose the declaration we added above -cat >> $UPGRADE_SCRIPT << 'ENDOFCAT' -#!/bin/bash -### DO NOT MODIFY THIS FILE -### This file is automatically delivered to the ceph-storage nodes as part of the -### tripleo upgrades workflow -set -eu - -echo INFO: starting $(basename "$0") - -# Exit if not running -if ! pidof ceph-osd &> /dev/null; then - echo INFO: ceph-osd is not running, skipping - exit 0 -fi - -# Exit if not Hammer -INSTALLED_VERSION=$(ceph --version | awk '{print $3}') -if ! [[ "$INSTALLED_VERSION" =~ ^0\.94.* ]]; then - echo INFO: version of Ceph installed is not 0.94, skipping - exit 0 -fi - -OSD_PIDS=$(pidof ceph-osd) -OSD_IDS=$(ls /var/lib/ceph/osd | awk 'BEGIN { FS = "-" } ; { print $2 }') - -# "so that mirrors aren't rebalanced as if the OSD died" - gfidente / leseb -ceph osd set noout -ceph osd set norebalance -ceph osd set nodeep-scrub -ceph osd set noscrub - -# Stop daemon using Hammer sysvinit script -for OSD_ID in $OSD_IDS; do - service ceph stop osd.${OSD_ID} -done - -# Nice guy will return non-0 only when all failed -timeout 60 bash -c "while kill -0 ${OSD_PIDS} 2> /dev/null; do - sleep 2; -done" - -special_case_ovs_upgrade_if_needed - -# Update (Ceph to Jewel) -yum -y install python-zaqarclient # needed for os-collect-config -yum -y update - -# Restart/Exit if not on Jewel, only in that case we need the changes -UPDATED_VERSION=$(ceph --version | awk '{print $3}') -if [[ "$UPDATED_VERSION" =~ ^0\.94.* ]]; then - echo WARNING: Ceph was not upgraded, restarting daemon - for OSD_ID in $OSD_IDS; do - service ceph start osd.${OSD_ID} - done -elif [[ "$UPDATED_VERSION" =~ ^10\.2.* ]]; then - # RPM could own some of these but we can't take risks on the pre-existing files - for d in /var/lib/ceph/osd /var/log/ceph /var/run/ceph /etc/ceph; do - chown -L -R ceph:ceph $d || echo WARNING: chown of $d failed - done - - # Replay udev events with newer rules - udevadm trigger && udevadm settle - - # If on ext4, we need to enforce lower values for name and namespace len - # or ceph-osd will refuse to start, see: http://tracker.ceph.com/issues/16187 - for OSD_ID in $OSD_IDS; do - OSD_FS=$(df -l --output=fstype /var/lib/ceph/osd/ceph-${OSD_ID} | tail -n +2) - if [ ${OSD_FS} = ext4 ]; then - crudini --set /etc/ceph/ceph.conf global osd_max_object_name_len 256 - crudini --set /etc/ceph/ceph.conf global osd_max_object_namespace_len 64 - fi - done - - # Enable systemd unit - systemctl enable ceph-osd.target - for OSD_ID in $OSD_IDS; do - systemctl enable ceph-osd@${OSD_ID} - systemctl start ceph-osd@${OSD_ID} - done - - echo INFO: Ceph was upgraded to Jewel -else - echo ERROR: Ceph was upgraded to an unknown release, daemon is stopped, need manual intervention - exit 1 -fi - -ceph osd unset noout -ceph osd unset norebalance -ceph osd unset nodeep-scrub -ceph osd unset noscrub -ENDOFCAT - -# ensure the permissions are OK -chmod 0755 $UPGRADE_SCRIPT 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.yaml b/extraconfig/tasks/major_upgrade_pacemaker.yaml index b63aafbd..8c91027d 100644 --- a/extraconfig/tasks/major_upgrade_pacemaker.yaml +++ b/extraconfig/tasks/major_upgrade_pacemaker.yaml @@ -33,33 +33,6 @@ resources: # map_merge with input_values instead of feeding params into scripts # via str_replace on bash snippets - CephMonUpgradeConfig: - type: OS::Heat::SoftwareConfig - properties: - group: script - config: - list_join: - - '' - - - str_replace: - template: | - #!/bin/bash - ignore_ceph_upgrade_warnings='IGNORE_CEPH_UPGRADE_WARNINGS' - params: - IGNORE_CEPH_UPGRADE_WARNINGS: {get_param: IgnoreCephUpgradeWarnings} - - get_file: major_upgrade_ceph_mon.sh - - CephMonUpgradeDeployment: - type: OS::Heat::SoftwareDeploymentGroup - properties: - servers: {get_param: [servers, Controller]} - config: {get_resource: CephMonUpgradeConfig} - input_values: {get_param: input_values} - update_policy: - batch_create: - max_batch_size: 1 - rolling_update: - max_batch_size: 1 - ControllerPacemakerUpgradeConfig_Step1: type: OS::Heat::SoftwareConfig properties: @@ -86,30 +59,11 @@ resources: ControllerPacemakerUpgradeDeployment_Step1: type: OS::Heat::SoftwareDeploymentGroup - depends_on: CephMonUpgradeDeployment properties: servers: {get_param: [servers, Controller]} config: {get_resource: ControllerPacemakerUpgradeConfig_Step1} input_values: {get_param: input_values} - BlockStorageUpgradeConfig: - type: OS::Heat::SoftwareConfig - depends_on: ControllerPacemakerUpgradeDeployment_Step1 - properties: - group: script - config: - list_join: - - '' - - - get_file: pacemaker_common_functions.sh - - get_file: major_upgrade_block_storage.sh - - BlockStorageUpgradeDeployment: - type: OS::Heat::SoftwareDeploymentGroup - properties: - servers: {get_param: [servers, BlockStorage]} - config: {get_resource: BlockStorageUpgradeConfig} - input_values: {get_param: input_values} - ControllerPacemakerUpgradeConfig_Step2: type: OS::Heat::SoftwareConfig properties: @@ -136,7 +90,7 @@ resources: ControllerPacemakerUpgradeDeployment_Step2: type: OS::Heat::SoftwareDeploymentGroup - depends_on: BlockStorageUpgradeDeployment + depends_on: ControllerPacemakerUpgradeDeployment_Step1 properties: servers: {get_param: [servers, Controller]} config: {get_resource: ControllerPacemakerUpgradeConfig_Step2} 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..27ba33a8 --- /dev/null +++ b/extraconfig/tasks/tripleo_upgrade_node.sh @@ -0,0 +1,54 @@ +#!/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 + +$(declare -f special_case_ovs_upgrade_if_needed) +special_case_ovs_upgrade_if_needed + +yum -y install python-zaqarclient # needed for os-collect-config +systemctl_swift stop +yum -y update +systemctl_swift start + +# Due to bug#1640177 we need to restart compute agent +if [[ -n \$NOVA_COMPUTE ]]; then + echo "Restarting openstack ceilometer agent compute" + systemctl restart openstack-ceilometer-compute +fi + +# Apply puppet manifest to converge just right after the \$ROLE upgrade +puppet apply /root/${ROLE}_puppet_config.pp + +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 b879fafa..3362a01f 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 +49,45 @@ 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' + ROLE='ROLE_NAME' + params: + SERVICE_PASSWORD: { get_param: NovaPassword } + REGION_NAME: { get_param: KeystoneRegion } + AUTH_URL: { get_param: [EndpointMap, KeystoneAdmin, uri_no_suffix]} + ROLE_NAME: {{role.name}} + - 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 @@ -46,11 +99,22 @@ resources: {{role.name}}UpgradeBatchConfig_Step{{step}}: type: OS::TripleO::UpgradeConfig {%- if step > 0 %} + condition: {{role.name}}UpgradeBatchConfigEnabled + {% if role.name in enabled_roles %} depends_on: - {%- for dep in enabled_roles %} - - {{dep.name}}UpgradeBatch_Step{{step -1}} - {%- endfor %} - {%- endif %} + - {{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 %} + {% 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}} @@ -61,12 +125,13 @@ resources: {{role.name}}UpgradeBatch_Step{{step}}: type: OS::Heat::SoftwareDeploymentGroup condition: {{role.name}}UpgradeBatchConfigEnabled - {%- if step > 0 %} + {%- if step > 0 %} depends_on: - {%- for dep in enabled_roles %} - - {{dep.name}}UpgradeBatch_Step{{step -1}} - {%- endfor %} - {%- endif %} + - {{role.name}}UpgradeBatch_Step{{step -1}} + {% else %} + depends_on: + - {{role.name}}UpgradeBatchConfig_Step{{step}} + {%- endif %} update_policy: batch_create: max_batch_size: {{role.upgrade_batch_size|default(1)}} @@ -82,6 +147,32 @@ resources: {%- endfor %} {%- endfor %} +# Dump the puppet manifests to be apply later when disable_upgrade_deployment +# is to true +{% for role in roles if role.disable_upgrade_deployment|default(false) %} + {{role.name}}DeliverPuppetConfig: + type: OS::Heat::SoftwareConfig + properties: + group: script + config: + list_join: + - '' + - - str_replace: + template: | + #!/bin/bash + cat > /root/{{role.name}}_puppet_config.pp << ENDOFCAT + PUPPET_CLASSES + ENDOFCAT + params: + PUPPET_CLASSES: {get_param: [role_data, {{role.name}}, step_config]} + + {{role.name}}DeliverPuppetDeployment: + type: OS::Heat::SoftwareDeploymentGroup + properties: + servers: {get_param: [servers, {{role.name}}]} + config: {get_resource: {{role.name}}DeliverPuppetConfig} +{% endfor %} + # Upgrade Steps for all roles {%- for step in range(0, upgrade_steps_max) %} # Config resources for step {{step}} @@ -92,15 +183,12 @@ resources: # serialization, but the event output is easier to follow if we # do, and there should be minimal performance hit (creating the # config is cheap compared to the time to apply the deployment). - depends_on: {%- if step > 0 %} - {%- for dep in enabled_roles %} - - {{dep.name}}Upgrade_Step{{step -1}} - {%- endfor %} - {%- else %} - {%- for dep in enabled_roles %} - - {{dep.name}}UpgradeBatch_Step{{batch_upgrade_steps_max -1}} - {%- endfor %} + condition: {{role.name}}UpgradeConfigEnabled + {% if role.name in enabled_roles %} + depends_on: + - {{role.name}}Upgrade_Step{{step -1}} + {% endif %} {%- endif %} properties: UpgradeStepConfig: {get_param: [role_data, {{role.name}}, upgrade_tasks]} @@ -111,17 +199,11 @@ resources: {%- for role in enabled_roles %} {{role.name}}Upgrade_Step{{step}}: type: OS::Heat::SoftwareDeploymentGroup + {%- if step > 0 %} condition: {{role.name}}UpgradeConfigEnabled depends_on: - {%- if step > 0 %} - {%- for dep in enabled_roles %} - - {{dep.name}}Upgrade_Step{{step -1}} - {%- endfor %} - {%- else %} - {%- for dep in enabled_roles %} - - {{dep.name}}UpgradeBatch_Step{{batch_upgrade_steps_max -1}} - {%- endfor %} - {%- endif %} + - {{role.name}}Upgrade_Step{{step -1}} + {%- endif %} properties: name: {{role.name}}Upgrade_Step{{step}} servers: {get_param: [servers, {{role.name}}]} @@ -154,4 +236,3 @@ outputs: {% for role in roles %} {{role.name.lower()}}: {get_attr: [{{role.name}}UpgradeConfig_Step1, upgrade_config]} {% endfor %} - 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/puppet/services/aodh-api.yaml b/puppet/services/aodh-api.yaml index 62c4b093..4bd9fc47 100644 --- a/puppet/services/aodh-api.yaml +++ b/puppet/services/aodh-api.yaml @@ -89,6 +89,3 @@ outputs: - name: Stop aodh_api service (running under httpd) tags: step2 service: name=httpd state=stopped - - name: Run aodh dbsync - tags: step5 - command: aodh-dbsync diff --git a/puppet/services/gnocchi-api.yaml b/puppet/services/gnocchi-api.yaml index 2a1ed2a3..23fcb2f6 100644 --- a/puppet/services/gnocchi-api.yaml +++ b/puppet/services/gnocchi-api.yaml @@ -131,6 +131,3 @@ outputs: - name: Stop gnocchi_api service (running under httpd) tags: step2 service: name=httpd state=stopped - - name: Run gnocchi upgrade - tags: step5 - command: gnocchi-upgrade diff --git a/puppet/services/neutron-plugin-ml2.yaml b/puppet/services/neutron-plugin-ml2.yaml index 71a0076f..3abd04f3 100644 --- a/puppet/services/neutron-plugin-ml2.yaml +++ b/puppet/services/neutron-plugin-ml2.yaml @@ -60,12 +60,6 @@ parameters: default: 'vxlan' description: The tenant network type for Neutron. type: comma_delimited_list - NeutronSupportedPCIVendorDevs: - description: | - List of supported pci vendor devices in the format VendorID:ProductID. - By default Intel & Mellanox SR-IOV capable NICs are supported. - type: comma_delimited_list - default: ['15b3:1004','8086:10ca'] resources: NeutronBase: @@ -91,7 +85,6 @@ outputs: neutron::plugins::ml2::tunnel_id_ranges: {get_param: NeutronTunnelIdRanges} neutron::plugins::ml2::vni_ranges: {get_param: NeutronVniRanges} neutron::plugins::ml2::tenant_network_types: {get_param: NeutronNetworkType} - neutron::plugins::ml2::supported_pci_vendor_devs: {get_param: NeutronSupportedPCIVendorDevs} step_config: | include ::tripleo::profile::base::neutron::plugins::ml2 diff --git a/puppet/services/pacemaker.yaml b/puppet/services/pacemaker.yaml index c47229f1..ca21cfbe 100644 --- a/puppet/services/pacemaker.yaml +++ b/puppet/services/pacemaker.yaml @@ -87,6 +87,10 @@ parameters: \[(?<pid>[^ ]*)\] (?<host>[^ ]*) (?<message>.*)$/ + PacemakerResources: + type: comma_delimited_list + description: List of resources managed by pacemaker + default: ['rabbitmq','haproxy'] outputs: role_data: @@ -127,3 +131,17 @@ outputs: tripleo::profile::base::pacemaker::remote_authkey: {get_param: PacemakerRemoteAuthkey} step_config: | include ::tripleo::profile::base::pacemaker + upgrade_tasks: + - name: Check pacemaker cluster running before upgrade + tags: step0,validation + pacemaker_cluster: state=online check_and_fail=true + - name: Stop pacemaker cluster + tags: step1 + pacemaker_cluster: state=offline + - name: Start pacemaker cluster + tags: step4 + pacemaker_cluster: state=online + - name: Check pacemaker resource + tags: step4 + pacemaker_resource: state=started resource={{item}} check_mode=true wait_for_resource=true timeout=200 + with_items: {get_param: PacemakerResources} diff --git a/roles_data.yaml b/roles_data.yaml index 4b13baf9..31b12986 100644 --- a/roles_data.yaml +++ b/roles_data.yaml @@ -185,7 +185,6 @@ - OS::TripleO::Services::Collectd - name: CephStorage - disable_upgrade_deployment: True ServicesDefault: - OS::TripleO::Services::CACerts - OS::TripleO::Services::CephOSD |