Mvs local mode timeout

CHANGELOG.rst

@@ -9,6 +9,7 @@ CHANGELOG
 * doc-fix: add deprecation warning for current MXNet training script format
 * doc-fix: add docs on deploying TensorFlow model directly from existing model
 * doc-fix: fix code example for using Gzip compression for TensorFlow training data
+* bug-fix: Setting health check timeout limit on local mode to 30s
 
 1.10.0
 ======

src/sagemaker/local/local_session.py

@@ -25,6 +25,8 @@
 from sagemaker.session import Session
 from sagemaker.utils import get_config_value
 
+HEALTH_CHECK_TIMEOUT_LIMIT = 30
+
 logger = logging.getLogger(__name__)
 logger.setLevel(logging.WARNING)
 
@@ -123,7 +125,7 @@ def create_endpoint(self, EndpointName, EndpointConfigName):
         endpoint_url = "http://localhost:%s/ping" % serving_port
         while True:
             i += 1
-            if i >= 10:
+            if i >= HEALTH_CHECK_TIMEOUT_LIMIT:
                 raise RuntimeError("Giving up, endpoint: %s didn't launch correctly" % EndpointName)
 
             logger.info("Checking if endpoint is up, attempt: %s" % i)

tests/unit/test_local_session.py

@@ -204,7 +204,7 @@ def test_create_endpoint(serve, request, LocalSession):
 @patch('urllib3.PoolManager.request', return_value=BAD_RESPONSE)
 @patch('sagemaker.local.local_session.LocalSession')
 @patch('time.sleep')
-def test_create_endpoint_fails(*args):
+def test_create_endpoint_fails(sleep, *args):
     local_sagemaker_client = sagemaker.local.local_session.LocalSagemakerClient()
     local_sagemaker_client.variants = [{'InstanceType': 'ml.c4.99xlarge', 'InitialInstanceCount': 10}]
     local_sagemaker_client.primary_container = {'ModelDataUrl': '/some/model/path',
@@ -214,6 +214,8 @@ def test_create_endpoint_fails(*args):
     with pytest.raises(RuntimeError):
         local_sagemaker_client.create_endpoint('my-endpoint', 'some-endpoint-config')
 
+    assert 29 == sleep.call_count
+
 
 def test_file_input_all_defaults():
     prefix = 'pre'