Switch entity values for all tests using push sources to not affect other tests

felixwang9817 · felixwang9817 · commit e54ea6eb1446 · 2022-06-30T13:00:51.000-07:00
Signed-off-by: Felix Wang &lt;wangfelix98@gmail.com&gt;
diff --git a/sdk/python/tests/integration/e2e/test_python_feature_server.py b/sdk/python/tests/integration/e2e/test_python_feature_server.py
@@ -63,13 +63,16 @@ def test_get_online_features(python_fs_client):
 @pytest.mark.integration
 @pytest.mark.universal_online_stores
 def test_push(python_fs_client):
-    initial_temp = get_temperatures(python_fs_client, location_ids=[1])[0]
+    # TODO(felixwang9817): Note that we choose an entity value of 102 here since it is not included
+    # in the existing range of entity values (1-49). This allows us to push data for this test
+    # without affecting other tests. This decision is tech debt, and should be resolved by finding a
+    # better way to isolate data sources across tests.
     json_data = json.dumps(
         {
             "push_source_name": "location_stats_push_source",
             "df": {
-                "location_id": [1],
-                "temperature": [initial_temp * 100],
+                "location_id": [102],
+                "temperature": [4],
                 "event_timestamp": [str(datetime.utcnow())],
                 "created": [str(datetime.utcnow())],
             },
@@ -79,7 +82,7 @@ def test_push(python_fs_client):
 
     # Check new pushed temperature is fetched
     assert response.status_code == 200
-    assert get_temperatures(python_fs_client, location_ids=[1]) == [initial_temp * 100]
+    assert get_temperatures(python_fs_client, location_ids=[102]) == [4]
 
 
 def get_temperatures(client, location_ids: List[int]):
diff --git a/sdk/python/tests/integration/offline_store/test_push_offline_retrieval.py b/sdk/python/tests/integration/offline_store/test_push_offline_retrieval.py
@@ -26,17 +26,21 @@ def test_push_features_and_read_from_offline_store(environment, universal_data_s
     now = pd.Timestamp(datetime.datetime.utcnow()).round("ms")
 
     store.apply([driver(), customer(), location(), *feature_views.values()])
-    entity_df = pd.DataFrame.from_dict({"location_id": [1], "event_timestamp": [now]})
+    entity_df = pd.DataFrame.from_dict({"location_id": [100], "event_timestamp": [now]})
 
     before_df = store.get_historical_features(
         entity_df=entity_df,
         features=["pushable_location_stats:temperature"],
         full_feature_names=False,
     ).to_df()
 
+    # TODO(felixwang9817): Note that we choose an entity value of 100 here since it is not included
+    # in the existing range of entity values (1-49). This allows us to push data for this test
+    # without affecting other tests. This decision is tech debt, and should be resolved by finding a
+    # better way to isolate data sources across tests.
     data = {
         "event_timestamp": [now],
-        "location_id": [1],
+        "location_id": [100],
         "temperature": [4],
         "created": [now],
     }
diff --git a/sdk/python/tests/integration/online_store/test_push_online_retrieval.py b/sdk/python/tests/integration/online_store/test_push_online_retrieval.py
@@ -22,8 +22,13 @@ def test_push_features_and_read(environment, universal_data_sources):
     feature_views = construct_universal_feature_views(data_sources)
 
     store.apply([driver(), customer(), location(), *feature_views.values()])
+
+    # TODO(felixwang9817): Note that we choose an entity value of 101 here since it is not included
+    # in the existing range of entity values (1-49). This allows us to push data for this test
+    # without affecting other tests. This decision is tech debt, and should be resolved by finding a
+    # better way to isolate data sources across tests.
     data = {
-        "location_id": [1],
+        "location_id": [101],
         "temperature": [4],
         "event_timestamp": [pd.Timestamp(datetime.datetime.utcnow()).round("ms")],
         "created": [pd.Timestamp(datetime.datetime.utcnow()).round("ms")],
@@ -34,8 +39,8 @@ def test_push_features_and_read(environment, universal_data_sources):
 
     online_resp = store.get_online_features(
         features=["pushable_location_stats:temperature"],
-        entity_rows=[{"location_id": 1}],
+        entity_rows=[{"location_id": 101}],
     )
     online_resp_dict = online_resp.to_dict()
-    assert online_resp_dict["location_id"] == [1]
+    assert online_resp_dict["location_id"] == [101]
     assert online_resp_dict["temperature"] == [4]
