feat(config): Add WORKFLOW_PARALLEL_DEPTH_LIMIT and set default to 3

Signed-off-by: -LAN- <[email protected]>

api/.env.example

@@ -389,6 +389,7 @@ INDEXING_MAX_SEGMENTATION_TOKENS_LENGTH=4000
 WORKFLOW_MAX_EXECUTION_STEPS=500
 WORKFLOW_MAX_EXECUTION_TIME=1200
 WORKFLOW_CALL_MAX_DEPTH=5
+WORKFLOW_PARALLEL_DEPTH_LIMIT=3
 MAX_VARIABLE_SIZE=204800
 
 # App configuration

api/configs/feature/__init__.py

@@ -433,6 +433,11 @@ class WorkflowConfig(BaseSettings):
         default=5,
     )
 
+    WORKFLOW_PARALLEL_DEPTH_LIMIT: PositiveInt = Field(
+        description="Maximum allowed depth for nested parallel executions",
+        default=3,
+    )
+
     MAX_VARIABLE_SIZE: PositiveInt = Field(
         description="Maximum size in bytes for a single variable in workflows. Default to 200 KB.",
         default=200 * 1024,

api/core/workflow/graph_engine/entities/graph.py

@@ -4,6 +4,7 @@
 
 from pydantic import BaseModel, Field
 
+from configs import dify_config
 from core.workflow.graph_engine.entities.run_condition import RunCondition
 from core.workflow.nodes import NodeType
 from core.workflow.nodes.answer.answer_stream_generate_router import AnswerStreamGeneratorRouter
@@ -161,7 +162,9 @@ def init(cls, graph_config: Mapping[str, Any], root_node_id: Optional[str] = Non
         for parallel in parallel_mapping.values():
             if parallel.parent_parallel_id:
                 cls._check_exceed_parallel_limit(
-                    parallel_mapping=parallel_mapping, level_limit=3, parent_parallel_id=parallel.parent_parallel_id
+                    parallel_mapping=parallel_mapping,
+                    level_limit=dify_config.WORKFLOW_PARALLEL_DEPTH_LIMIT,
+                    parent_parallel_id=parallel.parent_parallel_id,
                 )
 
         # init answer stream generate routes

docker/.env.example

@@ -712,6 +712,7 @@ WORKFLOW_MAX_EXECUTION_STEPS=500
 WORKFLOW_MAX_EXECUTION_TIME=1200
 WORKFLOW_CALL_MAX_DEPTH=5
 MAX_VARIABLE_SIZE=204800
+WORKFLOW_PARALLEL_DEPTH_LIMIT=3
 WORKFLOW_FILE_UPLOAD_LIMIT=10
 
 # HTTP request node in workflow configuration