Heat templates for deploying OpenStack
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

563 lines
24 KiB

#!/usr/bin/env python
# Licensed under the Apache License, Version 2.0 (the "License"); you may
# not use this file except in compliance with the License. You may obtain
# a copy of the License at
# http://www.apache.org/licenses/LICENSE-2.0
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
# License for the specific language governing permissions and limitations
# under the License.
import os
import sys
import traceback
import yaml
required_params = ['EndpointMap', 'ServiceNetMap', 'DefaultPasswords',
'RoleName', 'RoleParameters', 'ServiceData']
# NOTE(bnemec): The duplication in this list is intentional. For the
# transition to generated environments we have two copies of these files,
# so they need to be listed twice. Once the deprecated version can be removed
# the duplicate entries can be as well.
envs_containing_endpoint_map = ['tls-endpoints-public-dns.yaml',
ENDPOINT_MAP_FILE = 'endpoint_map.yaml'
OPTIONAL_SECTIONS = ['service_workflow_tasks']
REQUIRED_DOCKER_SECTIONS = ['service_name', 'docker_config', 'puppet_config',
'config_settings', 'step_config']
OPTIONAL_DOCKER_SECTIONS = ['docker_puppet_tasks', 'upgrade_tasks',
'service_config_settings', 'host_prep_tasks',
'metadata_settings', 'kolla_config',
REQUIRED_DOCKER_PUPPET_CONFIG_SECTIONS = ['config_volume', 'step_config',
OPTIONAL_DOCKER_PUPPET_CONFIG_SECTIONS = [ 'puppet_tags', 'volumes' ]
# Mapping of parameter names to a list of the fields we should _not_ enforce
# consistency across files on. This should only contain parameters whose
# definition we cannot change for backwards compatibility reasons. New
# parameters to the templates should not be added to this list.
PARAMETER_DEFINITION_EXCLUSIONS = {'ManagementNetCidr': ['default'],
'ManagementAllocationPools': ['default'],
'ExternalNetCidr': ['default'],
'ExternalAllocationPools': ['default'],
'StorageNetCidr': ['default'],
'StorageAllocationPools': ['default'],
'StorageMgmtNetCidr': ['default'],
'StorageMgmtAllocationPools': ['default'],
'TenantNetCidr': ['default'],
'TenantAllocationPools': ['default'],
'InternalApiNetCidr': ['default'],
'InternalApiAllocationPools': ['default'],
'UpdateIdentifier': ['description'],
'key_name': ['default'],
# There's one template that defines this
# differently, and I'm not sure if we can
# safely change it.
'EC2MetadataIp': ['default'],
# Same as EC2MetadataIp
'ControlPlaneDefaultRoute': ['default'],
# TODO(bnemec): Address these existing
# inconsistencies.
'ServiceNetMap': ['description', 'default'],
'network': ['default'],
'ControlPlaneIP': ['default',
'ControlPlaneIp': ['default',
'NeutronBigswitchLLDPEnabled': ['default'],
'NeutronWorkers': ['description'],
'ServerMetadata': ['description'],
'server': ['description'],
'servers': ['description'],
'ExtraConfig': ['description'],
'DefaultPasswords': ['description',
'BondInterfaceOvsOptions': ['description',
'KeyName': ['constraints'],
'OVNSouthboundServerPort': ['description'],
['description', 'default'],
'IPPool': ['description'],
'SSLCertificate': ['description',
'HostCpusList': ['default', 'constraints'],
'NodeIndex': ['description'],
'name': ['description', 'default'],
'image': ['description', 'default'],
'NeutronBigswitchAgentEnabled': ['default'],
'EndpointMap': ['description', 'default'],
'DockerManilaConfigImage': ['description',
'replacement_policy': ['default'],
'CloudDomain': ['description', 'default'],
'EnableLoadBalancer': ['description'],
'ControllerExtraConfig': ['description'],
'NovaComputeExtraConfig': ['description'],
'controllerExtraConfig': ['description'],
'DockerSwiftConfigImage': ['default']
'ec2api': 'Ec2Api',
'haproxy': 'HAProxy',
def exit_usage():
print('Usage %s <yaml file or directory>' % sys.argv[0])
def to_camel_case(string):
return PREFERRED_CAMEL_CASE.get(string, ''.join(s.capitalize() or '_' for
s in string.split('_')))
def get_base_endpoint_map(filename):
tpl = yaml.load(open(filename).read())
return tpl['parameters']['EndpointMap']['default']
except Exception:
return None
def get_endpoint_map_from_env(filename):
tpl = yaml.load(open(filename).read())
return {
'file': filename,
'map': tpl['parameter_defaults']['EndpointMap']
except Exception:
return None
def validate_endpoint_map(base_map, env_map):
return sorted(base_map.keys()) == sorted(env_map.keys())
def validate_hci_compute_services_default(env_filename, env_tpl):
env_services_list = env_tpl['parameter_defaults']['ComputeServices']
roles_filename = os.path.join(os.path.dirname(env_filename),
roles_tpl = yaml.load(open(roles_filename).read())
for role in roles_tpl:
if role['name'] == 'Compute':
roles_services_list = role['ServicesDefault']
if sorted(env_services_list) != sorted(roles_services_list):
print('ERROR: ComputeServices in %s is different from '
'ServicesDefault in roles/Compute.yaml' % env_filename)
return 1
return 0
def validate_hci_computehci_role(hci_role_filename, hci_role_tpl):
compute_role_filename = os.path.join(os.path.dirname(hci_role_filename),
compute_role_tpl = yaml.load(open(compute_role_filename).read())
compute_role_services = compute_role_tpl[0]['ServicesDefault']
for role in hci_role_tpl:
if role['name'] == 'ComputeHCI':
hci_role_services = role['ServicesDefault']
if sorted(hci_role_services) != sorted(compute_role_services):
print('ERROR: ServicesDefault in %s is different from'
'ServicesDefault in roles/Compute.yaml' % hci_role_filename)
return 1
return 0
def search(item, check_item, check_key):
if check_item(item):
return True
elif isinstance(item, list):
for i in item:
if search(i, check_item, check_key):
return True
elif isinstance(item, dict):
for k in item.keys():
if check_key(k, item[k]):
return True
elif search(item[k], check_item, check_key):
return True
return False
def validate_mysql_connection(settings):
no_op = lambda *args: False
error_status = [0]
def mysql_protocol(items):
return items == ['EndpointMap', 'MysqlInternal', 'protocol']
def client_bind_address(item):
return 'read_default_file' in item and \
'read_default_group' in item
def validate_mysql_uri(key, items):
# Only consider a connection if it targets mysql
if key.endswith('connection') and \
search(items, mysql_protocol, no_op):
# Assume the "bind_address" option is one of
# the token that made up the uri
if not search(items, client_bind_address, no_op):
error_status[0] = 1
return False
search(settings, no_op, validate_mysql_uri)
return error_status[0]
def validate_docker_service_mysql_usage(filename, tpl):
no_op = lambda *args: False
included_res = []
def match_included_res(item):
is_config_setting = isinstance(item, list) and len(item) > 1 and \
item[1:] == ['role_data', 'config_settings']
if is_config_setting:
return is_config_setting
def match_use_mysql_protocol(items):
return items == ['EndpointMap', 'MysqlInternal', 'protocol']
all_content = []
def read_all(incfile, inctpl):
# search for included content
content = inctpl['outputs']['role_data']['value'].get('config_settings',{})
included_res[:] = []
if search(content, match_included_res, no_op):
files = [inctpl['resources'][x]['type'] for x in included_res]
# parse included content
for r, f in zip(included_res, files):
# disregard class names, only consider file names
if 'OS::' in f:
newfile = os.path.normpath(os.path.dirname(incfile)+'/'+f)
newtmp = yaml.load(open(newfile).read())
read_all(newfile, newtmp)
read_all(filename, tpl)
if search(all_content, match_use_mysql_protocol, no_op):
# ensure this service includes the mysqlclient service
resources = tpl['resources']
mysqlclient = [x for x in resources
if resources[x]['type'].endswith('mysql-client.yaml')]
if len(mysqlclient) == 0:
print("ERROR: containerized service %s uses mysql but "
"resource mysql-client.yaml is not used"
% filename)
return 1
# and that mysql::client puppet module is included in puppet-config
match_mysqlclient = \
lambda x: x == [mysqlclient[0], 'role_data', 'step_config']
role_data = tpl['outputs']['role_data']
puppet_config = role_data['value']['puppet_config']['step_config']
if not search(puppet_config, match_mysqlclient, no_op):
print("ERROR: containerized service %s uses mysql but "
"puppet_config section does not include "
% filename)
return 1
return 0
def validate_docker_service(filename, tpl):
if 'outputs' in tpl and 'role_data' in tpl['outputs']:
if 'value' not in tpl['outputs']['role_data']:
print('ERROR: invalid role_data for filename: %s'
% filename)
return 1
role_data = tpl['outputs']['role_data']['value']
for section_name in REQUIRED_DOCKER_SECTIONS:
if section_name not in role_data:
print('ERROR: %s is required in role_data for %s.'
% (section_name, filename))
return 1
for section_name in role_data.keys():
if section_name in REQUIRED_DOCKER_SECTIONS:
if section_name in OPTIONAL_DOCKER_SECTIONS:
elif section_name in OPTIONAL_SECTIONS:
print('ERROR: %s is extra in role_data for %s.'
% (section_name, filename))
return 1
if 'puppet_config' in role_data:
if validate_docker_service_mysql_usage(filename, tpl):
print('ERROR: could not validate use of mysql service for %s.'
% filename)
return 1
puppet_config = role_data['puppet_config']
for key in puppet_config:
print('ERROR: %s should not be in puppet_config section.'
% key)
return 1
if key not in puppet_config:
print('ERROR: %s is required in puppet_config for %s.'
% (key, filename))
return 1
config_volume = puppet_config.get('config_volume')
expected_config_image_parameter = "Docker%sConfigImage" % to_camel_case(config_volume)
if config_volume and not expected_config_image_parameter in tpl.get('parameters', []):
print('ERROR: Missing %s heat parameter for %s config_volume.'
% (expected_config_image_parameter, config_volume))
return 1
if 'docker_config' in role_data:
docker_config = role_data['docker_config']
for _, step in docker_config.items():
if not isinstance(step, dict):
# NOTE(mandre) this skips everything that is not a dict
# so we may ignore some containers definitions if they
# are in a map_merge for example
for _, container in step.items():
if not isinstance(container, dict):
command = container.get('command', '')
if isinstance(command, list):
command = ' '.join(map(str, command))
if 'bootstrap_host_exec' in command \
and container.get('user') != 'root':
print('ERROR: bootstrap_host_exec needs to run as the root user.')
return 1
if 'parameters' in tpl:
for param in required_params:
if param not in tpl['parameters']:
print('ERROR: parameter %s is required for %s.'
% (param, filename))
return 1
return 0
def validate_service(filename, tpl):
if 'outputs' in tpl and 'role_data' in tpl['outputs']:
if 'value' not in tpl['outputs']['role_data']:
print('ERROR: invalid role_data for filename: %s'
% filename)
return 1
role_data = tpl['outputs']['role_data']['value']
if 'service_name' not in role_data:
print('ERROR: service_name is required in role_data for %s.'
% filename)
return 1
# service_name must match the filename, but with an underscore
if (role_data['service_name'] !=
os.path.basename(filename).split('.')[0].replace("-", "_")):
print('ERROR: service_name should match file name for service: %s.'
% filename)
return 1
# if service connects to mysql, the uri should use option
# bind_address to avoid issues with VIP failover
if 'config_settings' in role_data and \
print('ERROR: mysql connection uri should use option bind_address')
return 1
if 'parameters' in tpl:
for param in required_params:
if param not in tpl['parameters']:
print('ERROR: parameter %s is required for %s.'
% (param, filename))
return 1
return 0
def validate(filename, param_map):
"""Validate a Heat template
:param filename: The path to the file to validate
:param param_map: A dict which will be populated with the details of the
parameters in the template. The dict will have the
following structure:
{'ParameterName': [
{'filename': ./file1.yaml,
'data': {'description': '',
'type': string,
'default': '',
{'filename': ./file2.yaml,
'data': {'description': '',
'type': string,
'default': '',
print('Validating %s' % filename)
retval = 0
tpl = yaml.load(open(filename).read())
# The template alias version should be used instead a date, this validation
# will be applied to all templates not just for those in the services folder.
if 'heat_template_version' in tpl and not str(tpl['heat_template_version']).isalpha():
print('ERROR: heat_template_version needs to be the release alias not a date: %s'
% filename)
return 1
# qdr aliases rabbitmq service to provide alternative messaging backend
if (filename.startswith('./puppet/services/') and
filename not in ['./puppet/services/qdr.yaml']):
retval = validate_service(filename, tpl)
if filename.startswith('./docker/services/'):
retval = validate_docker_service(filename, tpl)
if filename.endswith('hyperconverged-ceph.yaml'):
retval = validate_hci_compute_services_default(filename, tpl)
if filename.startswith('./roles/ComputeHCI.yaml'):
retval = validate_hci_computehci_role(filename, tpl)
except Exception:
return 1
# yaml is OK, now walk the parameters and output a warning for unused ones
if 'heat_template_version' in tpl:
for p, data in tpl.get('parameters', {}).items():
definition = {'data': data, 'filename': filename}
param_map.setdefault(p, []).append(definition)
if p in required_params:
str_p = '\'%s\'' % p
in_resources = str_p in str(tpl.get('resources', {}))
in_outputs = str_p in str(tpl.get('outputs', {}))
if not in_resources and not in_outputs:
print('Warning: parameter %s in template %s '
'appears to be unused' % (p, filename))
return retval
if len(sys.argv) < 2:
path_args = sys.argv[1:]
exit_val = 0
failed_files = []
base_endpoint_map = None
env_endpoint_maps = list()
param_map = {}
for base_path in path_args:
if os.path.isdir(base_path):
for subdir, dirs, files in os.walk(base_path):
if '.tox' in dirs:
for f in files:
if f.endswith('.yaml') and not f.endswith('.j2.yaml'):
file_path = os.path.join(subdir, f)
failed = validate(file_path, param_map)
if failed:
exit_val |= failed
base_endpoint_map = get_base_endpoint_map(file_path)
if f in envs_containing_endpoint_map:
env_endpoint_map = get_endpoint_map_from_env(file_path)
if env_endpoint_map:
elif os.path.isfile(base_path) and base_path.endswith('.yaml'):
failed = validate(base_path, param_map)
if failed:
exit_val |= failed
print('Unexpected argument %s' % base_path)
if base_endpoint_map and \
len(env_endpoint_maps) == len(envs_containing_endpoint_map):
for env_endpoint_map in env_endpoint_maps:
matches = validate_endpoint_map(base_endpoint_map,
if not matches:
print("ERROR: %s needs to be updated to match changes in base "
"endpoint map" % env_endpoint_map['file'])
exit_val |= 1
print("%s matches base endpoint map" % env_endpoint_map['file'])
print("ERROR: Did not find expected number of environments containing the "
"EndpointMap parameter. If you meant to add or remove one of these "
"environments then you also need to update this tool.")
if not base_endpoint_map:
if len(env_endpoint_maps) != len(envs_containing_endpoint_map):
matched_files = set(os.path.basename(matched_env_file['file'])
for matched_env_file in env_endpoint_maps)
failed_files.extend(set(envs_containing_endpoint_map) - matched_files)
exit_val |= 1
# Validate that duplicate parameters defined in multiple files all have the
# same definition.
mismatch_count = 0
for p, defs in param_map.items():
# Nothing to validate if the parameter is only defined once
if len(defs) == 1:
check_data = [d['data'] for d in defs]
# Override excluded fields so they don't affect the result
ex_dict = {}
for field in exclusions:
ex_dict[field] = 'IGNORED'
for d in check_data:
# If all items in the list are not == the first, then the check fails
if check_data.count(check_data[0]) != len(check_data):
mismatch_count += 1
exit_val |= 1
failed_files.extend([d['filename'] for d in defs])
print('Mismatched parameter definitions found for "%s"' % p)
print('Definitions found:')
for d in defs:
print(' %s:\n %s' % (d['filename'], d['data']))
print('Mismatched parameter definitions: %d' % mismatch_count)
if failed_files:
print('Validation failed on:')
for f in failed_files:
print('Validation successful!')