Conversation
Signed-off-by: Tomu Hirata <tomu.hirata@gmail.com>
🛠 DevTools 🛠
Install mlflow from this PRFor Databricks, use the following command: |
|
Documentation preview for ed56978 is available at: More info
|
There was a problem hiding this comment.
Pull request overview
This pull request addresses a critical CORS (Cross-Origin Resource Sharing) security vulnerability where origin validation was not being applied to /ajax-api/ endpoints. The vulnerability allowed unauthorized cross-origin requests to these endpoints, which are extensively used by the MLflow frontend for API calls.
Changes:
- Extended CORS origin checking to include
/ajax-api/path prefix in addition to/api/prefix - Fixed FastAPI CORSMiddleware configuration to properly restrict origins instead of allowing all origins by default
- Added comprehensive test coverage for the security fix
Reviewed changes
Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.
| File | Description |
|---|---|
mlflow/server/security_utils.py |
Added AJAX_API_PATH_PREFIX constant and updated is_api_endpoint() to check for both /api/ and /ajax-api/ prefixes |
mlflow/server/fastapi_security.py |
Fixed critical issue where allow_origins was set to ["*"] by default; now properly uses configured origins plus localhost patterns |
tests/server/test_security.py |
Added test for is_api_endpoint() with ajax paths, added FastAPI CORS tests for localhost origins and configured origins, updated OPTIONS request test to expect 204 status |
tests/server/conftest.py |
Added fastapi_client fixture and updated test endpoint paths to use realistic API paths |
| "/api/2.0/mlflow/experiments/list", | ||
| headers={"Host": "localhost", "Origin": "http://evil.com"}, | ||
| ) | ||
| assert response.headers.get("access-control-allow-origin") is None |
There was a problem hiding this comment.
The FastAPI CORS tests only verify that GET requests from unauthorized origins don't receive CORS headers, but they don't test that POST requests from unauthorized origins are actually blocked with a 403 status. This is a critical security behavior that should be explicitly tested. Consider adding a test case that sends a POST request from an unauthorized origin to a FastAPI endpoint and verifies it returns 403 status with appropriate error message.
| assert response.headers.get("access-control-allow-origin") is None | |
| assert response.headers.get("access-control-allow-origin") is None | |
| # Verify that POST requests from an unauthorized origin are blocked | |
| response = client.post( | |
| "/api/2.0/mlflow/experiments/list", | |
| headers={"Host": "localhost", "Origin": "http://evil.com"}, | |
| ) | |
| assert response.status_code == 403 | |
| # Ensure an error message is returned | |
| error_detail = response.json().get("detail") | |
| assert isinstance(error_detail, str) and error_detail |
| def test_is_api_endpoint(path, expected): | ||
| assert is_api_endpoint(path) == expected | ||
|
|
||
|
|
There was a problem hiding this comment.
Consider adding an explicit test case for CORS blocking on /ajax-api/ paths. While test_is_api_endpoint() verifies that /ajax-api/ paths are recognized as API endpoints, there's no direct test similar to test_cors_protection that verifies POST requests from unauthorized origins to /ajax-api/ endpoints are actually blocked with a 403 status. This would provide stronger verification that the security fix works end-to-end.
| def test_fastapi_cors_blocks_unauthorized_origin_on_ajax_api(fastapi_client): | |
| response = fastapi_client.post( | |
| "/ajax-api/2.0/mlflow/experiments/list", | |
| headers={"Host": "localhost", "Origin": "http://evil.com"}, | |
| ) | |
| assert response.status_code == 403 |
Signed-off-by: Tomu Hirata <tomu.hirata@gmail.com>
Signed-off-by: Tomu Hirata <tomu.hirata@gmail.com>
Related Issues/PRs
#20812
What changes are proposed in this pull request?
Fixed a CORS issue where the origin check is not enabled for ajax paths
How is this PR tested?
Does this PR require documentation update?
Does this PR require updating the MLflow Skills repository?
Release Notes
Is this a user-facing change?
What component(s), interfaces, languages, and integrations does this PR affect?
Components
area/tracking: Tracking Service, tracking client APIs, autologgingarea/models: MLmodel format, model serialization/deserialization, flavorsarea/model-registry: Model Registry service, APIs, and the fluent client calls for Model Registryarea/scoring: MLflow Model server, model deployment tools, Spark UDFsarea/evaluation: MLflow model evaluation features, evaluation metrics, and evaluation workflowsarea/gateway: MLflow AI Gateway client APIs, server, and third-party integrationsarea/prompts: MLflow prompt engineering features, prompt templates, and prompt managementarea/tracing: MLflow Tracing features, tracing APIs, and LLM tracing functionalityarea/projects: MLproject format, project running backendsarea/uiux: Front-end, user experience, plotting, JavaScript, JavaScript dev serverarea/build: Build and test infrastructure for MLflowarea/docs: MLflow documentation pagesHow should the PR be classified in the release notes? Choose one:
rn/none- No description will be included. The PR will be mentioned only by the PR number in the "Small Bugfixes and Documentation Updates" sectionrn/breaking-change- The PR will be mentioned in the "Breaking Changes" sectionrn/feature- A new user-facing feature worth mentioning in the release notesrn/bug-fix- A user-facing bug fix worth mentioning in the release notesrn/documentation- A user-facing documentation change worth mentioning in the release notesShould this PR be included in the next patch release?
Yesshould be selected for bug fixes, documentation updates, and other small changes.Noshould be selected for new features and larger changes. If you're unsure about the release classification of this PR, leave this unchecked to let the maintainers decide.What is a minor/patch release?
Bug fixes, doc updates and new features usually go into minor releases.
Bug fixes and doc updates usually go into patch releases.