openstack/placement
Code Issues Proposed changes
Files
9a971dbb5dcf94ba12e4f2fc9b3423cb8375a061
placement/nova/conf/database.py
Roman Podoliaka bf31e145c9 Optional separate database for placement API 
If 'connection' is set in the 'placement_database' conf group use
that as the connection URL for the placement database. Otherwise if
it is None, the default, then use the entire api_database conf group
to configure a database connection.

This works by:

* adding a 'placement sync' and 'placement version' command to
  nova-manage

* adding placement_migrations that that sync will run

* adding a placement_database config group with the relevant
  database configuration settings

* adding a placement_context_manager. If
  CONF.placement_database.connection is None this is the same as
  the api_context_manager, otherwise it is a new one from its own config

* adjust nova/tests/fixtures to be aware of a 'placement' database
  and the placement_context_manager

This version of this change differs from others by not requiring
separate placement commands for migration, instead using existing
tooling, which makes the size of the change a bit smaller and also
addresses problems with the test fixtures needing to be too aware of
what migration code to run. Now it runs the same code.

This functionality is being provided to allow deployers to choose
between establishing a new database now or requiring a migration
later. The default is migration later.

This is a modification of Id93cb93a0f4e8667c8e7848aa8cff1d994c2c364
and I3290e26d0a212911f8ef386418b9fa08c685380b.

Change-Id: Ice03144376c9868c064e4393d531510615fc6fc1
Co-Authored-By: Chris Dent <cdent@anticdent.org>
Partially-Implements: blueprint generic-resource-pools
2016-08-11 13:30:25 +00:00

201 lines
7.0 KiB
Python
Raw Blame History

# Copyright 2015 OpenStack Foundation
# All Rights Reserved.
#
# 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.
from nova.conf import paths
from oslo_config import cfg
from oslo_db import options as oslo_db_options
_DEFAULT_SQL_CONNECTION = 'sqlite:///' + paths.state_path_def('nova.sqlite')
# NOTE(sdague): we know of at least 1 instance of out of tree usage
# for this config in RAX. They used this because of performance issues
# with some queries. We think the right path forward is fixing the
# SQLA queries to be more performant for everyone.
db_driver_opt = cfg.StrOpt(
'db_driver',
default='nova.db',
help='DEPRECATED: The driver to use for database access',
deprecated_for_removal=True)
# NOTE(markus_z): We cannot simply do:
# conf.register_opts(oslo_db_options.database_opts, 'api_database')
# If we reuse a db config option for two different groups ("api_database"
# and "database") and deprecate or rename a config option in one of these
# groups, "oslo.config" cannot correctly determine which one to update.
# That's why we copied & pasted these config options for the "api_database"
# group here. See commit ba407e3 ("Add support for multiple database engines")
# for more details.
api_db_group = cfg.OptGroup('api_database',
title='API Database Options',
help="""
The *Nova API Database* is a separate database which is used for information
which is used across *cells*. This database is mandatory since the Mitaka
release (13.0.0).
""")
api_db_opts = [
# TODO(markus_z): This should probably have a required=True attribute
cfg.StrOpt('connection',
secret=True,
help=''),
cfg.BoolOpt('sqlite_synchronous',
default=True,
help=''),
cfg.StrOpt('slave_connection',
secret=True,
help=''),
cfg.StrOpt('mysql_sql_mode',
default='TRADITIONAL',
help=''),
cfg.IntOpt('idle_timeout',
default=3600,
help=''),
# TODO(markus_z): We should probably default this to 5 to not rely on the
# SQLAlchemy default. Otherwise we wouldn't provide a stable default.
cfg.IntOpt('max_pool_size',
help=''),
cfg.IntOpt('max_retries',
default=10,
help=''),
# TODO(markus_z): This should have a minimum attribute of 0
cfg.IntOpt('retry_interval',
default=10,
help=''),
# TODO(markus_z): We should probably default this to 10 to not rely on the
# SQLAlchemy default. Otherwise we wouldn't provide a stable default.
cfg.IntOpt('max_overflow',
help=''),
# TODO(markus_z): This should probably make use of the "choices" attribute.
# "oslo.db" uses only the values [<0, 0, 50, 100] see module
# /oslo_db/sqlalchemy/engines.py method "_setup_logging"
cfg.IntOpt('connection_debug',
default=0,
help=''),
cfg.BoolOpt('connection_trace',
default=False,
help=''),
# TODO(markus_z): We should probably default this to 30 to not rely on the
# SQLAlchemy default. Otherwise we wouldn't provide a stable default.
cfg.IntOpt('pool_timeout',
help='')
] # noqa
def enrich_help_text(alt_db_opts):
def get_db_opts():
for group_name, db_opts in oslo_db_options.list_opts():
if group_name == 'database':
return db_opts
return []
for db_opt in get_db_opts():
for alt_db_opt in alt_db_opts:
if alt_db_opt.name == db_opt.name:
# NOTE(markus_z): We can append alternative DB specific help
# texts here if needed.
alt_db_opt.help = db_opt.help + alt_db_opt.help
# NOTE(cdent): See the note above on api_db_opts. The same issues
# apply here.
placement_db_group = cfg.OptGroup('placement_database',
title='Placement API database options',
help="""
The *Placement API Database* is a separate database which is used for the new
placement-api service. In Ocata release (14.0.0) this database is optional: if
connection option is not set, api database will be used instead. However, this
is not recommended, as it implies a potentially lengthy data migration in the
future. Operators are advised to use a separate database for Placement API from
the start.
""")
# TODO(rpodolyaka): see the notes on help messages on api_db_opts above, those
# also apply here
placement_db_opts = [
cfg.StrOpt('connection',
default=None,
help='',
secret=True),
cfg.BoolOpt('sqlite_synchronous',
default=True,
help=''),
cfg.StrOpt('slave_connection',
secret=True,
help=''),
cfg.StrOpt('mysql_sql_mode',
default='TRADITIONAL',
help=''),
cfg.IntOpt('idle_timeout',
default=3600,
help=''),
cfg.IntOpt('max_pool_size',
help=''),
cfg.IntOpt('max_retries',
default=10,
help=''),
cfg.IntOpt('retry_interval',
default=10,
help=''),
cfg.IntOpt('max_overflow',
help=''),
cfg.IntOpt('connection_debug',
default=0,
help=''),
cfg.BoolOpt('connection_trace',
default=False,
help=''),
cfg.IntOpt('pool_timeout',
help=''),
] # noqa
def register_opts(conf):
oslo_db_options.set_defaults(conf, connection=_DEFAULT_SQL_CONNECTION,
sqlite_db='nova.sqlite')
conf.register_opt(db_driver_opt)
conf.register_opts(api_db_opts, group=api_db_group)
conf.register_opts(placement_db_opts, group=placement_db_group)
def list_opts():
# NOTE(markus_z): 2016-04-04: If we list the oslo_db_options here, they
# get emitted twice(!) in the "sample.conf" file. First under the
# namespace "nova.conf" and second under the namespace "oslo.db". This
# is due to the setting in file "etc/nova/nova-config-generator.conf".
# As I think it is useful to have the "oslo.db" namespace information
# in the "sample.conf" file, I omit the listing of the "oslo_db_options"
# here.
enrich_help_text(api_db_opts)
enrich_help_text(placement_db_opts)
return {'DEFAULT': [db_driver_opt],
api_db_group: api_db_opts,
placement_db_group: placement_db_opts,
}
View Git Blame Copy Permalink