{AzureML} fix for the LocalEndpointImageBuildError (#28461)

navba-MSFT · web-flow · commit 59a58378d045 · 2023-01-30T09:38:25.000-08:00
LocalEndpointImageBuildError: - Local deployment with custom Dockerfile due to wrong path

On this line, absolute_build_directory should be referenced instead of absolute_dockerfile_path

Reproduce:
env = Environment( build=BuildContext(path=".", dockerfile_path="Dockerfile") ) deployment_config=ManagedOnlineDeployment( environment=env, ... ) MLClient.online_deployments.begin_create_or_update(deployment_config, local=True)

AFAIK, LocalEndpointImageBuildError can occur when deploying an endpoint locally and using a custom Dockerfile, if the path to the Dockerfile is incorrect. To resolve this error, you need to check that the path to the Dockerfile specified in the deployment configuration is correct and that the Dockerfile is in the correct location
diff --git a/sdk/ml/azure-ai-ml/azure/ai/ml/_local_endpoints/validators/environment_validator.py b/sdk/ml/azure-ai-ml/azure/ai/ml/_local_endpoints/validators/environment_validator.py
@@ -128,7 +128,7 @@ def _get_local_environment_artifacts(base_path: str, environment: Environment):
             None,
             None,
             None,
-            absolute_dockerfile_path,
+            absolute_build_directory,
             dockerfile_contents,
             environment.inference_config,
         )

Original file line number	Diff line number	Diff line change
`@@ -128,7 +128,7 @@ def _get_local_environment_artifacts(base_path: str, environment: Environment):`
`128`	`128`	`None,`
`129`	`129`	`None,`
`130`	`130`	`None,`
`131`		`- absolute_dockerfile_path,`
	`131`	`+ absolute_build_directory,`
`132`	`132`	`dockerfile_contents,`
`133`	`133`	`environment.inference_config,`
`134`	`134`	`)`