This is an automated email from the ASF dual-hosted git repository. dongjoon pushed a commit to branch branch-3.0 in repository https://gitbox.apache.org/repos/asf/spark.git
The following commit(s) were added to refs/heads/branch-3.0 by this push: new 808dd8f [SPARK-33371][PYTHON][3.0] Update setup.py and tests for Python 3.9 808dd8f is described below commit 808dd8fe40564e8f4f25c9faf1897a9efbb2cbe1 Author: HyukjinKwon <gurwls...@apache.org> AuthorDate: Sun Nov 8 21:23:46 2020 -0800 [SPARK-33371][PYTHON][3.0] Update setup.py and tests for Python 3.9 ### What changes were proposed in this pull request? This PR is a backport of https://github.com/apache/spark/pull/30277 This PR proposes to fix PySpark to officially support Python 3.9. The main codes already work. We should just note that we support Python 3.9. Also, this PR fixes some minor fixes into the test codes. - `Thread.isAlive` is removed in Python 3.9, and `Thread.is_alive` exists in Python 3.6+, see https://docs.python.org/3/whatsnew/3.9.html#removed - Fixed `TaskContextTestsWithWorkerReuse.test_barrier_with_python_worker_reuse` and `TaskContextTests.test_barrier` to be less flaky. This becomes more flaky in Python 3.9 for some reasons. NOTE that PyArrow does not support Python 3.9 yet. ### Why are the changes needed? To officially support Python 3.9. ### Does this PR introduce _any_ user-facing change? Yes, it officially supports Python 3.9. ### How was this patch tested? Manually ran the tests: ``` $ ./run-tests --python-executable=python Running PySpark tests. Output is in /.../spark/python/unit-tests.log Will test against the following Python executables: ['python'] Will test the following Python modules: ['pyspark-core', 'pyspark-ml', 'pyspark-mllib', 'pyspark-resource', 'pyspark-sql', 'pyspark-streaming'] python python_implementation is CPython python version is: Python 3.9.0 Starting test(python): pyspark.ml.tests.test_base Starting test(python): pyspark.ml.tests.test_evaluation Starting test(python): pyspark.ml.tests.test_algorithms Starting test(python): pyspark.ml.tests.test_feature Finished test(python): pyspark.ml.tests.test_base (12s) Starting test(python): pyspark.ml.tests.test_image Finished test(python): pyspark.ml.tests.test_evaluation (15s) Starting test(python): pyspark.ml.tests.test_linalg Finished test(python): pyspark.ml.tests.test_feature (25s) Starting test(python): pyspark.ml.tests.test_param Finished test(python): pyspark.ml.tests.test_image (17s) Starting test(python): pyspark.ml.tests.test_persistence Finished test(python): pyspark.ml.tests.test_param (17s) Starting test(python): pyspark.ml.tests.test_pipeline Finished test(python): pyspark.ml.tests.test_linalg (30s) Starting test(python): pyspark.ml.tests.test_stat Finished test(python): pyspark.ml.tests.test_pipeline (6s) Starting test(python): pyspark.ml.tests.test_training_summary Finished test(python): pyspark.ml.tests.test_stat (12s) Starting test(python): pyspark.ml.tests.test_tuning Finished test(python): pyspark.ml.tests.test_algorithms (68s) Starting test(python): pyspark.ml.tests.test_wrapper Finished test(python): pyspark.ml.tests.test_persistence (51s) Starting test(python): pyspark.mllib.tests.test_algorithms Finished test(python): pyspark.ml.tests.test_training_summary (33s) Starting test(python): pyspark.mllib.tests.test_feature Finished test(python): pyspark.ml.tests.test_wrapper (19s) Starting test(python): pyspark.mllib.tests.test_linalg Finished test(python): pyspark.mllib.tests.test_feature (26s) Starting test(python): pyspark.mllib.tests.test_stat Finished test(python): pyspark.mllib.tests.test_stat (22s) Starting test(python): pyspark.mllib.tests.test_streaming_algorithms Finished test(python): pyspark.mllib.tests.test_algorithms (53s) Starting test(python): pyspark.mllib.tests.test_util Finished test(python): pyspark.mllib.tests.test_linalg (54s) Starting test(python): pyspark.sql.tests.test_arrow Finished test(python): pyspark.sql.tests.test_arrow (0s) ... 61 tests were skipped Starting test(python): pyspark.sql.tests.test_catalog Finished test(python): pyspark.mllib.tests.test_util (11s) Starting test(python): pyspark.sql.tests.test_column Finished test(python): pyspark.sql.tests.test_catalog (16s) Starting test(python): pyspark.sql.tests.test_conf Finished test(python): pyspark.sql.tests.test_column (17s) Starting test(python): pyspark.sql.tests.test_context Finished test(python): pyspark.sql.tests.test_context (6s) ... 3 tests were skipped Starting test(python): pyspark.sql.tests.test_dataframe Finished test(python): pyspark.sql.tests.test_conf (11s) Starting test(python): pyspark.sql.tests.test_datasources Finished test(python): pyspark.sql.tests.test_datasources (19s) Starting test(python): pyspark.sql.tests.test_functions Finished test(python): pyspark.sql.tests.test_dataframe (35s) ... 3 tests were skipped Starting test(python): pyspark.sql.tests.test_group Finished test(python): pyspark.sql.tests.test_functions (32s) Starting test(python): pyspark.sql.tests.test_pandas_cogrouped_map Finished test(python): pyspark.sql.tests.test_pandas_cogrouped_map (1s) ... 15 tests were skipped Starting test(python): pyspark.sql.tests.test_pandas_grouped_map Finished test(python): pyspark.sql.tests.test_group (19s) Starting test(python): pyspark.sql.tests.test_pandas_map Finished test(python): pyspark.sql.tests.test_pandas_grouped_map (0s) ... 21 tests were skipped Starting test(python): pyspark.sql.tests.test_pandas_udf Finished test(python): pyspark.sql.tests.test_pandas_map (0s) ... 6 tests were skipped Starting test(python): pyspark.sql.tests.test_pandas_udf_grouped_agg Finished test(python): pyspark.sql.tests.test_pandas_udf (0s) ... 6 tests were skipped Starting test(python): pyspark.sql.tests.test_pandas_udf_scalar Finished test(python): pyspark.sql.tests.test_pandas_udf_grouped_agg (0s) ... 13 tests were skipped Starting test(python): pyspark.sql.tests.test_pandas_udf_typehints Finished test(python): pyspark.sql.tests.test_pandas_udf_scalar (0s) ... 50 tests were skipped Starting test(python): pyspark.sql.tests.test_pandas_udf_window Finished test(python): pyspark.sql.tests.test_pandas_udf_typehints (0s) ... 10 tests were skipped Starting test(python): pyspark.sql.tests.test_readwriter Finished test(python): pyspark.sql.tests.test_pandas_udf_window (0s) ... 14 tests were skipped Starting test(python): pyspark.sql.tests.test_serde Finished test(python): pyspark.sql.tests.test_serde (19s) Starting test(python): pyspark.sql.tests.test_session Finished test(python): pyspark.mllib.tests.test_streaming_algorithms (120s) Starting test(python): pyspark.sql.tests.test_streaming Finished test(python): pyspark.sql.tests.test_readwriter (25s) Starting test(python): pyspark.sql.tests.test_types Finished test(python): pyspark.ml.tests.test_tuning (208s) Starting test(python): pyspark.sql.tests.test_udf Finished test(python): pyspark.sql.tests.test_session (31s) Starting test(python): pyspark.sql.tests.test_utils Finished test(python): pyspark.sql.tests.test_streaming (35s) Starting test(python): pyspark.streaming.tests.test_context Finished test(python): pyspark.sql.tests.test_types (34s) Starting test(python): pyspark.streaming.tests.test_dstream Finished test(python): pyspark.sql.tests.test_utils (14s) Starting test(python): pyspark.streaming.tests.test_kinesis Finished test(python): pyspark.streaming.tests.test_kinesis (0s) ... 2 tests were skipped Starting test(python): pyspark.streaming.tests.test_listener Finished test(python): pyspark.streaming.tests.test_listener (11s) Starting test(python): pyspark.tests.test_appsubmit Finished test(python): pyspark.sql.tests.test_udf (39s) Starting test(python): pyspark.tests.test_broadcast Finished test(python): pyspark.streaming.tests.test_context (23s) Starting test(python): pyspark.tests.test_conf Finished test(python): pyspark.tests.test_conf (15s) Starting test(python): pyspark.tests.test_context Finished test(python): pyspark.tests.test_broadcast (33s) Starting test(python): pyspark.tests.test_daemon Finished test(python): pyspark.tests.test_daemon (5s) Starting test(python): pyspark.tests.test_install_spark Finished test(python): pyspark.tests.test_context (44s) Starting test(python): pyspark.tests.test_join Finished test(python): pyspark.tests.test_appsubmit (68s) Starting test(python): pyspark.tests.test_profiler Finished test(python): pyspark.tests.test_join (7s) Starting test(python): pyspark.tests.test_rdd Finished test(python): pyspark.tests.test_profiler (9s) Starting test(python): pyspark.tests.test_rddbarrier Finished test(python): pyspark.tests.test_rddbarrier (7s) Starting test(python): pyspark.tests.test_readwrite Finished test(python): pyspark.streaming.tests.test_dstream (107s) Starting test(python): pyspark.tests.test_serializers Finished test(python): pyspark.tests.test_serializers (8s) Starting test(python): pyspark.tests.test_shuffle Finished test(python): pyspark.tests.test_readwrite (14s) Starting test(python): pyspark.tests.test_taskcontext Finished test(python): pyspark.tests.test_install_spark (65s) Starting test(python): pyspark.tests.test_util Finished test(python): pyspark.tests.test_shuffle (8s) Starting test(python): pyspark.tests.test_worker Finished test(python): pyspark.tests.test_util (5s) Starting test(python): pyspark.accumulators Finished test(python): pyspark.accumulators (5s) Starting test(python): pyspark.broadcast Finished test(python): pyspark.broadcast (6s) Starting test(python): pyspark.conf Finished test(python): pyspark.tests.test_worker (14s) Starting test(python): pyspark.context Finished test(python): pyspark.conf (4s) Starting test(python): pyspark.ml.classification Finished test(python): pyspark.tests.test_rdd (60s) Starting test(python): pyspark.ml.clustering Finished test(python): pyspark.context (21s) Starting test(python): pyspark.ml.evaluation Finished test(python): pyspark.tests.test_taskcontext (69s) Starting test(python): pyspark.ml.feature Finished test(python): pyspark.ml.evaluation (26s) Starting test(python): pyspark.ml.fpm Finished test(python): pyspark.ml.clustering (45s) Starting test(python): pyspark.ml.functions Finished test(python): pyspark.ml.fpm (24s) Starting test(python): pyspark.ml.image Finished test(python): pyspark.ml.functions (17s) Starting test(python): pyspark.ml.linalg.__init__ Finished test(python): pyspark.ml.linalg.__init__ (0s) Starting test(python): pyspark.ml.recommendation Finished test(python): pyspark.ml.classification (74s) Starting test(python): pyspark.ml.regression Finished test(python): pyspark.ml.image (8s) Starting test(python): pyspark.ml.stat Finished test(python): pyspark.ml.stat (29s) Starting test(python): pyspark.ml.tuning Finished test(python): pyspark.ml.regression (53s) Starting test(python): pyspark.mllib.classification Finished test(python): pyspark.ml.tuning (35s) Starting test(python): pyspark.mllib.clustering Finished test(python): pyspark.ml.feature (103s) Starting test(python): pyspark.mllib.evaluation Finished test(python): pyspark.mllib.classification (33s) Starting test(python): pyspark.mllib.feature Finished test(python): pyspark.mllib.evaluation (21s) Starting test(python): pyspark.mllib.fpm Finished test(python): pyspark.ml.recommendation (103s) Starting test(python): pyspark.mllib.linalg.__init__ Finished test(python): pyspark.mllib.linalg.__init__ (1s) Starting test(python): pyspark.mllib.linalg.distributed Finished test(python): pyspark.mllib.feature (26s) Starting test(python): pyspark.mllib.random Finished test(python): pyspark.mllib.fpm (23s) Starting test(python): pyspark.mllib.recommendation Finished test(python): pyspark.mllib.clustering (50s) Starting test(python): pyspark.mllib.regression Finished test(python): pyspark.mllib.random (13s) Starting test(python): pyspark.mllib.stat.KernelDensity Finished test(python): pyspark.mllib.stat.KernelDensity (1s) Starting test(python): pyspark.mllib.stat._statistics Finished test(python): pyspark.mllib.linalg.distributed (42s) Starting test(python): pyspark.mllib.tree Finished test(python): pyspark.mllib.stat._statistics (19s) Starting test(python): pyspark.mllib.util Finished test(python): pyspark.mllib.regression (33s) Starting test(python): pyspark.profiler Finished test(python): pyspark.mllib.recommendation (36s) Starting test(python): pyspark.rdd Finished test(python): pyspark.profiler (9s) Starting test(python): pyspark.resource.tests.test_resources Finished test(python): pyspark.mllib.tree (19s) Starting test(python): pyspark.serializers Finished test(python): pyspark.mllib.util (21s) Starting test(python): pyspark.shuffle Finished test(python): pyspark.resource.tests.test_resources (9s) Starting test(python): pyspark.sql.avro.functions Finished test(python): pyspark.shuffle (1s) Starting test(python): pyspark.sql.catalog Finished test(python): pyspark.rdd (22s) Starting test(python): pyspark.sql.column Finished test(python): pyspark.serializers (12s) Starting test(python): pyspark.sql.conf Finished test(python): pyspark.sql.conf (6s) Starting test(python): pyspark.sql.context Finished test(python): pyspark.sql.catalog (14s) Starting test(python): pyspark.sql.dataframe Finished test(python): pyspark.sql.avro.functions (15s) Starting test(python): pyspark.sql.functions Finished test(python): pyspark.sql.column (24s) Starting test(python): pyspark.sql.group Finished test(python): pyspark.sql.context (20s) Starting test(python): pyspark.sql.pandas.conversion Finished test(python): pyspark.sql.pandas.conversion (13s) Starting test(python): pyspark.sql.pandas.group_ops Finished test(python): pyspark.sql.group (36s) Starting test(python): pyspark.sql.pandas.map_ops Finished test(python): pyspark.sql.pandas.group_ops (21s) Starting test(python): pyspark.sql.pandas.serializers Finished test(python): pyspark.sql.pandas.serializers (0s) Starting test(python): pyspark.sql.pandas.typehints Finished test(python): pyspark.sql.pandas.typehints (0s) Starting test(python): pyspark.sql.pandas.types Finished test(python): pyspark.sql.pandas.types (0s) Starting test(python): pyspark.sql.pandas.utils Finished test(python): pyspark.sql.pandas.utils (0s) Starting test(python): pyspark.sql.readwriter Finished test(python): pyspark.sql.dataframe (56s) Starting test(python): pyspark.sql.session Finished test(python): pyspark.sql.functions (57s) Starting test(python): pyspark.sql.streaming Finished test(python): pyspark.sql.pandas.map_ops (12s) Starting test(python): pyspark.sql.types Finished test(python): pyspark.sql.types (10s) Starting test(python): pyspark.sql.udf Finished test(python): pyspark.sql.streaming (16s) Starting test(python): pyspark.sql.window Finished test(python): pyspark.sql.session (19s) Starting test(python): pyspark.streaming.util Finished test(python): pyspark.streaming.util (0s) Starting test(python): pyspark.util Finished test(python): pyspark.util (0s) Finished test(python): pyspark.sql.readwriter (24s) Finished test(python): pyspark.sql.udf (13s) Finished test(python): pyspark.sql.window (14s) Tests passed in 780 seconds ``` Closes #30288 from HyukjinKwon/SPARK-33371-3.0. Authored-by: HyukjinKwon <gurwls...@apache.org> Signed-off-by: Dongjoon Hyun <dh...@apple.com> --- python/pyspark/tests/test_taskcontext.py | 8 ++++---- python/pyspark/tests/test_worker.py | 2 +- python/setup.py | 1 + 3 files changed, 6 insertions(+), 5 deletions(-) diff --git a/python/pyspark/tests/test_taskcontext.py b/python/pyspark/tests/test_taskcontext.py index 57c880a..b2fa991 100644 --- a/python/pyspark/tests/test_taskcontext.py +++ b/python/pyspark/tests/test_taskcontext.py @@ -127,12 +127,12 @@ class TaskContextTests(PySparkTestCase): def context_barrier(x): tc = BarrierTaskContext.get() - time.sleep(random.randint(1, 10)) + time.sleep(random.randint(1, 5) * 2) tc.barrier() return time.time() times = rdd.barrier().mapPartitions(f).map(context_barrier).collect() - self.assertTrue(max(times) - min(times) < 1) + self.assertTrue(max(times) - min(times) < 2) def test_all_gather(self): """ @@ -235,7 +235,7 @@ class TaskContextTestsWithWorkerReuse(unittest.TestCase): def context_barrier(x): tc = BarrierTaskContext.get() - time.sleep(random.randint(1, 10)) + time.sleep(random.randint(1, 5) * 2) tc.barrier() return (time.time(), os.getpid()) @@ -243,7 +243,7 @@ class TaskContextTestsWithWorkerReuse(unittest.TestCase): times = list(map(lambda x: x[0], result)) pids = list(map(lambda x: x[1], result)) # check both barrier and worker reuse effect - self.assertTrue(max(times) - min(times) < 1) + self.assertTrue(max(times) - min(times) < 2) for pid in pids: self.assertTrue(pid in worker_pids) diff --git a/python/pyspark/tests/test_worker.py b/python/pyspark/tests/test_worker.py index 9d7deb2..bfcbc43 100644 --- a/python/pyspark/tests/test_worker.py +++ b/python/pyspark/tests/test_worker.py @@ -137,7 +137,7 @@ class WorkerTests(ReusedPySparkTestCase): t.daemon = True t.start() t.join(5) - self.assertTrue(not t.isAlive()) + self.assertTrue(not t.is_alive()) self.assertEqual(100000, rdd.count()) def test_with_different_versions_of_python(self): diff --git a/python/setup.py b/python/setup.py index a1cde31..78b5a69 100755 --- a/python/setup.py +++ b/python/setup.py @@ -226,6 +226,7 @@ try: 'Programming Language :: Python :: 3.6', 'Programming Language :: Python :: 3.7', 'Programming Language :: Python :: 3.8', + 'Programming Language :: Python :: 3.9', 'Programming Language :: Python :: Implementation :: CPython', 'Programming Language :: Python :: Implementation :: PyPy'] ) --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@spark.apache.org For additional commands, e-mail: commits-h...@spark.apache.org