fix additional p2 and job/endpoint naming issues (#544)

Author: jesterhazy

tests/integ/__init__.py

@@ -23,9 +23,11 @@
 TUNING_DEFAULT_TIMEOUT_MINUTES = 20
 TRANSFORM_DEFAULT_TIMEOUT_MINUTES = 20
 PYTHON_VERSION = 'py' + str(sys.version_info.major)
-HOSTING_P2_UNAVAILABLE_REGIONS = ['ca-central-1', 'us-west-1', 'eu-west-2']
-HOSTING_P3_UNAVAILABLE_REGIONS = ['ap-southeast-1', 'ap-southeast-2', 'ap-south-1', 'ca-central-1',
-                                  'eu-west-2', 'us-west-1']
+
+# 'eu-central-1' has some p2, but no enough for continuous testing
+HOSTING_NO_P2_REGIONS = ['ca-central-1', 'eu-west-2', 'us-west-1', 'eu-central-1']
+HOSTING_NO_P3_REGIONS = ['ap-southeast-1', 'ap-southeast-2', 'ap-south-1', 'ca-central-1',
+                         'eu-west-2', 'us-west-1']
 
 logging.getLogger('boto3').setLevel(logging.INFO)
 logging.getLogger('botocore').setLevel(logging.INFO)

tests/integ/test_byo_estimator.py

@@ -23,7 +23,7 @@
 import sagemaker
 from sagemaker.amazon.amazon_estimator import registry
 from sagemaker.estimator import Estimator
-from sagemaker.utils import name_from_base
+from sagemaker.utils import unique_name_from_base
 from tests.integ import DATA_DIR, TRAINING_DEFAULT_TIMEOUT_MINUTES
 from tests.integ.timeout import timeout, timeout_and_delete_endpoint_by_name
 
@@ -81,7 +81,7 @@ def test_byo_estimator(sagemaker_session, region):
         # training labels must be 'float32'
         estimator.fit({'train': s3_train_data})
 
-    endpoint_name = name_from_base('byo')
+    endpoint_name = unique_name_from_base('byo')
 
     with timeout_and_delete_endpoint_by_name(endpoint_name, sagemaker_session):
         model = estimator.create_model()
@@ -99,7 +99,7 @@ def test_byo_estimator(sagemaker_session, region):
 
 def test_async_byo_estimator(sagemaker_session, region):
     image_name = registry(region) + "/factorization-machines:1"
-    endpoint_name = name_from_base('byo')
+    endpoint_name = unique_name_from_base('byo')
     training_data_path = os.path.join(DATA_DIR, 'dummy_tensor')
     training_job_name = ""
 

tests/integ/test_chainer_train.py

@@ -36,8 +36,8 @@ def test_distributed_cpu_training(sagemaker_session, chainer_full_version):
     _run_mnist_training_job(sagemaker_session, "ml.c4.xlarge", 2, chainer_full_version)
 
 
-@pytest.mark.skipif(tests.integ.test_region() in ['us-west-1', 'eu-west-2', 'ca-central-1'],
-                    reason='No ml.p2.xlarge supported in these regions')
+@pytest.mark.skipif(tests.integ.test_region() in tests.integ.HOSTING_NO_P2_REGIONS,
+                    reason='no ml.p2 instances in these regions')
 def test_distributed_gpu_training(sagemaker_session, chainer_full_version):
     _run_mnist_training_job(sagemaker_session, "ml.p2.xlarge", 2, chainer_full_version)
 

tests/integ/test_factorization_machines.py

@@ -21,7 +21,7 @@
 import pytest
 
 from sagemaker import FactorizationMachines, FactorizationMachinesModel
-from sagemaker.utils import name_from_base
+from sagemaker.utils import unique_name_from_base
 from tests.integ import DATA_DIR, TRAINING_DEFAULT_TIMEOUT_MINUTES
 from tests.integ.timeout import timeout, timeout_and_delete_endpoint_by_name
 
@@ -45,7 +45,7 @@ def test_factorization_machines(sagemaker_session):
         # training labels must be 'float32'
         fm.fit(fm.record_set(train_set[0][:200], train_set[1][:200].astype('float32')))
 
-    endpoint_name = name_from_base('fm')
+    endpoint_name = unique_name_from_base('fm')
     with timeout_and_delete_endpoint_by_name(endpoint_name, sagemaker_session):
         model = FactorizationMachinesModel(fm.model_data, role='SageMakerRole', sagemaker_session=sagemaker_session)
         predictor = model.deploy(1, 'ml.c4.xlarge', endpoint_name=endpoint_name)
@@ -58,7 +58,7 @@ def test_factorization_machines(sagemaker_session):
 
 def test_async_factorization_machines(sagemaker_session):
     training_job_name = ""
-    endpoint_name = name_from_base('factorizationMachines')
+    endpoint_name = unique_name_from_base('factorizationMachines')
 
     with timeout(minutes=5):
         data_path = os.path.join(DATA_DIR, 'one_p_mnist', 'mnist.pkl.gz')

tests/integ/test_ipinsights.py

@@ -17,7 +17,7 @@
 
 from sagemaker import IPInsights, IPInsightsModel
 from sagemaker.predictor import RealTimePredictor
-from sagemaker.utils import name_from_base
+from sagemaker.utils import unique_name_from_base
 from tests.integ import DATA_DIR, TRAINING_DEFAULT_TIMEOUT_MINUTES
 from tests.integ.record_set import prepare_record_set_from_local_files
 from tests.integ.timeout import timeout, timeout_and_delete_endpoint_by_name
@@ -47,7 +47,7 @@ def test_ipinsights(sagemaker_session):
                                                          num_records, FEATURE_DIM, sagemaker_session)
         ipinsights.fit(record_set, None)
 
-    endpoint_name = name_from_base('ipinsights')
+    endpoint_name = unique_name_from_base('ipinsights')
     with timeout_and_delete_endpoint_by_name(endpoint_name, sagemaker_session):
         model = IPInsightsModel(ipinsights.model_data, role='SageMakerRole', sagemaker_session=sagemaker_session)
         predictor = model.deploy(1, 'ml.c4.xlarge', endpoint_name=endpoint_name)

tests/integ/test_kmeans.py

@@ -21,7 +21,7 @@
 import pytest
 
 from sagemaker import KMeans, KMeansModel
-from sagemaker.utils import name_from_base
+from sagemaker.utils import unique_name_from_base
 from tests.integ import DATA_DIR, TRAINING_DEFAULT_TIMEOUT_MINUTES
 from tests.integ.timeout import timeout, timeout_and_delete_endpoint_by_name
 
@@ -64,7 +64,7 @@ def test_kmeans(sagemaker_session):
 
         kmeans.fit(kmeans.record_set(train_set[0][:100]))
 
-    endpoint_name = name_from_base('kmeans')
+    endpoint_name = unique_name_from_base('kmeans')
     with timeout_and_delete_endpoint_by_name(endpoint_name, sagemaker_session):
         model = KMeansModel(kmeans.model_data, role='SageMakerRole', sagemaker_session=sagemaker_session)
         predictor = model.deploy(1, 'ml.c4.xlarge', endpoint_name=endpoint_name)
@@ -78,7 +78,7 @@ def test_kmeans(sagemaker_session):
 
 def test_async_kmeans(sagemaker_session):
     training_job_name = ""
-    endpoint_name = name_from_base('kmeans')
+    endpoint_name = unique_name_from_base('kmeans')
 
     with timeout(minutes=5):
         data_path = os.path.join(DATA_DIR, 'one_p_mnist', 'mnist.pkl.gz')

tests/integ/test_knn.py

@@ -21,7 +21,7 @@
 import pytest
 
 from sagemaker import KNN, KNNModel
-from sagemaker.utils import name_from_base
+from sagemaker.utils import unique_name_from_base
 from tests.integ import DATA_DIR, TRAINING_DEFAULT_TIMEOUT_MINUTES
 from tests.integ.timeout import timeout, timeout_and_delete_endpoint_by_name
 
@@ -44,7 +44,7 @@ def test_knn_regressor(sagemaker_session):
         # training labels must be 'float32'
         knn.fit(knn.record_set(train_set[0][:200], train_set[1][:200].astype('float32')))
 
-    endpoint_name = name_from_base('knn')
+    endpoint_name = unique_name_from_base('knn')
     with timeout_and_delete_endpoint_by_name(endpoint_name, sagemaker_session):
         model = KNNModel(knn.model_data, role='SageMakerRole', sagemaker_session=sagemaker_session)
         predictor = model.deploy(1, 'ml.c4.xlarge', endpoint_name=endpoint_name)
@@ -57,7 +57,7 @@ def test_knn_regressor(sagemaker_session):
 
 def test_async_knn_classifier(sagemaker_session):
     training_job_name = ""
-    endpoint_name = name_from_base('knn')
+    endpoint_name = unique_name_from_base('knn')
 
     with timeout(minutes=5):
         data_path = os.path.join(DATA_DIR, 'one_p_mnist', 'mnist.pkl.gz')

tests/integ/test_lda.py

@@ -19,7 +19,7 @@
 
 from sagemaker import LDA, LDAModel
 from sagemaker.amazon.common import read_records
-from sagemaker.utils import name_from_base
+from sagemaker.utils import unique_name_from_base
 from tests.integ import DATA_DIR, TRAINING_DEFAULT_TIMEOUT_MINUTES
 from tests.integ.timeout import timeout, timeout_and_delete_endpoint_by_name
 from tests.integ.record_set import prepare_record_set_from_local_files
@@ -44,7 +44,7 @@ def test_lda(sagemaker_session):
                                                          len(all_records), feature_num, sagemaker_session)
         lda.fit(record_set, 100)
 
-    endpoint_name = name_from_base('lda')
+    endpoint_name = unique_name_from_base('lda')
     with timeout_and_delete_endpoint_by_name(endpoint_name, sagemaker_session):
         model = LDAModel(lda.model_data, role='SageMakerRole', sagemaker_session=sagemaker_session)
         predictor = model.deploy(1, 'ml.c4.xlarge', endpoint_name=endpoint_name)

tests/integ/test_linear_learner.py

@@ -22,7 +22,7 @@
 import pytest
 
 from sagemaker.amazon.linear_learner import LinearLearner, LinearLearnerModel
-from sagemaker.utils import name_from_base, sagemaker_timestamp
+from sagemaker.utils import unique_name_from_base, sagemaker_timestamp
 from tests.integ import DATA_DIR, TRAINING_DEFAULT_TIMEOUT_MINUTES
 from tests.integ.timeout import timeout, timeout_and_delete_endpoint_by_name
 
@@ -80,7 +80,7 @@ def test_linear_learner(sagemaker_session):
         ll.early_stopping_patience = 3
         ll.fit(ll.record_set(train_set[0][:200], train_set[1][:200]))
 
-    endpoint_name = name_from_base('linear-learner')
+    endpoint_name = unique_name_from_base('linear-learner')
     with timeout_and_delete_endpoint_by_name(endpoint_name, sagemaker_session):
 
         predictor = ll.deploy(1, 'ml.c4.xlarge', endpoint_name=endpoint_name)
@@ -109,7 +109,7 @@ def test_linear_learner_multiclass(sagemaker_session):
         ll.epochs = 1
         ll.fit(ll.record_set(train_set[0][:200], train_set[1][:200]))
 
-    endpoint_name = name_from_base('linear-learner')
+    endpoint_name = unique_name_from_base('linear-learner')
     with timeout_and_delete_endpoint_by_name(endpoint_name, sagemaker_session):
 
         predictor = ll.deploy(1, 'ml.c4.xlarge', endpoint_name=endpoint_name)

tests/integ/test_ntm.py

@@ -19,7 +19,7 @@
 
 from sagemaker import NTM, NTMModel
 from sagemaker.amazon.common import read_records
-from sagemaker.utils import name_from_base
+from sagemaker.utils import unique_name_from_base
 from tests.integ import DATA_DIR, TRAINING_DEFAULT_TIMEOUT_MINUTES
 from tests.integ.timeout import timeout, timeout_and_delete_endpoint_by_name
 from tests.integ.record_set import prepare_record_set_from_local_files
@@ -44,7 +44,7 @@ def test_ntm(sagemaker_session):
                                                          len(all_records), feature_num, sagemaker_session)
         ntm.fit(record_set, None)
 
-    endpoint_name = name_from_base('ntm')
+    endpoint_name = unique_name_from_base('ntm')
     with timeout_and_delete_endpoint_by_name(endpoint_name, sagemaker_session):
         model = NTMModel(ntm.model_data, role='SageMakerRole', sagemaker_session=sagemaker_session)
         predictor = model.deploy(1, 'ml.c4.xlarge', endpoint_name=endpoint_name)

tests/integ/test_object2vec.py

@@ -17,7 +17,7 @@
 
 from sagemaker.predictor import RealTimePredictor
 from sagemaker import Object2Vec, Object2VecModel
-from sagemaker.utils import name_from_base
+from sagemaker.utils import unique_name_from_base
 from tests.integ import DATA_DIR, TRAINING_DEFAULT_TIMEOUT_MINUTES
 from tests.integ.timeout import timeout, timeout_and_delete_endpoint_by_name
 from tests.integ.record_set import prepare_record_set_from_local_files
@@ -52,7 +52,7 @@ def test_object2vec(sagemaker_session):
 
         object2vec.fit(record_set, None)
 
-    endpoint_name = name_from_base('object2vec')
+    endpoint_name = unique_name_from_base('object2vec')
 
     with timeout_and_delete_endpoint_by_name(endpoint_name, sagemaker_session):
         model = Object2VecModel(object2vec.model_data, role='SageMakerRole', sagemaker_session=sagemaker_session)

tests/integ/test_pca.py

@@ -21,7 +21,7 @@
 import pytest
 
 import sagemaker.amazon.pca
-from sagemaker.utils import name_from_base
+from sagemaker.utils import unique_name_from_base
 from tests.integ import DATA_DIR, TRAINING_DEFAULT_TIMEOUT_MINUTES
 from tests.integ.timeout import timeout, timeout_and_delete_endpoint_by_name
 
@@ -45,7 +45,7 @@ def test_pca(sagemaker_session):
         pca.extra_components = 5
         pca.fit(pca.record_set(train_set[0][:100]))
 
-    endpoint_name = name_from_base('pca')
+    endpoint_name = unique_name_from_base('pca')
     with timeout_and_delete_endpoint_by_name(endpoint_name, sagemaker_session):
         pca_model = sagemaker.amazon.pca.PCAModel(model_data=pca.model_data, role='SageMakerRole',
                                                   sagemaker_session=sagemaker_session)
@@ -61,7 +61,7 @@ def test_pca(sagemaker_session):
 
 def test_async_pca(sagemaker_session):
     training_job_name = ""
-    endpoint_name = name_from_base('pca')
+    endpoint_name = unique_name_from_base('pca')
 
     with timeout(minutes=5):
         data_path = os.path.join(DATA_DIR, 'one_p_mnist', 'mnist.pkl.gz')

tests/integ/test_pytorch_train.py

@@ -75,8 +75,8 @@ def test_deploy_model(pytorch_training_job, sagemaker_session):
         assert output.shape == (batch_size, 10)
 
 
-@pytest.mark.skipif(tests.integ.test_region() in ['us-west-1', 'eu-west-2', 'ca-central-1'],
-                    reason='No ml.p2.xlarge supported in these regions')
+@pytest.mark.skipif(tests.integ.test_region() in tests.integ.HOSTING_NO_P2_REGIONS,
+                    reason='no ml.p2 instances in these regions')
 def test_async_fit_deploy(sagemaker_session, pytorch_full_version):
     training_job_name = ""
     # TODO: add tests against local mode when it's ready to be used

tests/integ/test_randomcutforest.py

@@ -16,7 +16,7 @@
 import pytest
 
 from sagemaker import RandomCutForest, RandomCutForestModel
-from sagemaker.utils import name_from_base
+from sagemaker.utils import unique_name_from_base
 from tests.integ import TRAINING_DEFAULT_TIMEOUT_MINUTES
 from tests.integ.timeout import timeout, timeout_and_delete_endpoint_by_name
 
@@ -34,7 +34,7 @@ def test_randomcutforest(sagemaker_session):
 
         rcf.fit(rcf.record_set(train_input))
 
-    endpoint_name = name_from_base('randomcutforest')
+    endpoint_name = unique_name_from_base('randomcutforest')
     with timeout_and_delete_endpoint_by_name(endpoint_name, sagemaker_session):
         model = RandomCutForestModel(rcf.model_data, role='SageMakerRole', sagemaker_session=sagemaker_session)
         predictor = model.deploy(1, 'ml.c4.xlarge', endpoint_name=endpoint_name)

tests/integ/test_tf_cifar.py

@@ -37,8 +37,8 @@ def __call__(self, data):
 @pytest.mark.continuous_testing
 @pytest.mark.skipif(tests.integ.PYTHON_VERSION != 'py2',
                     reason="TensorFlow image supports only python 2.")
-@pytest.mark.skipif(tests.integ.test_region() in ['us-west-1', 'eu-west-2', 'ca-central-1'],
-                    reason='No ml.p2.xlarge supported in these regions')
+@pytest.mark.skipif(tests.integ.test_region() in tests.integ.HOSTING_NO_P2_REGIONS,
+                    reason='no ml.p2 instances in these regions')
 def test_cifar(sagemaker_session, tf_full_version):
     with timeout(minutes=45):
         script_path = os.path.join(tests.integ.DATA_DIR, 'cifar_10', 'source')

tests/integ/test_tf_keras.py

@@ -26,8 +26,8 @@
 @pytest.mark.continuous_testing
 @pytest.mark.skipif(tests.integ.PYTHON_VERSION != 'py2',
                     reason="TensorFlow image supports only python 2.")
-@pytest.mark.skipif(tests.integ.test_region() in ['us-west-1', 'eu-west-2', 'ca-central-1'],
-                    reason='No ml.p2.xlarge supported in these regions')
+@pytest.mark.skipif(tests.integ.test_region() in tests.integ.HOSTING_NO_P2_REGIONS,
+                    reason='no ml.p2 instances in these regions')
 def test_keras(sagemaker_session, tf_full_version):
     script_path = os.path.join(tests.integ.DATA_DIR, 'cifar_10', 'source')
     dataset_path = os.path.join(tests.integ.DATA_DIR, 'cifar_10', 'data')

tests/integ/test_tfs.py

@@ -26,7 +26,7 @@
     'ml.c5.xlarge',
     pytest.param('ml.p3.2xlarge',
                  marks=pytest.mark.skipif(
-                     tests.integ.test_region() in tests.integ.HOSTING_P3_UNAVAILABLE_REGIONS,
+                     tests.integ.test_region() in tests.integ.HOSTING_NO_P3_REGIONS,
                      reason='no ml.p3 instances in this region'))])
 def instance_type(request):
     return request.param