test overlapping connectors (but using a source that is way too big a… (#3152)

* test overlapping connectors (but using a source that is way too big and slow, fix that next)

* pass thru secrets

* rename

* rename again

* now we are fixing it

---------

Co-authored-by: Richard Kuo <rkuo@rkuo.com>

.github/workflows/pr-integration-tests.yml

@@ -13,7 +13,10 @@ on:
 env:
   OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
   SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
-
+  CONFLUENCE_TEST_SPACE_URL: ${{ secrets.CONFLUENCE_TEST_SPACE_URL }}
+  CONFLUENCE_USER_NAME: ${{ secrets.CONFLUENCE_USER_NAME }}
+  CONFLUENCE_ACCESS_TOKEN: ${{ secrets.CONFLUENCE_ACCESS_TOKEN }}
+  
 jobs:
   integration-tests:
     # See https://runs-on.com/runners/linux/
@@ -195,6 +198,9 @@ jobs:
             -e API_SERVER_HOST=api_server \
             -e OPENAI_API_KEY=${OPENAI_API_KEY} \
             -e SLACK_BOT_TOKEN=${SLACK_BOT_TOKEN} \
+            -e CONFLUENCE_TEST_SPACE_URL=${CONFLUENCE_TEST_SPACE_URL} \
+            -e CONFLUENCE_USER_NAME=${CONFLUENCE_USER_NAME} \
+            -e CONFLUENCE_ACCESS_TOKEN=${CONFLUENCE_ACCESS_TOKEN} \
             -e TEST_WEB_HOSTNAME=test-runner \
             danswer/danswer-integration:test \
             /app/tests/integration/tests \

backend/tests/integration/tests/connector/test_connector_creation.py

@@ -0,0 +1,84 @@
+import os
+from datetime import datetime
+from datetime import timezone
+
+from danswer.server.documents.models import DocumentSource
+from tests.integration.common_utils.managers.cc_pair import CCPairManager
+from tests.integration.common_utils.managers.user import UserManager
+from tests.integration.common_utils.test_models import DATestUser
+
+
+def test_connector_creation(reset: None) -> None:
+    # Creating an admin user (first user created is automatically an admin)
+    admin_user: DATestUser = UserManager.create(name="admin_user")
+
+    # create connectors
+    cc_pair_1 = CCPairManager.create_from_scratch(
+        source=DocumentSource.INGESTION_API,
+        user_performing_action=admin_user,
+    )
+
+    cc_pair_info = CCPairManager.get_single(
+        cc_pair_1.id, user_performing_action=admin_user
+    )
+    assert cc_pair_info
+    assert cc_pair_info.creator
+    assert str(cc_pair_info.creator) == admin_user.id
+    assert cc_pair_info.creator_email == admin_user.email
+
+
+def test_overlapping_connector_creation(reset: None) -> None:
+    """Tests that connectors indexing the same documents don't interfere with each other.
+    A previous bug involved document by cc pair entries not being added for new connectors
+    when the docs existed already via another connector and were up to date relative to the source.
+    """
+    admin_user: DATestUser = UserManager.create(name="admin_user")
+
+    config = {
+        "wiki_base": os.environ["CONFLUENCE_TEST_SPACE_URL"],
+        "space": "DailyConne",
+        "is_cloud": True,
+        "page_id": "",
+    }
+
+    credential = {
+        "confluence_username": os.environ["CONFLUENCE_USER_NAME"],
+        "confluence_access_token": os.environ["CONFLUENCE_ACCESS_TOKEN"],
+    }
+
+    # store the time before we create the connector so that we know after
+    # when the indexing should have started
+    now = datetime.now(timezone.utc)
+
+    # create connector
+    cc_pair_1 = CCPairManager.create_from_scratch(
+        source=DocumentSource.CONFLUENCE,
+        connector_specific_config=config,
+        credential_json=credential,
+        user_performing_action=admin_user,
+    )
+
+    CCPairManager.wait_for_indexing(
+        cc_pair_1, now, timeout=120, user_performing_action=admin_user
+    )
+
+    now = datetime.now(timezone.utc)
+
+    cc_pair_2 = CCPairManager.create_from_scratch(
+        source=DocumentSource.CONFLUENCE,
+        connector_specific_config=config,
+        credential_json=credential,
+        user_performing_action=admin_user,
+    )
+
+    CCPairManager.wait_for_indexing(
+        cc_pair_2, now, timeout=120, user_performing_action=admin_user
+    )
+
+    info_1 = CCPairManager.get_single(cc_pair_1.id, user_performing_action=admin_user)
+    assert info_1
+
+    info_2 = CCPairManager.get_single(cc_pair_2.id, user_performing_action=admin_user)
+    assert info_2
+
+    assert info_1.num_docs_indexed == info_2.num_docs_indexed

backend/tests/integration/tests/connector/test_connector_deletion.py

@@ -29,78 +29,6 @@ from tests.integration.common_utils.test_models import DATestUserGroup
 from tests.integration.common_utils.vespa import vespa_fixture
 
 
-def test_connector_creation(reset: None) -> None:
-    # Creating an admin user (first user created is automatically an admin)
-    admin_user: DATestUser = UserManager.create(name="admin_user")
-
-    # create connectors
-    cc_pair_1 = CCPairManager.create_from_scratch(
-        source=DocumentSource.INGESTION_API,
-        user_performing_action=admin_user,
-    )
-
-    cc_pair_info = CCPairManager.get_single(
-        cc_pair_1.id, user_performing_action=admin_user
-    )
-    assert cc_pair_info
-    assert cc_pair_info.creator
-    assert str(cc_pair_info.creator) == admin_user.id
-    assert cc_pair_info.creator_email == admin_user.email
-
-
-# TODO(rkuo): will enable this once i have credentials on github
-# def test_overlapping_connector_creation(reset: None) -> None:
-#     # Creating an admin user (first user created is automatically an admin)
-#     admin_user: DATestUser = UserManager.create(name="admin_user")
-
-#     config = {
-#         "wiki_base": os.environ["CONFLUENCE_TEST_SPACE_URL"],
-#         "space": os.environ["CONFLUENCE_TEST_SPACE"],
-#         "is_cloud": True,
-#         "page_id": "",
-#     }
-
-#     credential = {
-#         "confluence_username": os.environ["CONFLUENCE_USER_NAME"],
-#         "confluence_access_token": os.environ["CONFLUENCE_ACCESS_TOKEN"],
-#     }
-
-#     # store the time before we create the connector so that we know after
-#     # when the indexing should have started
-#     now = datetime.now(timezone.utc)
-
-#     # create connector
-#     cc_pair_1 = CCPairManager.create_from_scratch(
-#         source=DocumentSource.CONFLUENCE,
-#         connector_specific_config=config,
-#         credential_json=credential,
-#         user_performing_action=admin_user,
-#     )
-
-#     CCPairManager.wait_for_indexing(
-#         cc_pair_1, now, timeout=60, user_performing_action=admin_user
-#     )
-
-#     cc_pair_2 = CCPairManager.create_from_scratch(
-#         source=DocumentSource.CONFLUENCE,
-#         connector_specific_config=config,
-#         credential_json=credential,
-#         user_performing_action=admin_user,
-#     )
-
-#     CCPairManager.wait_for_indexing(
-#         cc_pair_2, now, timeout=60, user_performing_action=admin_user
-#     )
-
-#     info_1 = CCPairManager.get_single(cc_pair_1.id)
-#     assert info_1
-
-#     info_2 = CCPairManager.get_single(cc_pair_2.id)
-#     assert info_2
-
-#     assert info_1.num_docs_indexed == info_2.num_docs_indexed
-
-
 def test_connector_deletion(reset: None, vespa_client: vespa_fixture) -> None:
     # Creating an admin user (first user created is automatically an admin)
     admin_user: DATestUser = UserManager.create(name="admin_user")