Fix Bug#131: 0 rows returns when server return 429 on first page of results.

azure/cosmos/execution_context/base_execution_context.py

@@ -135,13 +135,7 @@ def _fetch_items_helper_no_retries(self, fetch_function):
             if fetched_items:
                 break
         return fetched_items   
-
-    def _fetch_items_helper_with_retries(self, fetch_function):
-        def callback():
-            return self._fetch_items_helper_no_retries(fetch_function)
-
-        return retry_utility._Execute(self._client, self._client._global_endpoint_manager, callback)
-
+
 
 class _DefaultQueryExecutionContext(_QueryExecutionContextBase):
     """
@@ -166,7 +160,7 @@ def __init__(self, client, options, fetch_function):
 
     def _fetch_next_block(self):
         while super(_DefaultQueryExecutionContext, self)._has_more_pages() and len(self._buffer) == 0:
-            return self._fetch_items_helper_with_retries(self._fetch_function)
+            return self._fetch_items_helper_no_retries(self._fetch_function)
 
 class _MultiCollectionQueryExecutionContext(_QueryExecutionContextBase):
     """
@@ -235,7 +229,7 @@ def _fetch_next_block(self):
         :rtype: list
         """
         # Fetch next block of results by executing the query against the current document collection
-        fetched_items = self._fetch_items_helper_with_retries(self._fetch_function)
+        fetched_items = self._fetch_items_helper_no_retries(self._fetch_function)
 
         # If there are multiple document collections to query for(in case of partitioning), keep looping through each one of them,
         # creating separate feed queries for each collection and fetching the items
@@ -255,7 +249,7 @@ def fetch_fn(options):
 
                 self._fetch_function = fetch_fn
 
-                fetched_items = self._fetch_items_helper_with_retries(self._fetch_function)
+                fetched_items = self._fetch_items_helper_no_retries(self._fetch_function)
                 self._current_collection_index += 1
             else:
                 break

azure/cosmos/retry_options.py

@@ -32,7 +32,7 @@ class RetryOptions(object):
     :ivar int MaxWaitTimeInSeconds:
         Max wait time in seconds to wait for a request while the retries are happening. Default value 30 seconds.
     """
-    def __init__(self, max_retry_attempt_count = 9, fixed_retry_interval_in_milliseconds = None, max_wait_time_in_seconds = 30):
+    def __init__(self, max_retry_attempt_count = 17, fixed_retry_interval_in_milliseconds = None, max_wait_time_in_seconds = 60):
         self._max_retry_attempt_count = max_retry_attempt_count
         self._fixed_retry_interval_in_milliseconds = fixed_retry_interval_in_milliseconds
         self._max_wait_time_in_seconds = max_wait_time_in_seconds
@@ -47,4 +47,4 @@ def FixedRetryIntervalInMilliseconds(self):
 
     @property
     def MaxWaitTimeInSeconds(self):
-        return self._max_wait_time_in_seconds
+        return self._max_wait_time_in_seconds

test/retry_policy_tests.py

@@ -1,4 +1,4 @@
-#The MIT License (MIT)
+#The MIT License (MIT)
 #Copyright (c) 2014 Microsoft Corporation
 
 #Permission is hereby granted, free of charge, to any person obtaining a copy
@@ -222,7 +222,7 @@ def test_default_retry_policy_for_query(self):
         result_docs = list(docs)
         self.assertEqual(result_docs[0]['id'], 'doc1')
         self.assertEqual(result_docs[1]['id'], 'doc2')
-        self.assertEqual(self.counter, 12)
+        self.assertEqual(self.counter, 6)
 
         self.counter = 0
         retry_utility._ExecuteFunction = self.OriginalExecuteFunction
@@ -278,6 +278,38 @@ def test_default_retry_policy_for_create(self):
 
         retry_utility._ExecuteFunction = self.OriginalExecuteFunction
 
+    def test_429_on_first_page(self):
+        client = document_client.DocumentClient(Test_retry_policy_tests.host, {'masterKey': Test_retry_policy_tests.masterKey})
+
+        document_definition = { 'id': 'doc429',
+                                'name': 'sample document',
+                                'key': 'value'}
+
+        created_document = client.CreateDocument(self.created_collection['_self'], document_definition)
+
+        # Mock an overloaded server which always returns 429 Too Many
+        # Requests, by hooking the client's POST method.
+        original_post_function = client._DocumentClient__Post
+        client._DocumentClient__Post = self._MockPost429
+
+        # Test: query for the document.  Expect the mock overloaded server
+        # to raise a 429 exception.
+        try:
+            query = client.QueryDocuments(self.created_collection['_self'], "SELECT * FROM c")
+            docs = list(query) # force execution now
+            self.assertFalse(True, 'function should raise HTTPFailure.')
+
+        except errors.HTTPFailure as ex:
+            self.assertEqual(ex.status_code, StatusCodes.TOO_MANY_REQUESTS)
+
+        client._DocumentClient__Post = original_post_function
+        client.DeleteDocument(created_document['_self'])
+
+    def _MockPost429(self, url, path, body, headers):
+        raise errors.HTTPFailure(StatusCodes.TOO_MANY_REQUESTS,
+                                 "Request rate is too large",
+                                 {HttpHeaders.RetryAfterInMilliseconds: 500})
+
     def _MockExecuteFunction(self, function, *args, **kwargs):
         raise errors.HTTPFailure(StatusCodes.TOO_MANY_REQUESTS, "Request rate is too large", {HttpHeaders.RetryAfterInMilliseconds: self.retry_after_in_milliseconds})