vellum-workflow-server 0.14.73.post1__tar.gz → 0.14.73.post2__tar.gz

{vellum_workflow_server-0.14.73.post1 → vellum_workflow_server-0.14.73.post2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vellum-workflow-server
-Version: 0.14.73.post1
+Version: 0.14.73.post2
 Summary: 
 License: AGPL
 Requires-Python: >=3.9.0,<4

{vellum_workflow_server-0.14.73.post1 → vellum_workflow_server-0.14.73.post2}/pyproject.toml

@@ -3,7 +3,7 @@ name = "vellum-workflow-server"
 
 [tool.poetry]
 name = "vellum-workflow-server"
-version = "0.14.73.post1"
+version = "0.14.73.post2"
 description = ""
 readme = "README.md"
 authors = []

{vellum_workflow_server-0.14.73.post1 → vellum_workflow_server-0.14.73.post2}/src/workflow_server/api/tests/test_workflow_view_stream_workflow_route.py

@@ -10,6 +10,7 @@ import requests_mock
 
 from workflow_server.code_exec_runner import run_code_exec_stream
 from workflow_server.server import create_app
+from workflow_server.utils.system_utils import get_active_process_count
 
 
 def flask_stream(request_body: dict) -> tuple[int, list]:
@@ -775,44 +776,15 @@ class Workflow(BaseWorkflow):
     # WHEN we call the stream route
     status_code, events = flask_stream(request_body)
 
-    # THEN we get a 200 response
-    assert status_code == 200, events
+    # THEN we get a 429 response
+    assert status_code == 429, events
 
-    # THEN we get the expected events
+    # AND we get a simple JSON error response
+    assert len(events) == 1
     assert events[0] == {
-        "id": mock.ANY,
-        "trace_id": mock.ANY,
-        "span_id": str(span_id),
-        "timestamp": mock.ANY,
-        "api_version": "2024-10-25",
-        "parent": None,
-        "name": "vembda.execution.initiated",
-        "body": {
-            "sdk_version": version("vellum-ai"),
-            "server_version": "local",
-        },
+        "detail": f"Workflow server concurrent request rate exceeded. Process count: {get_active_process_count()}"
     }
 
-    # AND we get a vembda.execution.fulfilled event with error
-    assert events[1] == {
-        "id": mock.ANY,
-        "trace_id": events[0]["trace_id"],
-        "span_id": str(span_id),
-        "timestamp": mock.ANY,
-        "api_version": "2024-10-25",
-        "parent": None,
-        "name": "vembda.execution.fulfilled",
-        "body": {
-            "log": "",
-            "exit_code": -1,
-            "stderr": "Workflow server concurrent request rate exceeded. Process count: 0",
-            "container_overhead_latency": mock.ANY,
-            "timed_out": False,
-        },
-    }
-
-    assert len(events) == 2
-
 
 def test_stream_workflow_route__with_environment_variables(both_stream_types):
     # GIVEN a valid request body with environment variables

{vellum_workflow_server-0.14.73.post1 → vellum_workflow_server-0.14.73.post2}/src/workflow_server/api/workflow_view.py

@@ -97,16 +97,14 @@ def stream_workflow_route() -> Response:
     # if we detect a memory problem just exit us early
     if not wait_for_available_process():
         return Response(
-            stream_with_context(
-                startup_error_generator(
-                    context=context,
-                    message=f"Workflow server concurrent request rate exceeded. "
-                    f"Process count: {get_active_process_count()}",
-                    vembda_initiated_event=vembda_initiated_event,
-                )
+            json.dumps(
+                {
+                    "detail": f"Workflow server concurrent request rate exceeded. "
+                    f"Process count: {get_active_process_count()}"
+                }
             ),
-            status=200,
-            content_type="application/x-ndjson",
+            status=429,
+            content_type="application/json",
             headers=headers,
         )