tripleo-heat-templates/deployment/octavia/octavia-health-manager-container-puppet.yaml
Gregory Thiemonge 15744039dc Increase stop_grace_period for Octavia controller services
The stop_grace_period for Octavia should be greater than the longest
task executed by Octavia.
The timeout of longest task in Octavia is currently set to 10 minutes
[0].

Updating the stop_grace_period from 5min to 10min30 will prevent
failures during updates. If a really long task is running Octavia (for
instance a failover during a network outage), it will timeout in Octavia
and Octavia will exit gracefully.

[0] https://review.opendev.org/c/openstack/octavia/+/637077

Change-Id: I05cd4a54c35479ca06580f541042948f69c7f532
2022-03-14 12:00:21 +01:00

236 lines
9.2 KiB
YAML

heat_template_version: wallaby
description: >
OpenStack Octavia health-manager service configured with Puppet
parameters:
ContainerOctaviaHealthManagerImage:
description: image
type: string
tags:
- role_specific
ContainerOctaviaConfigImage:
description: The container image to use for the octavia config_volume
type: string
tags:
- role_specific
ContainerOctaviaRsyslogImage:
description: The container image to use for the rsyslog-sidecar
type: string
tags:
- role_specific
OctaviaHealthManagerLoggingSource:
type: json
default:
tag: openstack.octavia.health-manager
file: /var/log/containers/octavia/health-manager.log
EndpointMap:
default: {}
description: Mapping of service endpoint -> protocol. Typically set
via parameter_defaults in the resource registry.
type: json
ServiceData:
default: {}
description: Dictionary packing service data
type: json
ServiceNetMap:
default: {}
description: Mapping of service_name -> network name. Typically set
via parameter_defaults in the resource registry. Use
parameter_merge_strategies to merge it with the defaults.
type: json
RoleName:
default: ''
description: Role name on which the service is applied
type: string
RoleParameters:
default: {}
description: Parameters specific to the role
type: json
MonitoringSubscriptionOctaviaHealthManager:
default: 'overcloud-octavia-health-manager'
type: string
OctaviaHeartbeatKey:
type: string
description: Key to identify heartbeat messages for amphorae.
hidden: true
OctaviaMgmtPortDevName:
type: string
default: "o-hm0"
description: Name of the octavia management network interface using
for communication between octavia worker/health-manager
with the amphora machine.
OctaviaLogOffload:
default: true
description: When true, log messages from the amphora will be forwarded
to the administrative log endponts and will be stored with
the controller logs.
type: boolean
resources:
ContainersCommon:
type: ../containers-common.yaml
MySQLClient:
type: ../database/mysql-client.yaml
OctaviaBase:
type: ./octavia-base.yaml
properties:
EndpointMap: {get_param: EndpointMap}
ServiceData: {get_param: ServiceData}
ServiceNetMap: {get_param: ServiceNetMap}
RoleName: {get_param: RoleName}
RoleParameters: {get_param: RoleParameters}
RoleParametersValue:
type: OS::Heat::Value
properties:
type: json
value:
map_replace:
- map_replace:
- ContainerOctaviaHealthManagerImage: ContainerOctaviaHealthManagerImage
ContainerOctaviaConfigImage: ContainerOctaviaConfigImage
ContainerOctaviaRsyslogImage: ContainerOctaviaRsyslogImage
- values: {get_param: [RoleParameters]}
- values:
ContainerOctaviaHealthManagerImage: {get_param: ContainerOctaviaHealthManagerImage}
ContainerOctaviaConfigImage: {get_param: ContainerOctaviaConfigImage}
ContainerOctaviaRsyslogImage: {get_param: ContainerOctaviaConfigImage}
outputs:
role_data:
description: Role data for the Octavia health-manager role.
value:
service_name: octavia_health_manager
firewall_rules:
'200 octavia health manager interface':
proto: udp
dport: 5555
iniface: {get_param: OctaviaMgmtPortDevName}
'210 octavia lb-mgmt-net offload rsyslog':
proto: udp
dport: 514
iniface: {get_param: OctaviaMgmtPortDevName}
monitoring_subscription: {get_param: MonitoringSubscriptionOctaviaHealthManager}
config_settings:
map_merge:
- get_attr: [OctaviaBase, role_data, config_settings]
- octavia::health_manager::heartbeat_key: {get_param: OctaviaHeartbeatKey}
service_config_settings:
rsyslog:
tripleo_logging_sources_octavia_health_manager:
- {get_param: OctaviaHealthManagerLoggingSource}
# BEGIN DOCKER SETTINGS #
puppet_config:
config_volume: octavia
puppet_tags: octavia_config
step_config:
list_join:
- "\n"
- - "include tripleo::profile::base::octavia::health_manager"
- {get_attr: [MySQLClient, role_data, step_config]}
config_image: {get_attr: [RoleParametersValue, value, ContainerOctaviaConfigImage]}
kolla_config:
/var/lib/kolla/config_files/octavia_health_manager.json:
command: /usr/bin/octavia-health-manager --config-file /etc/octavia/octavia.conf --config-file /etc/octavia/post-deploy.conf --log-file /var/log/octavia/health-manager.log --config-dir /etc/octavia/conf.d/octavia-health-manager
config_files:
- source: "/var/lib/kolla/config_files/src/*"
dest: "/"
merge: true
preserve_properties: true
permissions:
- path: /var/log/octavia
owner: octavia:octavia
recurse: true
/var/lib/kolla/config_files/octavia_rsyslog.json:
command: /usr/sbin/rsyslogd -n
config_files:
- source: "/var/lib/kolla/config_files/src/*"
dest: "/"
merge: true
preserve_properties: true
permissions:
- path: /var/log/octavia
owner: root:root
recurse: true
docker_config:
step_2:
octavia_health_manager_init_dirs:
start_order: 0
image: &octavia_health_manager_image {get_attr: [RoleParametersValue, value, ContainerOctaviaHealthManagerImage]}
user: root
net: none
volumes:
# NOTE(mandre) we need extra dir for the service in /etc/octavia/conf.d
# It is normally created as part of the RPM install, but it is
# missing here because we use the same config_volume for all
# octavia services, hence the same container image to generate
# configuration.
- /var/lib/config-data/puppet-generated/octavia/etc/octavia:/etc/octavia:z
command: ['/bin/bash', '-c', 'mkdir -p /etc/octavia/conf.d/octavia-health-manager; chown -R octavia:octavia /etc/octavia/conf.d/octavia-health-manager']
step_5:
map_merge:
- octavia_health_manager:
start_order: 2
stop_grace_period: 630
image: *octavia_health_manager_image
net: host
privileged: false
restart: always
healthcheck:
test: /openstack/healthcheck
volumes:
list_concat:
- {get_attr: [ContainersCommon, volumes]}
- - /var/lib/kolla/config_files/octavia_health_manager.json:/var/lib/kolla/config_files/config.json:ro
- /var/lib/config-data/puppet-generated/octavia:/var/lib/kolla/config_files/src:ro
- /var/log/containers/octavia:/var/log/octavia:z
environment:
KOLLA_CONFIG_STRATEGY: COPY_ALWAYS
- if:
- {get_param: OctaviaLogOffload}
- octavia_rsyslog:
start_order: 2
image: {get_attr: [RoleParametersValue, value, ContainerOctaviaRsyslogImage]}
net: host
privileged: false
restart: always
volumes:
list_concat:
- {get_attr: [ContainersCommon, volumes]}
-
- /var/lib/kolla/config_files/octavia_rsyslog.json:/var/lib/kolla/config_files/config.json:ro
- /var/lib/config-data/puppet-generated/octavia:/var/lib/kolla/config_files/src:ro
- /var/log/containers/octavia-amphorae:/var/log/octavia:z
environment:
KOLLA_CONFIG_STRATEGY: COPY_ALWAYS
update_tasks: {get_attr: [OctaviaBase, role_data, update_tasks]}
upgrade_tasks: {get_attr: [OctaviaBase, role_data, upgrade_tasks]}
host_prep_tasks:
- name: create persistent directories
file:
path: "{{ item.path }}"
state: directory
setype: "{{ item.setype }}"
mode: "{{ item.mode }}"
with_items:
- { 'path': /var/log/containers/octavia, 'setype': container_file_t, 'mode': '0750' }
- { 'path': /var/log/containers/octavia-amphorae, 'setype': container_file_t, 'mode': '0750' }
external_upgrade_tasks:
- when:
- step|int == 1
tags:
- never
- system_upgrade_transfer_data
- system_upgrade_stop_services
block:
- name: Stop octavia health manager container
import_role:
name: tripleo_container_stop
vars:
tripleo_containers_to_stop:
- octavia_health_manager
tripleo_delegate_to: "{{ groups['octavia_health_manager'] | default([]) }}"