Rework caching

Replace caching solution:
- cache on the proxy layer (API communication)
- support caching all GET calls
- add possibility to bypass cache (important for _wait_for_ operations)

Cheery-Picked-From: https://review.opendev.org/c/openstack/openstacksdk/+/805851

Change-Id: I2c8ae2c59d15c750ea8ebd3031ffdd2ced2421ed

doc/source/user/config/configuration.rst

@@ -223,11 +223,32 @@ Different cloud behaviors are also differently expensive to deal with. If you
 want to get really crazy and tweak stuff, you can specify different expiration
 times on a per-resource basis by passing values, in seconds to an expiration
 mapping keyed on the singular name of the resource. A value of `-1` indicates
-that the resource should never expire.
+that the resource should never expire. Not specifying a value (same as
+specifying `0`) indicates that no caching for this resource should be done.
+`openstacksdk` only caches `GET` request responses for the queries which have
+non-zero expiration time defined. Caching key contains url and request
+parameters, therefore no collisions are expected.
 
-`openstacksdk` does not actually cache anything itself, but it collects
-and presents the cache information so that your various applications that
-are connecting to OpenStack can share a cache should you desire.
+The expiration time key is constructed (joined with `.`) in the same way as the
+metrics are emmited:
+
+* service type
+* meaningful resource url segments (i.e. `/servers` results in `servers`,
+  `/servers/ID` results in `server`, `/servers/ID/metadata/KEY` results in
+  `server.metadata`
+
+Non `GET` requests cause cache invalidation based on the caching key prefix so
+that i.e. `PUT` request to `/images/ID` will invalidate all images cache (list
+and all individual entries). Moreover it is possible to explicitly pass
+`_sdk_skip_cache` parameter to the `proxy._get` function to bypass cache and
+invalidate what is already there. This is happening automatically in the
+`wait_for_status` methods where it is expected that resource is going to change
+some of the attributes over the time. Forcing complete cache invalidation can
+be achieved calling `conn._cache.invalidate`.
+
+`openstacksdk` does not actually cache anything itself, but it collects and
+presents the cache information so that your various applications that are
+connecting to OpenStack can share a cache should you desire.
 
 .. code-block:: yaml
 
@@ -240,6 +261,9 @@ are connecting to OpenStack can share a cache should you desire.
     expiration:
       server: 5
       flavor: -1
+      compute.servers: 5
+      compute.flavors: -1
+      image.images: 5
   clouds:
     mtvexx:
       profile: vexxhost

openstack/cloud/openstackcloud.py

@@ -102,26 +102,15 @@ class _OpenStackCloudMixin:
         cache_class = self.config.get_cache_class()
         cache_arguments = self.config.get_cache_arguments()
 
-        self._resource_caches = {}
+        self._cache_expirations = dict()
 
         if cache_class != 'dogpile.cache.null':
             self.cache_enabled = True
-            self._cache = self._make_cache(
-                cache_class, cache_expiration_time, cache_arguments)
-            expirations = self.config.get_cache_expirations()
-            for expire_key in expirations.keys():
-                # Only build caches for things we have list operations for
-                if getattr(
-                        self, 'list_{0}'.format(expire_key), None):
-                    self._resource_caches[expire_key] = self._make_cache(
-                        cache_class, expirations[expire_key], cache_arguments)
-
-            self._SERVER_AGE = DEFAULT_SERVER_AGE
-            self._PORT_AGE = DEFAULT_PORT_AGE
-            self._FLOAT_AGE = DEFAULT_FLOAT_AGE
         else:
             self.cache_enabled = False
 
+            # TODO(gtema): delete it with the standalone cloud layer caching
+
             def _fake_invalidate(unused):
                 pass
 
@@ -148,15 +137,20 @@ class _OpenStackCloudMixin:
                     new_func.invalidate = _fake_invalidate
                     setattr(self, method, new_func)
 
-        # If server expiration time is set explicitly, use that. Otherwise
-        # fall back to whatever it was before
-        self._SERVER_AGE = self.config.get_cache_resource_expiration(
-            'server', self._SERVER_AGE)
-        self._PORT_AGE = self.config.get_cache_resource_expiration(
-            'port', self._PORT_AGE)
-        self._FLOAT_AGE = self.config.get_cache_resource_expiration(
-            'floating_ip', self._FLOAT_AGE)
+        # Uncoditionally create cache even with a "null" backend
+        self._cache = self._make_cache(
+            cache_class, cache_expiration_time, cache_arguments)
+        expirations = self.config.get_cache_expirations()
+        for expire_key in expirations.keys():
+            self._cache_expirations[expire_key] = \
+                expirations[expire_key]
 
+        # TODO(gtema): delete in next change
+        self._SERVER_AGE = 0
+        self._PORT_AGE = 0
+        self._FLOAT_AGE = 0
+
+        self._api_cache_keys = set()
         self._container_cache = dict()
         self._file_hash_cache = dict()
 

openstack/compute/v2/limits.py

@@ -91,7 +91,7 @@ class Limits(resource.Resource):
     rate = resource.Body("rate", type=list, list_type=RateLimit)
 
     def fetch(self, session, requires_id=False, error_message=None,
-              base_path=None, **params):
+              base_path=None, skip_cache=False, **params):
         """Get the Limits resource.
 
         :param session: The session to use for making this request.
@@ -106,5 +106,5 @@ class Limits(resource.Resource):
             session=session, requires_id=requires_id,
             error_message=error_message,
             base_path=base_path,
-            **params
-        )
+            skip_cache=skip_cache,
+            **params)

openstack/key_manager/v1/secret.py

@@ -78,7 +78,7 @@ class Secret(resource.Resource):
     payload_content_encoding = resource.Body('payload_content_encoding')
 
     def fetch(self, session, requires_id=True,
-              base_path=None, error_message=None):
+              base_path=None, error_message=None, skip_cache=False):
         request = self._prepare_request(requires_id=requires_id,
                                         base_path=base_path)
 
@@ -93,8 +93,10 @@ class Secret(resource.Resource):
         # Only try to get the payload if a content type has been explicitly
         # specified or if one was found in the metadata response
         if content_type is not None:
-            payload = session.get(utils.urljoin(request.url, "payload"),
-                                  headers={"Accept": content_type})
+            payload = session.get(
+                utils.urljoin(request.url, "payload"),
+                headers={"Accept": content_type},
+                skip_cache=skip_cache)
             response["payload"] = payload.text
 
         # We already have the JSON here so don't call into _translate_response

openstack/message/v2/claim.py

@@ -82,7 +82,7 @@ class Claim(resource.Resource):
         return self
 
     def fetch(self, session, requires_id=True,
-              base_path=None, error_message=None):
+              base_path=None, error_message=None, skip_cache=False):
         request = self._prepare_request(requires_id=requires_id,
                                         base_path=base_path)
         headers = {
@@ -91,8 +91,8 @@ class Claim(resource.Resource):
         }
 
         request.headers.update(headers)
-        response = session.get(request.url,
-                               headers=request.headers)
+        response = session.get(
+            request.url, headers=request.headers, skip_cache=False)
         self._translate_response(response)
 
         return self

openstack/message/v2/message.py

@@ -112,7 +112,7 @@ class Message(resource.Resource):
             query_params["marker"] = new_marker
 
     def fetch(self, session, requires_id=True,
-              base_path=None, error_message=None):
+              base_path=None, error_message=None, skip_cache=False):
         request = self._prepare_request(requires_id=requires_id,
                                         base_path=base_path)
         headers = {
@@ -122,7 +122,8 @@ class Message(resource.Resource):
 
         request.headers.update(headers)
         response = session.get(request.url,
-                               headers=headers)
+                               headers=headers,
+                               skip_cache=skip_cache)
         self._translate_response(response)
 
         return self

openstack/message/v2/queue.py

@@ -111,7 +111,7 @@ class Queue(resource.Resource):
             query_params["marker"] = new_marker
 
     def fetch(self, session, requires_id=True,
-              base_path=None, error_message=None):
+              base_path=None, error_message=None, skip_cache=False):
         request = self._prepare_request(requires_id=requires_id,
                                         base_path=base_path)
         headers = {
@@ -119,8 +119,8 @@ class Queue(resource.Resource):
             "X-PROJECT-ID": self.project_id or session.get_project_id()
         }
         request.headers.update(headers)
-        response = session.get(request.url,
-                               headers=headers)
+        response = session.get(
+            request.url, headers=headers, skip_cache=skip_cache)
         self._translate_response(response)
 
         return self

openstack/message/v2/subscription.py

@@ -119,7 +119,7 @@ class Subscription(resource.Resource):
             query_params["marker"] = new_marker
 
     def fetch(self, session, requires_id=True,
-              base_path=None, error_message=None):
+              base_path=None, error_message=None, skip_cache=False):
         request = self._prepare_request(requires_id=requires_id,
                                         base_path=base_path)
         headers = {
@@ -128,8 +128,8 @@ class Subscription(resource.Resource):
         }
 
         request.headers.update(headers)
-        response = session.get(request.url,
-                               headers=request.headers)
+        response = session.get(
+            request.url, headers=request.headers, skip_cache=skip_cache)
         self._translate_response(response)
 
         return self

openstack/object_store/v1/info.py

@@ -33,8 +33,10 @@ class Info(resource.Resource):
     staticweb = resource.Body("staticweb", type=dict)
     tempurl = resource.Body("tempurl", type=dict)
 
-    def fetch(self, session, requires_id=False,
-              base_path=None, error_message=None):
+    def fetch(
+        self, session, requires_id=False,
+        base_path=None, skip_cache=False, error_message=None
+    ):
         """Get a remote resource based on this instance.
 
         :param session: The session to use for making this request.

openstack/orchestration/v1/stack.py

@@ -168,7 +168,8 @@ class Stack(resource.Resource):
         return resp.json()
 
     def fetch(self, session, requires_id=True,
-              base_path=None, error_message=None, resolve_outputs=True):
+              base_path=None, error_message=None,
+              skip_cache=False, resolve_outputs=True):
 
         if not self.allow_fetch:
             raise exceptions.MethodNotSupported(self, "fetch")
@@ -183,7 +184,8 @@ class Stack(resource.Resource):
         # apply parameters again, what results in them being set doubled
         if not resolve_outputs:
             request.url = request.url + '?resolve_outputs=False'
-        response = session.get(request.url, microversion=microversion)
+        response = session.get(
+            request.url, microversion=microversion, skip_cache=skip_cache)
         kwargs = {}
         if error_message:
             kwargs['error_message'] = error_message

openstack/proxy.py

@@ -10,6 +10,7 @@
 # License for the specific language governing permissions and limitations
 # under the License.
 
+import functools
 import urllib
 from urllib.parse import urlparse
 
@@ -84,21 +85,79 @@ class Proxy(adapter.Adapter):
             log_name = 'openstack'
         self.log = _log.setup_logging(log_name)
 
+    def _get_cache_key_prefix(self, url):
+        """Calculate cache prefix for the url"""
+        name_parts = self._extract_name(
+            url, self.service_type,
+            self.session.get_project_id())
+
+        return '.'.join(
+            [self.service_type]
+            + name_parts)
+
+    def _invalidate_cache(self, conn, key_prefix):
+        """Invalidate all cache entries starting with given prefix"""
+        for k in set(conn._api_cache_keys):
+            if k.startswith(key_prefix):
+                conn._cache.delete(k)
+                conn._api_cache_keys.remove(k)
+
     def request(
             self, url, method, error_message=None,
             raise_exc=False, connect_retries=1,
-            global_request_id=None, *args, **kwargs):
+            global_request_id=None,
+            *args, **kwargs):
+        conn = self._get_connection()
         if not global_request_id:
-            conn = self._get_connection()
-            if conn:
-                # Per-request setting should take precedence
-                global_request_id = conn._global_request_id
+            # Per-request setting should take precedence
+            global_request_id = conn._global_request_id
+
+        key = None
+        key_prefix = self._get_cache_key_prefix(url)
+        # The caller might want to force cache bypass.
+        skip_cache = kwargs.pop('skip_cache', False)
+        if conn.cache_enabled:
+            # Construct cache key. It consists of:
+            # service.name_parts.URL.str(kwargs)
+            key = '.'.join(
+                [key_prefix, url, str(kwargs)]
+            )
+
+            # Track cache key for invalidating possibility
+            conn._api_cache_keys.add(key)
+
         try:
-            response = super(Proxy, self).request(
-                url, method,
-                connect_retries=connect_retries, raise_exc=raise_exc,
-                global_request_id=global_request_id,
-                **kwargs)
+            if conn.cache_enabled and not skip_cache and method == 'GET':
+                # Get the object expiration time from config
+                # default to 0 to disable caching for this resource type
+                expiration_time = int(
+                    conn._cache_expirations.get(key_prefix, 0))
+                # Get from cache or execute and cache
+                response = conn._cache.get_or_create(
+                    key=key,
+                    creator=super(Proxy, self).request,
+                    creator_args=(
+                        [url, method],
+                        dict(
+                            connect_retries=connect_retries,
+                            raise_exc=raise_exc,
+                            global_request_id=global_request_id,
+                            **kwargs
+                        )
+                    ),
+                    expiration_time=expiration_time
+                )
+            else:
+                # invalidate cache if we send modification request or user
+                # asked for cache bypass
+                self._invalidate_cache(conn, key_prefix)
+                # Pass through the API request bypassing cache
+                response = super(Proxy, self).request(
+                    url, method,
+                    connect_retries=connect_retries, raise_exc=raise_exc,
+                    global_request_id=global_request_id,
+                    **kwargs)
+
             for h in response.history:
                 self._report_stats(h)
             self._report_stats(response)
@@ -111,6 +170,7 @@ class Proxy(adapter.Adapter):
             self._report_stats(None, url, method, e)
             raise
 
+    @functools.lru_cache(maxsize=256)
     def _extract_name(self, url, service_type=None, project_id=None):
         '''Produce a key name to use in logging/metrics from the URL path.
 
@@ -484,7 +544,7 @@ class Proxy(adapter.Adapter):
 
     @_check_resource(strict=False)
     def _get(self, resource_type, value=None, requires_id=True,
-             base_path=None, **attrs):
+             base_path=None, skip_cache=False, **attrs):
         """Fetch a resource
 
         :param resource_type: The type of resource to get.
@@ -495,6 +555,8 @@ class Proxy(adapter.Adapter):
         :param str base_path: Base part of the URI for fetching resources, if
             different from
             :data:`~openstack.resource.Resource.base_path`.
+        :param bool skip_cache: A boolean indicating whether optional API
+            cache should be skipped for this invocation.
         :param dict attrs: Attributes to be passed onto the
             :meth:`~openstack.resource.Resource.get`
             method. These should correspond
@@ -509,6 +571,7 @@ class Proxy(adapter.Adapter):
 
         return res.fetch(
             self, requires_id=requires_id, base_path=base_path,
+            skip_cache=skip_cache,
             error_message="No {resource_type} found for {value}".format(
                 resource_type=resource_type.__name__, value=value))
 

openstack/resource.py

@@ -1480,6 +1480,7 @@ class Resource(dict):
         requires_id=True,
         base_path=None,
         error_message=None,
+        skip_cache=False,
         **params,
     ):
         """Get a remote resource based on this instance.
@@ -1492,6 +1493,8 @@ class Resource(dict):
             different from :data:`~openstack.resource.Resource.base_path`.
         :param str error_message: An Error message to be returned if
             requested object does not exist.
+        :param bool skip_cache: A boolean indicating whether optional API
+            cache should be skipped for this invocation.
         :param dict params: Additional parameters that can be consumed.
         :return: This :class:`Resource` instance.
         :raises: :exc:`~openstack.exceptions.MethodNotSupported` if
@@ -1507,7 +1510,8 @@ class Resource(dict):
         session = self._get_session(session)
         microversion = self._get_microversion_for(session, 'fetch')
         response = session.get(request.url, microversion=microversion,
-                               params=params)
+                               params=params,
+                               skip_cache=skip_cache)
         kwargs = {}
         if error_message:
             kwargs['error_message'] = error_message
@@ -2078,7 +2082,7 @@ def wait_for_status(
             timeout=wait,
             message=msg,
             wait=interval):
-        resource = resource.fetch(session)
+        resource = resource.fetch(session, skip_cache=True)
 
         if not resource:
             raise exceptions.ResourceFailure(
@@ -2120,7 +2124,7 @@ def wait_for_delete(session, resource, interval, wait):
                 id=resource.id),
             wait=interval):
         try:
-            resource = resource.fetch(session)
+            resource = resource.fetch(session, skip_cache=True)
             if not resource:
                 return orig_resource
             if resource.status.lower() == 'deleted':

openstack/tests/unit/cloud/test_caching.py

@@ -535,12 +535,13 @@ class TestMemoryCache(base.TestCase):
         down_port = test_port.TestPort.mock_neutron_port_create_rep['port']
         active_port = down_port.copy()
         active_port['status'] = 'ACTIVE'
-        # We're testing to make sure a query string isn't passed when we're
-        # caching, but that the results are still filtered.
+        # We're testing to make sure a query string is passed when we're
+        # caching (cache by url), and that the results are still filtered.
         self.register_uris([
             dict(method='GET',
                  uri=self.get_mock_url(
-                     'network', 'public', append=['v2.0', 'ports']),
+                     'network', 'public', append=['v2.0', 'ports'],
+                     qs_elements=['status=DOWN']),
                  json={'ports': [
                      down_port,
                      active_port,

openstack/tests/unit/common/test_quota_set.py

@@ -89,7 +89,8 @@ class TestQuotaSet(base.TestCase):
         self.sess.get.assert_called_with(
             '/os-quota-sets/proj',
             microversion=1,
-            params={})
+            params={},
+            skip_cache=False)
 
         self.assertEqual(BASIC_EXAMPLE['backups'], sot.backups)
         self.assertEqual({}, sot.reservation)
@@ -110,7 +111,8 @@ class TestQuotaSet(base.TestCase):
         self.sess.get.assert_called_with(
             '/os-quota-sets/proj',
             microversion=1,
-            params={'usage': True})
+            params={'usage': True},
+            skip_cache=False)
 
         self.assertEqual(
             USAGE_EXAMPLE['backups']['limit'],

openstack/tests/unit/image/v2/test_image.py

@@ -407,7 +407,8 @@ class TestImage(base.TestCase):
         self.sess.get.assert_has_calls(
             [mock.call('images/IDENTIFIER/file',
                        stream=False),
-             mock.call('images/IDENTIFIER', microversion=None, params={})])
+             mock.call('images/IDENTIFIER', microversion=None, params={},
+                       skip_cache=False)])
 
         self.assertEqual(rv, resp1)
 
@@ -436,7 +437,8 @@ class TestImage(base.TestCase):
         self.sess.get.assert_has_calls(
             [mock.call('images/IDENTIFIER/file',
                        stream=False),
-             mock.call('images/IDENTIFIER', microversion=None, params={})])
+             mock.call('images/IDENTIFIER', microversion=None, params={},
+                       skip_cache=False)])
 
         self.assertEqual(rv, resp1)
 
@@ -536,7 +538,7 @@ class TestImage(base.TestCase):
 
         self.sess.get.assert_has_calls([
             mock.call('images/' + EXAMPLE['name'], microversion=None,
-                      params={}),
+                      params={}, skip_cache=False),
             mock.call('/images', headers={'Accept': 'application/json'},
                       microversion=None, params={'name': EXAMPLE['name']}),
             mock.call('/images', headers={'Accept': 'application/json'},

openstack/tests/unit/key_manager/v1/test_secret.py

@@ -113,8 +113,10 @@ class TestSecret(base.TestCase):
 
         sess.get.assert_has_calls(
             [mock.call("secrets/id",),
-             mock.call("secrets/id/payload",
-                       headers={"Accept": content_type})])
+             mock.call(
+                 "secrets/id/payload",
+                 headers={"Accept": content_type},
+                 skip_cache=False)])
 
         self.assertEqual(rv.payload, payload)
         self.assertEqual(rv.status, metadata["status"])

openstack/tests/unit/message/v2/test_claim.py

@@ -141,8 +141,8 @@ class TestClaim(base.TestCase):
             "queue": FAKE1["queue_name"], "claim": FAKE1["id"]}
         headers = {"Client-ID": "NEW_CLIENT_ID",
                    "X-PROJECT-ID": "NEW_PROJECT_ID"}
-        sess.get.assert_called_with(url,
-                                    headers=headers)
+        sess.get.assert_called_with(
+            url, headers=headers, skip_cache=False)
         sess.get_project_id.assert_called_once_with()
         sot._translate_response.assert_called_once_with(resp)
         self.assertEqual(sot, res)
@@ -160,8 +160,8 @@ class TestClaim(base.TestCase):
             "queue": FAKE2["queue_name"], "claim": FAKE2["id"]}
         headers = {"Client-ID": "OLD_CLIENT_ID",
                    "X-PROJECT-ID": "OLD_PROJECT_ID"}
-        sess.get.assert_called_with(url,
-                                    headers=headers)
+        sess.get.assert_called_with(
+            url, headers=headers, skip_cache=False)
         sot._translate_response.assert_called_once_with(resp)
         self.assertEqual(sot, res)
 

openstack/tests/unit/message/v2/test_message.py

@@ -151,8 +151,8 @@ class TestMessage(base.TestCase):
             'queue': FAKE1['queue_name'], 'message': FAKE1['id']}
         headers = {'Client-ID': 'NEW_CLIENT_ID',
                    'X-PROJECT-ID': 'NEW_PROJECT_ID'}
-        sess.get.assert_called_with(url,
-                                    headers=headers)
+        sess.get.assert_called_with(
+            url, headers=headers, skip_cache=False)
         sess.get_project_id.assert_called_once_with()
         sot._translate_response.assert_called_once_with(resp)
         self.assertEqual(sot, res)
@@ -173,8 +173,8 @@ class TestMessage(base.TestCase):
         res = sot.fetch(sess)
         headers = {'Client-ID': 'OLD_CLIENT_ID',
                    'X-PROJECT-ID': 'OLD_PROJECT_ID'}
-        sess.get.assert_called_with(url,
-                                    headers=headers)
+        sess.get.assert_called_with(
+            url, headers=headers, skip_cache=False)
         sot._translate_response.assert_called_once_with(resp)
         self.assertEqual(sot, res)
 

openstack/tests/unit/message/v2/test_queue.py

@@ -109,8 +109,8 @@ class TestQueue(base.TestCase):
         url = 'queues/%s' % FAKE1['name']
         headers = {'Client-ID': 'NEW_CLIENT_ID',
                    'X-PROJECT-ID': 'NEW_PROJECT_ID'}
-        sess.get.assert_called_with(url,
-                                    headers=headers)
+        sess.get.assert_called_with(
+            url, headers=headers, skip_cache=False)
         sess.get_project_id.assert_called_once_with()
         sot._translate_response.assert_called_once_with(resp)
         self.assertEqual(sot, res)
@@ -127,8 +127,8 @@ class TestQueue(base.TestCase):
         url = 'queues/%s' % FAKE2['name']
         headers = {'Client-ID': 'OLD_CLIENT_ID',
                    'X-PROJECT-ID': 'OLD_PROJECT_ID'}
-        sess.get.assert_called_with(url,
-                                    headers=headers)
+        sess.get.assert_called_with(
+            url, headers=headers, skip_cache=False)
         sot._translate_response.assert_called_once_with(resp)
         self.assertEqual(sot, res)
 

openstack/tests/unit/message/v2/test_subscription.py

@@ -127,8 +127,8 @@ class TestSubscription(base.TestCase):
             "queue": FAKE1["queue_name"], "subscription": FAKE1["id"]}
         headers = {"Client-ID": "NEW_CLIENT_ID",
                    "X-PROJECT-ID": "NEW_PROJECT_ID"}
-        sess.get.assert_called_with(url,
-                                    headers=headers)
+        sess.get.assert_called_with(
+            url, headers=headers, skip_cache=False)
         sess.get_project_id.assert_called_once_with()
         sot._translate_response.assert_called_once_with(resp)
         self.assertEqual(sot, res)
@@ -146,8 +146,8 @@ class TestSubscription(base.TestCase):
             "queue": FAKE2["queue_name"], "subscription": FAKE2["id"]}
         headers = {"Client-ID": "OLD_CLIENT_ID",
                    "X-PROJECT-ID": "OLD_PROJECT_ID"}
-        sess.get.assert_called_with(url,
-                                    headers=headers)
+        sess.get.assert_called_with(
+            url, headers=headers, skip_cache=False)
         sot._translate_response.assert_called_once_with(resp)
         self.assertEqual(sot, res)
 

openstack/tests/unit/orchestration/v1/test_stack.py

@@ -232,11 +232,13 @@ class TestStack(base.TestCase):
         self.assertEqual(sot, sot.fetch(sess))
         sess.get.assert_called_with(
             'stacks/{id}'.format(id=sot.id),
-            microversion=None)
+            microversion=None,
+            skip_cache=False)
         sot.fetch(sess, resolve_outputs=False)
         sess.get.assert_called_with(
             'stacks/{id}?resolve_outputs=False'.format(id=sot.id),
-            microversion=None)
+            microversion=None,
+            skip_cache=False)
         ex = self.assertRaises(exceptions.ResourceNotFound, sot.fetch, sess)
         self.assertEqual('oops', str(ex))
         ex = self.assertRaises(exceptions.ResourceNotFound, sot.fetch, sess)

openstack/tests/unit/test_proxy.py

@@ -10,6 +10,7 @@
 # License for the specific language governing permissions and limitations
 # under the License.
 
+import copy
 import queue
 from unittest import mock
 
@@ -35,7 +36,7 @@ class CreateableResource(resource.Resource):
 
 
 class RetrieveableResource(resource.Resource):
-    allow_retrieve = True
+    allow_fetch = True
 
 
 class ListableResource(resource.Resource):
@@ -380,6 +381,7 @@ class TestProxyGet(base.TestCase):
         self.res.fetch.assert_called_with(
             self.sot, requires_id=True,
             base_path=None,
+            skip_cache=mock.ANY,
             error_message=mock.ANY)
         self.assertEqual(rv, self.fake_result)
 
@@ -390,6 +392,7 @@ class TestProxyGet(base.TestCase):
         self.res._update.assert_called_once_with(**args)
         self.res.fetch.assert_called_with(
             self.sot, requires_id=True, base_path=None,
+            skip_cache=mock.ANY,
             error_message=mock.ANY)
         self.assertEqual(rv, self.fake_result)
 
@@ -400,6 +403,7 @@ class TestProxyGet(base.TestCase):
             connection=self.cloud, id=self.fake_id)
         self.res.fetch.assert_called_with(
             self.sot, requires_id=True, base_path=None,
+            skip_cache=mock.ANY,
             error_message=mock.ANY)
         self.assertEqual(rv, self.fake_result)
 
@@ -412,6 +416,7 @@ class TestProxyGet(base.TestCase):
             connection=self.cloud, id=self.fake_id)
         self.res.fetch.assert_called_with(
             self.sot, requires_id=True, base_path=base_path,
+            skip_cache=mock.ANY,
             error_message=mock.ANY)
         self.assertEqual(rv, self.fake_result)
 
@@ -521,6 +526,116 @@ class TestExtractName(base.TestCase):
         self.assertEqual(self.parts, results)
 
 
+class TestProxyCache(base.TestCase):
+
+    class Res(resource.Resource):
+        base_path = 'fake'
+
+        allow_commit = True
+        allow_fetch = True
+
+        foo = resource.Body('foo')
+
+    def setUp(self):
+        super(TestProxyCache, self).setUp(
+            cloud_config_fixture='clouds_cache.yaml')
+
+        self.session = mock.Mock()
+        self.session._sdk_connection = self.cloud
+        self.session.get_project_id = mock.Mock(return_value='fake_prj')
+
+        self.response = mock.Mock()
+        self.response.status_code = 200
+        self.response.history = []
+        self.response.headers = {}
+        self.response.body = {}
+        self.response.json = mock.Mock(
+            return_value=self.response.body)
+        self.session.request = mock.Mock(
+            return_value=self.response)
+
+        self.sot = proxy.Proxy(self.session)
+        self.sot._connection = self.cloud
+        self.sot.service_type = 'srv'
+
+    def _get_key(self, id):
+        return (
+            f"srv.fake.fake/{id}."
+            "{'microversion': None, 'params': {}}")
+
+    def test_get_not_in_cache(self):
+        self.cloud._cache_expirations['srv.fake'] = 5
+        self.sot._get(self.Res, '1')
+
+        self.session.request.assert_called_with(
+            'fake/1',
+            'GET',
+            connect_retries=mock.ANY, raise_exc=mock.ANY,
+            global_request_id=mock.ANY,
+            endpoint_filter=mock.ANY,
+            headers=mock.ANY,
+            microversion=mock.ANY, params=mock.ANY
+        )
+        self.assertIn(
+            self._get_key(1),
+            self.cloud._api_cache_keys)
+
+    def test_get_from_cache(self):
+        key = self._get_key(2)
+
+        self.cloud._cache.set(key, self.response)
+        # set expiration for the resource to respect cache
+        self.cloud._cache_expirations['srv.fake'] = 5
+
+        self.sot._get(self.Res, '2')
+        self.session.request.assert_not_called()
+
+    def test_modify(self):
+        key = self._get_key(3)
+
+        self.cloud._cache.set(key, self.response)
+        self.cloud._api_cache_keys.add(key)
+        self.cloud._cache_expirations['srv.fake'] = 5
+
+        # Ensure first call gets value from cache
+        self.sot._get(self.Res, '3')
+        self.session.request.assert_not_called()
+
+        # update call invalidates the cache and triggers API
+        rs = self.Res.existing(id='3')
+        self.sot._update(self.Res, rs, foo='bar')
+
+        self.session.request.assert_called()
+        self.assertIsNotNone(self.cloud._cache.get(key))
+        self.assertEqual(
+            'NoValue',
+            type(self.cloud._cache.get(key)).__name__)
+        self.assertNotIn(key, self.cloud._api_cache_keys)
+
+        # next get call again triggers API
+        self.sot._get(self.Res, '3')
+        self.session.request.assert_called()
+
+    def test_get_bypass_cache(self):
+        key = self._get_key(4)
+
+        resp = copy.deepcopy(self.response)
+        resp.body = {'foo': 'bar'}
+        self.cloud._api_cache_keys.add(key)
+        self.cloud._cache.set(key, resp)
+        # set expiration for the resource to respect cache
+        self.cloud._cache_expirations['srv.fake'] = 5
+
+        self.sot._get(self.Res, '4', skip_cache=True)
+        self.session.request.assert_called()
+        # validate we got empty body as expected, and not what is in cache
+        self.assertEqual(dict(), self.response.body)
+        self.assertNotIn(key, self.cloud._api_cache_keys)
+        self.assertEqual(
+            'NoValue',
+            type(self.cloud._cache.get(key)).__name__)
+
+
 class TestProxyCleanup(base.TestCase):
 
     def setUp(self):

openstack/tests/unit/test_proxy_base.py

@@ -158,7 +158,8 @@ class TestProxyBase(base.TestCase):
             res.fetch.assert_called_once_with(
                 proxy, requires_id=True,
                 base_path=None,
-                error_message=mock.ANY)
+                error_message=mock.ANY,
+                skip_cache=False)
 
     def verify_head(
         self, test_method, resource_type, base_path=None, *,

openstack/tests/unit/test_resource.py

@@ -1620,7 +1620,8 @@ class TestResourceActions(base.TestCase):
         self.sot._prepare_request.assert_called_once_with(
             requires_id=True, base_path=None)
         self.session.get.assert_called_once_with(
-            self.request.url, microversion=None, params={})
+            self.request.url, microversion=None, params={},
+            skip_cache=False)
 
         self.assertIsNone(self.sot.microversion)
         self.sot._translate_response.assert_called_once_with(self.response)
@@ -1632,7 +1633,8 @@ class TestResourceActions(base.TestCase):
         self.sot._prepare_request.assert_called_once_with(
             requires_id=True, base_path=None)
         self.session.get.assert_called_once_with(
-            self.request.url, microversion=None, params={'fields': 'a,b'})
+            self.request.url, microversion=None, params={'fields': 'a,b'},
+            skip_cache=False)
 
         self.assertIsNone(self.sot.microversion)
         self.sot._translate_response.assert_called_once_with(self.response)
@@ -1654,7 +1656,8 @@ class TestResourceActions(base.TestCase):
         sot._prepare_request.assert_called_once_with(
             requires_id=True, base_path=None)
         self.session.get.assert_called_once_with(
-            self.request.url, microversion='1.42', params={})
+            self.request.url, microversion='1.42', params={},
+            skip_cache=False)
 
         self.assertEqual(sot.microversion, '1.42')
         sot._translate_response.assert_called_once_with(self.response)
@@ -1666,7 +1669,8 @@ class TestResourceActions(base.TestCase):
         self.sot._prepare_request.assert_called_once_with(
             requires_id=False, base_path=None)
         self.session.get.assert_called_once_with(
-            self.request.url, microversion=None, params={})
+            self.request.url, microversion=None, params={},
+            skip_cache=False)
 
         self.sot._translate_response.assert_called_once_with(self.response)
         self.assertEqual(result, self.sot)
@@ -1678,7 +1682,8 @@ class TestResourceActions(base.TestCase):
             requires_id=False,
             base_path='dummy')
         self.session.get.assert_called_once_with(
-            self.request.url, microversion=None, params={})
+            self.request.url, microversion=None, params={},
+            skip_cache=False)
 
         self.sot._translate_response.assert_called_once_with(self.response)
         self.assertEqual(result, self.sot)

releasenotes/notes/basic-api-cache-4ad8cf2754b004d1.yaml

@@ -0,0 +1,4 @@
+---
+features:
+  - |
+    Add possibility to cache GET requests using dogpile cache.