consolidate all debug flags to show kubernates exception

Author: mollyheamazon

hyperpod-pytorch-job-template/hyperpod_pytorch_job_template/v1_1/model.py

@@ -481,16 +481,6 @@ def build_dict(**kwargs):
 
         result = HyperPodPytorchJob(**job_kwargs)
         return result
-    
-    def create_from_k8s_yaml(self, yaml_file_path: str) -> None:
-        """Create HyperPodPytorchJob from k8s YAML file."""
-        with open(yaml_file_path, 'r') as f:
-            yaml_data = yaml.safe_load(f)
-        
-        # Combine metadata and spec for full validation
-        full_data = {**yaml_data['spec'], 'metadata': yaml_data['metadata']}
-        job = HyperPodPytorchJob.model_validate(full_data, by_name=True)
-        job.create()
 
 
 # Volume-specific type handlers - only override what's needed

src/sagemaker/hyperpod/cli/commands/inference.py

@@ -21,7 +21,7 @@
 # CREATE
 @click.command("hyp-jumpstart-endpoint")
 @click.option("--version", default="1.1", help="Schema version to use")
-@click.option("--debug", default=False, help="Enable debug mode")
+@click.option("--debug", is_flag=True, help="Enable debug mode")
 @generate_click_command(
     schema_pkg="hyperpod_jumpstart_inference_template",
     registry=JS_REG,
@@ -38,7 +38,7 @@ def js_create(version, debug, js_endpoint):
 
 @click.command("hyp-custom-endpoint")
 @click.option("--version", default="1.1", help="Schema version to use")
-@click.option("--debug", default=False, help="Enable debug mode")
+@click.option("--debug", is_flag=True, help="Enable debug mode")
 @generate_click_command(
     schema_pkg="hyperpod_custom_inference_template",
     registry=C_REG,

src/sagemaker/hyperpod/cli/commands/init.py

@@ -274,8 +274,9 @@ def validate():
 @click.command(name="_default_create")
 @click.option("--region", "-r", default=None, help="Region to create cluster stack for, default to your region in aws configure. Not available for other templates.")
 @click.option("--template-version", type=click.INT, help="Version number of cluster creation template. Not available for other templates.")
+@click.option("--debug", is_flag=True, help="Enable debug logging")
 @_hyperpod_telemetry_emitter(Feature.HYPERPOD_CLI, "init_create_cli")
-def _default_create(region, template_version):
+def _default_create(region, template_version, debug):
     """
     Validate configuration and render template files for deployment.
     
@@ -384,7 +385,7 @@ def _default_create(region, template_version):
             else:
                 # Create from k8s.yaml
                 k8s_file = out_dir / 'k8s.yaml'
-                create_from_k8s_yaml(str(k8s_file))
+                create_from_k8s_yaml(str(k8s_file), debug=debug)
 
 
     except Exception as e:

src/sagemaker/hyperpod/cli/commands/space.py

@@ -13,15 +13,16 @@
 
 
 @click.command("hyp-space")
+@click.option("--debug", is_flag=True, help="Enable debug mode")
 @generate_click_command(
     schema_pkg="hyperpod_space_template",
     registry=SCHEMA_REGISTRY,
 )
-def space_create(version, config):
+def space_create(version, debug, config):
     """Create a space resource."""
     space_config = SpaceConfig(**config)
     space = HPSpace(config=space_config)
-    space.create()
+    space.create(debug=debug)
     click.echo(f"Space '{space_config.name}' created successfully in namespace '{space_config.namespace}'")
 
 

src/sagemaker/hyperpod/cli/commands/training.py

@@ -13,7 +13,7 @@
 
 @click.command("hyp-pytorch-job")
 @click.option("--version", default="1.0", help="Schema version to use")
-@click.option("--debug", default=False, help="Enable debug mode")
+@click.option("--debug", is_flag=True, help="Enable debug mode")
 @generate_click_command(
     schema_pkg="hyperpod_pytorch_job_template",
     registry=SCHEMA_REGISTRY,

src/sagemaker/hyperpod/cli/init_utils.py

@@ -536,7 +536,7 @@ def build_config_from_schema(template: str, version: str, model_config=None, exi
     return full_cfg, comment_map
 
 
-def create_from_k8s_yaml(yaml_file_path: str) -> None:
+def create_from_k8s_yaml(yaml_file_path: str, debug: bool = False) -> None:
     """Create HyperPod resource from K8s YAML file based on kind mapping."""
     from sagemaker.hyperpod.cli.constants.init_constants import K8S_KIND_MAPPING
 
@@ -563,4 +563,4 @@ def create_from_k8s_yaml(yaml_file_path: str) -> None:
         resource = resource_class.model_validate(yaml_data['spec'], by_name=True)
         resource.metadata = Metadata.model_validate(yaml_data['metadata'], by_name=True)
 
-    resource.create()
+    resource.create(debug=debug)

src/sagemaker/hyperpod/common/utils.py

@@ -39,7 +39,7 @@ def get_default_namespace():
         )
 
 def handle_exception(e: Exception, name: str, namespace: Optional[str],
-                    operation_type: str = 'unknown', resource_type: str = 'unknown'):
+                    operation_type: str = 'unknown', resource_type: str = 'unknown', debug: bool = False):
     """
     Handle various Kubernetes API exceptions for SDK usage (non-CLI).
 
@@ -52,13 +52,17 @@ def handle_exception(e: Exception, name: str, namespace: Optional[str],
         namespace: Kubernetes namespace
         operation_type: Operation type (legacy parameter, kept for backward compatibility)
         resource_type: Resource type (legacy parameter, kept for backward compatibility)
+        debug: If True, show full Kubernetes exception details
     """
 
     if isinstance(e, ApiException):
         if e.status == 401:
             raise Exception(f"Credentials unauthorized.") from e
         elif e.status == 403:
-            if namespace:
+            if debug and e.body:
+                # Show full Kubernetes error details in debug mode
+                raise Exception(f"Kubernetes API error: {e.body}") from e
+            elif namespace:
                 raise Exception(
                     f"Access denied to resource '{name}' in namespace '{namespace}'."
                 ) from e
@@ -67,7 +71,9 @@ def handle_exception(e: Exception, name: str, namespace: Optional[str],
                     f"Access denied to resource '{name}'."
                 ) from e
         elif e.status == 404:
-            if namespace:
+            if debug and e.body:
+                raise Exception(f"Kubernetes API error: {e.body}") from e
+            elif namespace:
                 # Basic 404 for SDK usage - CLI commands get enhanced 404 via decorator
                 raise Exception(
                     f"Resource '{name}' not found in namespace '{namespace}'. "
@@ -78,7 +84,9 @@ def handle_exception(e: Exception, name: str, namespace: Optional[str],
                     f"Resource '{name}' not found. Please check the resource name."
                 ) from e
         elif e.status == 409:
-            if namespace:
+            if debug and e.body:
+                raise Exception(f"Kubernetes API error: {e.body}") from e
+            elif namespace:
                 raise Exception(
                     f"Resource '{name}' already exists in namespace '{namespace}'."
                 ) from e
@@ -87,9 +95,15 @@ def handle_exception(e: Exception, name: str, namespace: Optional[str],
                     f"Resource '{name}' already exists."
                 ) from e
         elif 500 <= e.status < 600:
-            raise Exception("Kubernetes API internal server error.") from e
+            if debug and e.body:
+                raise Exception(f"Kubernetes API error: {e.body}") from e
+            else:
+                raise Exception("Kubernetes API internal server error.") from e
         else:
-            raise Exception(f"Unhandled Kubernetes error: {e.status} {e.reason}") from e
+            if debug and e.body:
+                raise Exception(f"Kubernetes API error: {e.body}") from e
+            else:
+                raise Exception(f"Unhandled Kubernetes error: {e.status} {e.reason}") from e
 
     if isinstance(e, ValidationError):
         raise Exception("Response did not match expected schema.") from e

src/sagemaker/hyperpod/inference/hp_endpoint_base.py

@@ -127,7 +127,7 @@ def call_create_api(
             )
         except Exception as e:
             logger.error(f"Failed to create endpoint in namespace {metadata.namespace}!")
-            handle_exception(e, metadata.name, metadata.namespace)
+            handle_exception(e, metadata.name, metadata.namespace, debug=debug)
 
     @classmethod
     def call_list_api(

src/sagemaker/hyperpod/space/hyperpod_space.py

@@ -301,7 +301,7 @@ def create(self, debug: bool = False):
             logger.debug(f"Successfully created HyperPod Space '{self.config.name}'!")
         except Exception as e:
             logger.error(f"Failed to create HyperPod Space {self.config.name}!")
-            handle_exception(e, self.config.name, self.config.namespace)
+            handle_exception(e, self.config.name, self.config.namespace, debug=debug)
 
     @classmethod
     @_hyperpod_telemetry_emitter(Feature.HYPERPOD, "list_spaces")

src/sagemaker/hyperpod/training/hyperpod_pytorch_job.py

@@ -275,7 +275,7 @@ def create(self, debug=False):
             logger.info(f"Successfully submitted HyperPodPytorchJob '{self.metadata.name}'!")
         except Exception as e:
             logger.error(f"Failed to create HyperPodPytorchJob {self.metadata.name}!")
-            handle_exception(e, self.metadata.name, self.metadata.namespace)
+            handle_exception(e, self.metadata.name, self.metadata.namespace, debug=debug)