Merge branch 'main' into feat/add-connector-gh-pages

Author: akhilender-bongirwar

Diff for: .github/workflows/pr-python-connector-tests.yml

@@ -45,11 +45,16 @@ env:
   SHAREPOINT_CLIENT_SECRET: ${{ secrets.SHAREPOINT_CLIENT_SECRET }}
   SHAREPOINT_CLIENT_DIRECTORY_ID: ${{ secrets.SHAREPOINT_CLIENT_DIRECTORY_ID }}
   SHAREPOINT_SITE: ${{ secrets.SHAREPOINT_SITE }}
+  # Github
+  ACCESS_TOKEN_GITHUB: ${{ secrets.ACCESS_TOKEN_GITHUB }}
   # Gitbook
   GITBOOK_SPACE_ID: ${{ secrets.GITBOOK_SPACE_ID }}
   GITBOOK_API_KEY: ${{ secrets.GITBOOK_API_KEY }}
   # Notion
   NOTION_INTEGRATION_TOKEN: ${{ secrets.NOTION_INTEGRATION_TOKEN }}
+  # Highspot
+  HIGHSPOT_KEY: ${{ secrets.HIGHSPOT_KEY }}
+  HIGHSPOT_SECRET: ${{ secrets.HIGHSPOT_SECRET }}
 
 jobs:
   connectors-check:

Diff for: backend/alembic.ini

@@ -84,7 +84,7 @@ keys = console
 keys = generic
 
 [logger_root]
-level = WARN
+level = INFO
 handlers = console
 qualname =
 

Diff for: backend/alembic/env.py

@@ -25,6 +25,9 @@
 from onyx.db.models import Base
 from celery.backends.database.session import ResultModelBase  # type: ignore
 
+# Make sure in alembic.ini [logger_root] level=INFO is set or most logging will be
+# hidden! (defaults to level=WARN)
+
 # Alembic Config object
 config = context.config
 
@@ -36,6 +39,7 @@
 target_metadata = [Base.metadata, ResultModelBase.metadata]
 
 EXCLUDE_TABLES = {"kombu_queue", "kombu_message"}
+
 logger = logging.getLogger(__name__)
 
 ssl_context: ssl.SSLContext | None = None
@@ -64,7 +68,7 @@ def include_object(
     return True
 
 
-def get_schema_options() -> tuple[str, bool, bool]:
+def get_schema_options() -> tuple[str, bool, bool, bool]:
     x_args_raw = context.get_x_argument()
     x_args = {}
     for arg in x_args_raw:
@@ -76,6 +80,10 @@ def get_schema_options() -> tuple[str, bool, bool]:
     create_schema = x_args.get("create_schema", "true").lower() == "true"
     upgrade_all_tenants = x_args.get("upgrade_all_tenants", "false").lower() == "true"
 
+    # continue on error with individual tenant
+    # only applies to online migrations
+    continue_on_error = x_args.get("continue", "false").lower() == "true"
+
     if (
         MULTI_TENANT
         and schema_name == POSTGRES_DEFAULT_SCHEMA
@@ -86,14 +94,12 @@ def get_schema_options() -> tuple[str, bool, bool]:
             "Please specify a tenant-specific schema."
         )
 
-    return schema_name, create_schema, upgrade_all_tenants
+    return schema_name, create_schema, upgrade_all_tenants, continue_on_error
 
 
 def do_run_migrations(
     connection: Connection, schema_name: str, create_schema: bool
 ) -> None:
-    logger.info(f"About to migrate schema: {schema_name}")
-
     if create_schema:
         connection.execute(text(f'CREATE SCHEMA IF NOT EXISTS "{schema_name}"'))
         connection.execute(text("COMMIT"))
@@ -134,7 +140,12 @@ def provide_iam_token_for_alembic(
 
 
 async def run_async_migrations() -> None:
-    schema_name, create_schema, upgrade_all_tenants = get_schema_options()
+    (
+        schema_name,
+        create_schema,
+        upgrade_all_tenants,
+        continue_on_error,
+    ) = get_schema_options()
 
     engine = create_async_engine(
         build_connection_string(),
@@ -151,9 +162,15 @@ def event_provide_iam_token_for_alembic(
 
     if upgrade_all_tenants:
         tenant_schemas = get_all_tenant_ids()
+
+        i_tenant = 0
+        num_tenants = len(tenant_schemas)
         for schema in tenant_schemas:
+            i_tenant += 1
+            logger.info(
+                f"Migrating schema: index={i_tenant} num_tenants={num_tenants} schema={schema}"
+            )
             try:
-                logger.info(f"Migrating schema: {schema}")
                 async with engine.connect() as connection:
                     await connection.run_sync(
                         do_run_migrations,
@@ -162,7 +179,12 @@ def event_provide_iam_token_for_alembic(
                     )
             except Exception as e:
                 logger.error(f"Error migrating schema {schema}: {e}")
-                raise
+                if not continue_on_error:
+                    logger.error("--continue is not set, raising exception!")
+                    raise
+
+                logger.warning("--continue is set, continuing to next schema.")
+
     else:
         try:
             logger.info(f"Migrating schema: {schema_name}")
@@ -180,7 +202,11 @@ def event_provide_iam_token_for_alembic(
 
 
 def run_migrations_offline() -> None:
-    schema_name, _, upgrade_all_tenants = get_schema_options()
+    """This doesn't really get used when we migrate in the cloud."""
+
+    logger.info("run_migrations_offline starting.")
+
+    schema_name, _, upgrade_all_tenants, continue_on_error = get_schema_options()
     url = build_connection_string()
 
     if upgrade_all_tenants:
@@ -230,6 +256,7 @@ def event_provide_iam_token_for_alembic_offline(
 
 
 def run_migrations_online() -> None:
+    logger.info("run_migrations_online starting.")
     asyncio.run(run_async_migrations())
 
 

Diff for: backend/ee/onyx/server/enterprise_settings/api.py

@@ -15,8 +15,8 @@
 
 from ee.onyx.server.enterprise_settings.models import AnalyticsScriptUpload
 from ee.onyx.server.enterprise_settings.models import EnterpriseSettings
-from ee.onyx.server.enterprise_settings.store import _LOGO_FILENAME
-from ee.onyx.server.enterprise_settings.store import _LOGOTYPE_FILENAME
+from ee.onyx.server.enterprise_settings.store import get_logo_filename
+from ee.onyx.server.enterprise_settings.store import get_logotype_filename
 from ee.onyx.server.enterprise_settings.store import load_analytics_script
 from ee.onyx.server.enterprise_settings.store import load_settings
 from ee.onyx.server.enterprise_settings.store import store_analytics_script
@@ -28,7 +28,7 @@
 from onyx.auth.users import UserManager
 from onyx.db.engine import get_session
 from onyx.db.models import User
-from onyx.file_store.file_store import get_default_file_store
+from onyx.file_store.file_store import PostgresBackedFileStore
 from onyx.utils.logger import setup_logger
 
 admin_router = APIRouter(prefix="/admin/enterprise-settings")
@@ -131,31 +131,49 @@ def put_logo(
     upload_logo(file=file, db_session=db_session, is_logotype=is_logotype)
 
 
-def fetch_logo_or_logotype(is_logotype: bool, db_session: Session) -> Response:
+def fetch_logo_helper(db_session: Session) -> Response:
     try:
-        file_store = get_default_file_store(db_session)
-        filename = _LOGOTYPE_FILENAME if is_logotype else _LOGO_FILENAME
-        file_io = file_store.read_file(filename, mode="b")
-        # NOTE: specifying "image/jpeg" here, but it still works for pngs
-        # TODO: do this properly
-        return Response(content=file_io.read(), media_type="image/jpeg")
+        file_store = PostgresBackedFileStore(db_session)
+        onyx_file = file_store.get_file_with_mime_type(get_logo_filename())
+        if not onyx_file:
+            raise ValueError("get_onyx_file returned None!")
     except Exception:
         raise HTTPException(
             status_code=404,
-            detail=f"No {'logotype' if is_logotype else 'logo'} file found",
+            detail="No logo file found",
         )
+    else:
+        return Response(content=onyx_file.data, media_type=onyx_file.mime_type)
+
+
+def fetch_logotype_helper(db_session: Session) -> Response:
+    try:
+        file_store = PostgresBackedFileStore(db_session)
+        onyx_file = file_store.get_file_with_mime_type(get_logotype_filename())
+        if not onyx_file:
+            raise ValueError("get_onyx_file returned None!")
+    except Exception:
+        raise HTTPException(
+            status_code=404,
+            detail="No logotype file found",
+        )
+    else:
+        return Response(content=onyx_file.data, media_type=onyx_file.mime_type)
 
 
 @basic_router.get("/logotype")
 def fetch_logotype(db_session: Session = Depends(get_session)) -> Response:
-    return fetch_logo_or_logotype(is_logotype=True, db_session=db_session)
+    return fetch_logotype_helper(db_session)
 
 
 @basic_router.get("/logo")
 def fetch_logo(
     is_logotype: bool = False, db_session: Session = Depends(get_session)
 ) -> Response:
-    return fetch_logo_or_logotype(is_logotype=is_logotype, db_session=db_session)
+    if is_logotype:
+        return fetch_logotype_helper(db_session)
+
+    return fetch_logo_helper(db_session)
 
 
 @admin_router.put("/custom-analytics-script")

Diff for: backend/ee/onyx/server/enterprise_settings/store.py

@@ -13,6 +13,7 @@
 from onyx.configs.constants import FileOrigin
 from onyx.configs.constants import KV_CUSTOM_ANALYTICS_SCRIPT_KEY
 from onyx.configs.constants import KV_ENTERPRISE_SETTINGS_KEY
+from onyx.configs.constants import ONYX_DEFAULT_APPLICATION_NAME
 from onyx.file_store.file_store import get_default_file_store
 from onyx.key_value_store.factory import get_kv_store
 from onyx.key_value_store.interface import KvKeyNotFoundError
@@ -21,8 +22,18 @@
 
 logger = setup_logger()
 
+_LOGO_FILENAME = "__logo__"
+_LOGOTYPE_FILENAME = "__logotype__"
+
 
 def load_settings() -> EnterpriseSettings:
+    """Loads settings data directly from DB. This should be used primarily
+    for checking what is actually in the DB, aka for editing and saving back settings.
+
+    Runtime settings actually used by the application should be checked with
+    load_runtime_settings as defaults may be applied at runtime.
+    """
+
     dynamic_config_store = get_kv_store()
     try:
         settings = EnterpriseSettings(
@@ -36,9 +47,24 @@ def load_settings() -> EnterpriseSettings:
 
 
 def store_settings(settings: EnterpriseSettings) -> None:
+    """Stores settings directly to the kv store / db."""
+
     get_kv_store().store(KV_ENTERPRISE_SETTINGS_KEY, settings.model_dump())
 
 
+def load_runtime_settings() -> EnterpriseSettings:
+    """Loads settings from DB and applies any defaults or transformations for use
+    at runtime.
+
+    Should not be stored back to the DB.
+    """
+    enterprise_settings = load_settings()
+    if not enterprise_settings.application_name:
+        enterprise_settings.application_name = ONYX_DEFAULT_APPLICATION_NAME
+
+    return enterprise_settings
+
+
 _CUSTOM_ANALYTICS_SECRET_KEY = os.environ.get("CUSTOM_ANALYTICS_SECRET_KEY")
 
 
@@ -60,10 +86,6 @@ def store_analytics_script(analytics_script_upload: AnalyticsScriptUpload) -> No
     get_kv_store().store(KV_CUSTOM_ANALYTICS_SCRIPT_KEY, analytics_script_upload.script)
 
 
-_LOGO_FILENAME = "__logo__"
-_LOGOTYPE_FILENAME = "__logotype__"
-
-
 def is_valid_file_type(filename: str) -> bool:
     valid_extensions = (".png", ".jpg", ".jpeg")
     return filename.endswith(valid_extensions)
@@ -116,3 +138,11 @@ def upload_logo(
         file_type=file_type,
     )
     return True
+
+
+def get_logo_filename() -> str:
+    return _LOGO_FILENAME
+
+
+def get_logotype_filename() -> str:
+    return _LOGOTYPE_FILENAME

Diff for: backend/ee/onyx/server/tenants/provisioning.py

@@ -87,11 +87,14 @@ async def get_or_provision_tenant(
             # If we have a pre-provisioned tenant, assign it to the user
             await assign_tenant_to_user(tenant_id, email, referral_source)
             logger.info(f"Assigned pre-provisioned tenant {tenant_id} to user {email}")
-            return tenant_id
         else:
             # If no pre-provisioned tenant is available, create a new one on-demand
             tenant_id = await create_tenant(email, referral_source)
-            return tenant_id
+
+        # Notify control plane if we have created / assigned a new tenant
+        if not DEV_MODE:
+            await notify_control_plane(tenant_id, email, referral_source)
+        return tenant_id
 
     except Exception as e:
         # If we've encountered an error, log and raise an exception
@@ -116,10 +119,6 @@ async def create_tenant(email: str, referral_source: str | None = None) -> str:
         # Provision tenant on data plane
         await provision_tenant(tenant_id, email)
 
-        # Notify control plane if not already done in provision_tenant
-        if not DEV_MODE and referral_source:
-            await notify_control_plane(tenant_id, email, referral_source)
-
     except Exception as e:
         logger.exception(f"Tenant provisioning failed: {str(e)}")
         # Attempt to rollback the tenant provisioning
@@ -271,6 +270,7 @@ def configure_default_api_keys(db_session: Session) -> None:
             fast_default_model_name="claude-3-5-sonnet-20241022",
             model_names=ANTHROPIC_MODEL_NAMES,
             display_model_names=["claude-3-5-sonnet-20241022"],
+            api_key_changed=True,
         )
         try:
             full_provider = upsert_llm_provider(anthropic_provider, db_session)
@@ -283,17 +283,18 @@ def configure_default_api_keys(db_session: Session) -> None:
         )
 
     if OPENAI_DEFAULT_API_KEY:
-        open_provider = LLMProviderUpsertRequest(
+        openai_provider = LLMProviderUpsertRequest(
             name="OpenAI",
             provider=OPENAI_PROVIDER_NAME,
             api_key=OPENAI_DEFAULT_API_KEY,
             default_model_name="gpt-4o",
             fast_default_model_name="gpt-4o-mini",
             model_names=OPEN_AI_MODEL_NAMES,
             display_model_names=["o1", "o3-mini", "gpt-4o", "gpt-4o-mini"],
+            api_key_changed=True,
         )
         try:
-            full_provider = upsert_llm_provider(open_provider, db_session)
+            full_provider = upsert_llm_provider(openai_provider, db_session)
             update_default_provider(full_provider.id, db_session)
         except Exception as e:
             logger.error(f"Failed to configure OpenAI provider: {e}")
@@ -559,7 +560,3 @@ async def assign_tenant_to_user(
     except Exception:
         logger.exception(f"Failed to assign tenant {tenant_id} to user {email}")
         raise Exception("Failed to assign tenant to user")
-
-    # Notify control plane with retry logic
-    if not DEV_MODE:
-        await notify_control_plane(tenant_id, email, referral_source)

Diff for: backend/model_server/main.py

@@ -65,11 +65,17 @@ async def lifespan(app: FastAPI) -> AsyncGenerator:
 
     app.state.gpu_type = gpu_type
 
-    if TEMP_HF_CACHE_PATH.is_dir():
-        logger.notice("Moving contents of temp_huggingface to huggingface cache.")
-        _move_files_recursively(TEMP_HF_CACHE_PATH, HF_CACHE_PATH)
-        shutil.rmtree(TEMP_HF_CACHE_PATH, ignore_errors=True)
-        logger.notice("Moved contents of temp_huggingface to huggingface cache.")
+    try:
+        if TEMP_HF_CACHE_PATH.is_dir():
+            logger.notice("Moving contents of temp_huggingface to huggingface cache.")
+            _move_files_recursively(TEMP_HF_CACHE_PATH, HF_CACHE_PATH)
+            shutil.rmtree(TEMP_HF_CACHE_PATH, ignore_errors=True)
+            logger.notice("Moved contents of temp_huggingface to huggingface cache.")
+    except Exception as e:
+        logger.warning(
+            f"Error moving contents of temp_huggingface to huggingface cache: {e}. "
+            "This is not a critical error and the model server will continue to run."
+        )
 
     torch.set_num_threads(max(MIN_THREADS_ML_MODELS, torch.get_num_threads()))
     logger.notice(f"Torch Threads: {torch.get_num_threads()}")