deb-cinder/cinder/api/common.py
scottda 6b11d276d1 cinder-api-microversions code
Many changes to the Cinder REST API require changes to the consumers of the API.
For example, If we need to add a required parameter to a method that is called
 by Nova, we'd need both the Nova calling code and the cinderclient that
Nova uses to change. But newer Cinder versions with the change must work with
older Nova versions, and there is no mechanism for this at the moment. Adding
microversions will solve this problem.
With microversions, the highest supported version will be negotiated by a field
in the HTTP header that is sent to the Cinder API. In the case where the field
'versions' is not sent (i.e. clients and scripts that pre-date this change),
then the lowest supported version would be used. In order to ensure that the
API consumer is explicitly asking for a microversioned API, a new endpoint v3
is added, which is identical to API version v2. This means that our new
Cinder API v3 would be the default, and consumers of the API that wished to
use a newer version could do so by using that endpoint and a microversion in
the HTTP header.
New tests for microversioned API features on endpoint /v3 should be added to
cinder/tests/unit/api/v3/ directory. Existing functionality will be tested via
the .../v2/ unit tests.

DocImpact
APIImpact
Implements: https://blueprints.launchpad.net/cinder/+spec/cinder-api-microversions
Change-Id: I48cdbbc900c2805e59ee9aebc3b1c64aed3212ae
2016-02-24 06:50:54 -07:00

443 lines
17 KiB
Python

# Copyright 2010 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.
import os
import re
import enum
from oslo_config import cfg
from oslo_log import log as logging
from six.moves import urllib
import webob
from cinder.api.openstack import wsgi
from cinder.api import xmlutil
from cinder.i18n import _
from cinder import utils
api_common_opts = [
cfg.IntOpt('osapi_max_limit',
default=1000,
help='The maximum number of items that a collection '
'resource returns in a single response'),
cfg.StrOpt('osapi_volume_base_URL',
help='Base URL that will be presented to users in links '
'to the OpenStack Volume API',
deprecated_name='osapi_compute_link_prefix'),
cfg.ListOpt('query_volume_filters',
default=['name', 'status', 'metadata',
'availability_zone',
'bootable'],
help="Volume filter options which "
"non-admin user could use to "
"query volumes. Default values "
"are: ['name', 'status', "
"'metadata', 'availability_zone' ,"
"'bootable']")
]
CONF = cfg.CONF
CONF.register_opts(api_common_opts)
LOG = logging.getLogger(__name__)
XML_NS_V1 = 'http://docs.openstack.org/api/openstack-block-storage/1.0/content'
XML_NS_V2 = 'http://docs.openstack.org/api/openstack-block-storage/2.0/content'
METADATA_TYPES = enum.Enum('METADATA_TYPES', 'user image')
# Regex that matches alphanumeric characters, periods, hyphens,
# colons and underscores:
# ^ assert position at start of the string
# [\w\.\-\:\_] match expression
# $ assert position at end of the string
VALID_KEY_NAME_REGEX = re.compile(r"^[\w\.\-\:\_]+$", re.UNICODE)
def validate_key_names(key_names_list):
"""Validate each item of the list to match key name regex."""
for key_name in key_names_list:
if not VALID_KEY_NAME_REGEX.match(key_name):
return False
return True
def get_pagination_params(params, max_limit=None):
"""Return marker, limit, offset tuple from request.
:param params: `wsgi.Request`'s GET dictionary, possibly containing
'marker', 'limit', and 'offset' variables. 'marker' is the
id of the last element the client has seen, 'limit' is the
maximum number of items to return and 'offset' is the number
of items to skip from the marker or from the first element.
If 'limit' is not specified, or > max_limit, we default to
max_limit. Negative values for either offset or limit will
cause exc.HTTPBadRequest() exceptions to be raised. If no
offset is present we'll default to 0 and if no marker is
present we'll default to None.
:max_limit: Max value 'limit' return value can take
:returns: Tuple (marker, limit, offset)
"""
max_limit = max_limit or CONF.osapi_max_limit
limit = _get_limit_param(params, max_limit)
marker = _get_marker_param(params)
offset = _get_offset_param(params)
return marker, limit, offset
def _get_limit_param(params, max_limit=None):
"""Extract integer limit from request's dictionary or fail.
Defaults to max_limit if not present and returns max_limit if present
'limit' is greater than max_limit.
"""
max_limit = max_limit or CONF.osapi_max_limit
try:
limit = int(params.pop('limit', max_limit))
except ValueError:
msg = _('limit param must be an integer')
raise webob.exc.HTTPBadRequest(explanation=msg)
if limit < 0:
msg = _('limit param must be positive')
raise webob.exc.HTTPBadRequest(explanation=msg)
limit = min(limit, max_limit)
return limit
def _get_marker_param(params):
"""Extract marker id from request's dictionary (defaults to None)."""
return params.pop('marker', None)
def _get_offset_param(params):
"""Extract offset id from request's dictionary (defaults to 0) or fail."""
try:
offset = int(params.pop('offset', 0))
except ValueError:
msg = _('offset param must be an integer')
raise webob.exc.HTTPBadRequest(explanation=msg)
if offset < 0:
msg = _('offset param must be positive')
raise webob.exc.HTTPBadRequest(explanation=msg)
return offset
def limited(items, request, max_limit=None):
"""Return a slice of items according to requested offset and limit.
:param items: A sliceable entity
:param request: ``wsgi.Request`` possibly containing 'offset' and 'limit'
GET variables. 'offset' is where to start in the list,
and 'limit' is the maximum number of items to return. If
'limit' is not specified, 0, or > max_limit, we default
to max_limit. Negative values for either offset or limit
will cause exc.HTTPBadRequest() exceptions to be raised.
:kwarg max_limit: The maximum number of items to return from 'items'
"""
max_limit = max_limit or CONF.osapi_max_limit
marker, limit, offset = get_pagination_params(request.GET.copy(),
max_limit)
range_end = offset + (limit or max_limit)
return items[offset:range_end]
def limited_by_marker(items, request, max_limit=None):
"""Return a slice of items according to the requested marker and limit."""
max_limit = max_limit or CONF.osapi_max_limit
marker, limit, __ = get_pagination_params(request.GET.copy(), max_limit)
start_index = 0
if marker:
start_index = -1
for i, item in enumerate(items):
if 'flavorid' in item:
if item['flavorid'] == marker:
start_index = i + 1
break
elif item['id'] == marker or item.get('uuid') == marker:
start_index = i + 1
break
if start_index < 0:
msg = _('marker [%s] not found') % marker
raise webob.exc.HTTPBadRequest(explanation=msg)
range_end = start_index + limit
return items[start_index:range_end]
def get_sort_params(params, default_key='created_at', default_dir='desc'):
"""Retrieves sort keys/directions parameters.
Processes the parameters to create a list of sort keys and sort directions
that correspond to either the 'sort' parameter or the 'sort_key' and
'sort_dir' parameter values. The value of the 'sort' parameter is a comma-
separated list of sort keys, each key is optionally appended with
':<sort_direction>'.
Note that the 'sort_key' and 'sort_dir' parameters are deprecated in kilo
and an exception is raised if they are supplied with the 'sort' parameter.
The sort parameters are removed from the request parameters by this
function.
:param params: webob.multidict of request parameters (from
cinder.api.openstack.wsgi.Request.params)
:param default_key: default sort key value, added to the list if no
sort keys are supplied
:param default_dir: default sort dir value, added to the list if the
corresponding key does not have a direction
specified
:returns: list of sort keys, list of sort dirs
:raise webob.exc.HTTPBadRequest: If both 'sort' and either 'sort_key' or
'sort_dir' are supplied parameters
"""
if 'sort' in params and ('sort_key' in params or 'sort_dir' in params):
msg = _("The 'sort_key' and 'sort_dir' parameters are deprecated and "
"cannot be used with the 'sort' parameter.")
raise webob.exc.HTTPBadRequest(explanation=msg)
sort_keys = []
sort_dirs = []
if 'sort' in params:
for sort in params.pop('sort').strip().split(','):
sort_key, _sep, sort_dir = sort.partition(':')
if not sort_dir:
sort_dir = default_dir
sort_keys.append(sort_key.strip())
sort_dirs.append(sort_dir.strip())
else:
sort_key = params.pop('sort_key', default_key)
sort_dir = params.pop('sort_dir', default_dir)
sort_keys.append(sort_key.strip())
sort_dirs.append(sort_dir.strip())
return sort_keys, sort_dirs
def get_request_url(request):
url = request.application_url
headers = request.headers
forwarded = headers.get('X-Forwarded-Host')
if forwarded:
url_parts = list(urllib.parse.urlsplit(url))
url_parts[1] = re.split(',\s?', forwarded)[-1]
url = urllib.parse.urlunsplit(url_parts).rstrip('/')
return url
def remove_version_from_href(href):
"""Removes the first api version from the href.
Given: 'http://www.cinder.com/v1.1/123'
Returns: 'http://www.cinder.com/123'
Given: 'http://www.cinder.com/v1.1'
Returns: 'http://www.cinder.com'
"""
parsed_url = urllib.parse.urlsplit(href)
url_parts = parsed_url.path.split('/', 2)
# NOTE: this should match vX.X or vX
expression = re.compile(r'^v([0-9]+|[0-9]+\.[0-9]+)(/.*|$)')
if expression.match(url_parts[1]):
del url_parts[1]
new_path = '/'.join(url_parts)
if new_path == parsed_url.path:
msg = 'href %s does not contain version' % href
LOG.debug(msg)
raise ValueError(msg)
parsed_url = list(parsed_url)
parsed_url[2] = new_path
return urllib.parse.urlunsplit(parsed_url)
class ViewBuilder(object):
"""Model API responses as dictionaries."""
_collection_name = None
def _get_links(self, request, identifier):
return [{"rel": "self",
"href": self._get_href_link(request, identifier), },
{"rel": "bookmark",
"href": self._get_bookmark_link(request, identifier), }]
def _get_next_link(self, request, identifier, collection_name):
"""Return href string with proper limit and marker params."""
params = request.params.copy()
params["marker"] = identifier
prefix = self._update_link_prefix(get_request_url(request),
CONF.osapi_volume_base_URL)
url = os.path.join(prefix,
request.environ["cinder.context"].project_id,
collection_name)
return "%s?%s" % (url, urllib.parse.urlencode(params))
def _get_href_link(self, request, identifier):
"""Return an href string pointing to this object."""
prefix = self._update_link_prefix(get_request_url(request),
CONF.osapi_volume_base_URL)
return os.path.join(prefix,
request.environ["cinder.context"].project_id,
self._collection_name,
str(identifier))
def _get_bookmark_link(self, request, identifier):
"""Create a URL that refers to a specific resource."""
base_url = remove_version_from_href(get_request_url(request))
base_url = self._update_link_prefix(base_url,
CONF.osapi_volume_base_URL)
return os.path.join(base_url,
request.environ["cinder.context"].project_id,
self._collection_name,
str(identifier))
def _get_collection_links(self, request, items, collection_name,
item_count=None, id_key="uuid"):
"""Retrieve 'next' link, if applicable.
The next link is included if we are returning as many items as we can,
given the restrictions of limit optional request parameter and
osapi_max_limit configuration parameter as long as we are returning
some elements.
So we return next link if:
1) 'limit' param is specified and equal to the number of items.
2) 'limit' param is NOT specified and the number of items is
equal to CONF.osapi_max_limit.
:param request: API request
:param items: List of collection items
:param collection_name: Name of collection, used to generate the
next link for a pagination query
:param item_count: Length of the list of the original collection
items
:param id_key: Attribute key used to retrieve the unique ID, used
to generate the next link marker for a pagination query
:returns: links
"""
item_count = item_count or len(items)
limit = _get_limit_param(request.GET.copy())
if len(items) and limit <= item_count:
return self._generate_next_link(items, id_key, request,
collection_name)
return []
def _generate_next_link(self, items, id_key, request,
collection_name):
links = []
last_item = items[-1]
if id_key in last_item:
last_item_id = last_item[id_key]
else:
last_item_id = last_item["id"]
links.append({
"rel": "next",
"href": self._get_next_link(request, last_item_id,
collection_name),
})
return links
def _update_link_prefix(self, orig_url, prefix):
if not prefix:
return orig_url
url_parts = list(urllib.parse.urlsplit(orig_url))
prefix_parts = list(urllib.parse.urlsplit(prefix))
url_parts[0:2] = prefix_parts[0:2]
url_parts[2] = prefix_parts[2] + url_parts[2]
return urllib.parse.urlunsplit(url_parts).rstrip('/')
class MetadataDeserializer(wsgi.MetadataXMLDeserializer):
def deserialize(self, text):
dom = utils.safe_minidom_parse_string(text)
metadata_node = self.find_first_child_named(dom, "metadata")
metadata = self.extract_metadata(metadata_node)
return {'body': {'metadata': metadata}}
class MetaItemDeserializer(wsgi.MetadataXMLDeserializer):
def deserialize(self, text):
dom = utils.safe_minidom_parse_string(text)
metadata_item = self.extract_metadata(dom)
return {'body': {'meta': metadata_item}}
class MetadataXMLDeserializer(wsgi.XMLDeserializer):
def extract_metadata(self, metadata_node):
"""Marshal the metadata attribute of a parsed request."""
if metadata_node is None:
return {}
metadata = {}
for meta_node in self.find_children_named(metadata_node, "meta"):
key = meta_node.getAttribute("key")
metadata[key] = self.extract_text(meta_node)
return metadata
def _extract_metadata_container(self, datastring):
dom = utils.safe_minidom_parse_string(datastring)
metadata_node = self.find_first_child_named(dom, "metadata")
metadata = self.extract_metadata(metadata_node)
return {'body': {'metadata': metadata}}
def create(self, datastring):
return self._extract_metadata_container(datastring)
def update_all(self, datastring):
return self._extract_metadata_container(datastring)
def update(self, datastring):
dom = utils.safe_minidom_parse_string(datastring)
metadata_item = self.extract_metadata(dom)
return {'body': {'meta': metadata_item}}
metadata_nsmap = {None: xmlutil.XMLNS_V11}
class MetaItemTemplate(xmlutil.TemplateBuilder):
def construct(self):
sel = xmlutil.Selector('meta', xmlutil.get_items, 0)
root = xmlutil.TemplateElement('meta', selector=sel)
root.set('key', 0)
root.text = 1
return xmlutil.MasterTemplate(root, 1, nsmap=metadata_nsmap)
class MetadataTemplateElement(xmlutil.TemplateElement):
def will_render(self, datum):
return True
class MetadataTemplate(xmlutil.TemplateBuilder):
def construct(self):
root = MetadataTemplateElement('metadata', selector='metadata')
elem = xmlutil.SubTemplateElement(root, 'meta',
selector=xmlutil.get_items)
elem.set('key', 0)
elem.text = 1
return xmlutil.MasterTemplate(root, 1, nsmap=metadata_nsmap)