[SPARK-41376][CORE][3.2] Correct the Netty preferDirectBufs check logic on executor start #38982
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
… executor start ### What changes were proposed in this pull request? Fix the condition for judging Netty prefer direct memory on executor start, the logic should match `org.apache.spark.network.client.TransportClientFactory`. ### Why are the changes needed? The check logical was added in SPARK-27991, the original intention is to avoid potential Netty OOM issue when Netty uses direct memory to consume shuffle data, but the condition is not sufficient, this PR updates the logic to match `org.apache.spark.network.client.TransportClientFactory` ### Does this PR introduce _any_ user-facing change? No. ### How was this patch tested? Manual testing. Closes apache#38901 from pan3793/SPARK-41376. Authored-by: Cheng Pan <[email protected]> Signed-off-by: Sean Owen <[email protected]>
Member
|
The linter failure is irrelevant, but there is PySpark failure. Could you re-trigger the failed job? |
Member
Author
|
Re-triggered. |
Member
Author
|
@dongjoon-hyun the pyspark and lint CI fail consistently, and I see there were also failed on previous commits. Sorry I'm not familiar w/ Python, cc @zhengruifeng @Yikun would you please take a look what happened on branch-3.2? Thanks. |
Member
Member
|
For the pyspark failare, let's see Yikun@3840beb works or not: https://github.com/Yikun/spark/actions/runs/3655117402/jobs/6176152837 If test passed, I can submit the PR to upstream. |
This was referenced Dec 9, 2022
Yikun
added a commit
that referenced
this pull request
Dec 9, 2022
### What changes were proposed in this pull request? According to https://pandas.pydata.org/docs/reference/api/pandas.io.formats.style.Styler.to_latex.html: `pandas.io.formats.style.Styler.to_latex` introduced since 1.3.0, so before panda 1.3.0, should skip the check ``` ERROR [0.180s]: test_style (pyspark.pandas.tests.test_dataframe.DataFrameTest) ---------------------------------------------------------------------- Traceback (most recent call last): File "/__w/spark/spark/python/pyspark/pandas/tests/test_dataframe.py", line 5795, in test_style check_style() File "/__w/spark/spark/python/pyspark/pandas/tests/test_dataframe.py", line 5793, in check_style self.assert_eq(pdf_style.to_latex(), psdf_style.to_latex()) AttributeError: 'Styler' object has no attribute 'to_latex' ``` Related: 58375a8 ### Why are the changes needed? This test break the 3.2 branch pyspark test (with python 3.6 + pandas 1.1.x), so I think better add the `skipIf` it. See also #38982 (comment) ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? - CI passed - Test on 3.2 branch: Yikun#194, https://github.com/Yikun/spark/actions/runs/3655564439/jobs/6177030747 Closes #39002 from Yikun/skip-check. Authored-by: Yikun Jiang <[email protected]> Signed-off-by: Yikun Jiang <[email protected]>
Yikun
added a commit
that referenced
this pull request
Dec 9, 2022
### What changes were proposed in this pull request? According to https://pandas.pydata.org/docs/reference/api/pandas.io.formats.style.Styler.to_latex.html: `pandas.io.formats.style.Styler.to_latex` introduced since 1.3.0, so before panda 1.3.0, should skip the check ``` ERROR [0.180s]: test_style (pyspark.pandas.tests.test_dataframe.DataFrameTest) ---------------------------------------------------------------------- Traceback (most recent call last): File "/__w/spark/spark/python/pyspark/pandas/tests/test_dataframe.py", line 5795, in test_style check_style() File "/__w/spark/spark/python/pyspark/pandas/tests/test_dataframe.py", line 5793, in check_style self.assert_eq(pdf_style.to_latex(), psdf_style.to_latex()) AttributeError: 'Styler' object has no attribute 'to_latex' ``` Related: 58375a8 ### Why are the changes needed? This test break the 3.2 branch pyspark test (with python 3.6 + pandas 1.1.x), so I think better add the `skipIf` it. See also #38982 (comment) ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? - CI passed Closes #39007 from Yikun/branch-3.3-check. Authored-by: Yikun Jiang <[email protected]> Signed-off-by: Yikun Jiang <[email protected]>
Yikun
added a commit
that referenced
this pull request
Dec 9, 2022
### What changes were proposed in this pull request? According to https://pandas.pydata.org/docs/reference/api/pandas.io.formats.style.Styler.to_latex.html: `pandas.io.formats.style.Styler.to_latex` introduced since 1.3.0, so before panda 1.3.0, should skip the check ``` ERROR [0.180s]: test_style (pyspark.pandas.tests.test_dataframe.DataFrameTest) ---------------------------------------------------------------------- Traceback (most recent call last): File "/__w/spark/spark/python/pyspark/pandas/tests/test_dataframe.py", line 5795, in test_style check_style() File "/__w/spark/spark/python/pyspark/pandas/tests/test_dataframe.py", line 5793, in check_style self.assert_eq(pdf_style.to_latex(), psdf_style.to_latex()) AttributeError: 'Styler' object has no attribute 'to_latex' ``` Related: 58375a8 ### Why are the changes needed? This test break the 3.2 branch pyspark test (with python 3.6 + pandas 1.1.x), so I think better add the `skipIf` it. See also #38982 (comment) ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? CI passed Closes #39008 from Yikun/branch-3.2-style-check. Authored-by: Yikun Jiang <[email protected]> Signed-off-by: Yikun Jiang <[email protected]>
dongjoon-hyun
pushed a commit
that referenced
this pull request
Dec 9, 2022
…ic on executor start ### What changes were proposed in this pull request? Backport #38901 to branch-3.2. Fix the condition for judging Netty prefer direct memory on executor start, the logic should match `org.apache.spark.network.client.TransportClientFactory`. ### Why are the changes needed? The check logical was added in SPARK-27991, the original intention is to avoid potential Netty OOM issue when Netty uses direct memory to consume shuffle data, but the condition is not sufficient, this PR updates the logic to match `org.apache.spark.network.client.TransportClientFactory` ### Does this PR introduce _any_ user-facing change? No. ### How was this patch tested? Manual testing. Closes #38982 from pan3793/SPARK-41376-3.2. Authored-by: Cheng Pan <[email protected]> Signed-off-by: Dongjoon Hyun <[email protected]>
Member
beliefer
pushed a commit
to beliefer/spark
that referenced
this pull request
Dec 18, 2022
### What changes were proposed in this pull request? According to https://pandas.pydata.org/docs/reference/api/pandas.io.formats.style.Styler.to_latex.html: `pandas.io.formats.style.Styler.to_latex` introduced since 1.3.0, so before panda 1.3.0, should skip the check ``` ERROR [0.180s]: test_style (pyspark.pandas.tests.test_dataframe.DataFrameTest) ---------------------------------------------------------------------- Traceback (most recent call last): File "/__w/spark/spark/python/pyspark/pandas/tests/test_dataframe.py", line 5795, in test_style check_style() File "/__w/spark/spark/python/pyspark/pandas/tests/test_dataframe.py", line 5793, in check_style self.assert_eq(pdf_style.to_latex(), psdf_style.to_latex()) AttributeError: 'Styler' object has no attribute 'to_latex' ``` Related: apache@58375a8 ### Why are the changes needed? This test break the 3.2 branch pyspark test (with python 3.6 + pandas 1.1.x), so I think better add the `skipIf` it. See also apache#38982 (comment) ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? - CI passed - Test on 3.2 branch: Yikun#194, https://github.com/Yikun/spark/actions/runs/3655564439/jobs/6177030747 Closes apache#39002 from Yikun/skip-check. Authored-by: Yikun Jiang <[email protected]> Signed-off-by: Yikun Jiang <[email protected]>
sunchao
pushed a commit
to sunchao/spark
that referenced
this pull request
Jun 2, 2023
### What changes were proposed in this pull request? According to https://pandas.pydata.org/docs/reference/api/pandas.io.formats.style.Styler.to_latex.html: `pandas.io.formats.style.Styler.to_latex` introduced since 1.3.0, so before panda 1.3.0, should skip the check ``` ERROR [0.180s]: test_style (pyspark.pandas.tests.test_dataframe.DataFrameTest) ---------------------------------------------------------------------- Traceback (most recent call last): File "/__w/spark/spark/python/pyspark/pandas/tests/test_dataframe.py", line 5795, in test_style check_style() File "/__w/spark/spark/python/pyspark/pandas/tests/test_dataframe.py", line 5793, in check_style self.assert_eq(pdf_style.to_latex(), psdf_style.to_latex()) AttributeError: 'Styler' object has no attribute 'to_latex' ``` Related: apache@58375a8 ### Why are the changes needed? This test break the 3.2 branch pyspark test (with python 3.6 + pandas 1.1.x), so I think better add the `skipIf` it. See also apache#38982 (comment) ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? CI passed Closes apache#39008 from Yikun/branch-3.2-style-check. Authored-by: Yikun Jiang <[email protected]> Signed-off-by: Yikun Jiang <[email protected]>
sunchao
pushed a commit
to sunchao/spark
that referenced
this pull request
Jun 2, 2023
…ic on executor start ### What changes were proposed in this pull request? Backport apache#38901 to branch-3.2. Fix the condition for judging Netty prefer direct memory on executor start, the logic should match `org.apache.spark.network.client.TransportClientFactory`. ### Why are the changes needed? The check logical was added in SPARK-27991, the original intention is to avoid potential Netty OOM issue when Netty uses direct memory to consume shuffle data, but the condition is not sufficient, this PR updates the logic to match `org.apache.spark.network.client.TransportClientFactory` ### Does this PR introduce _any_ user-facing change? No. ### How was this patch tested? Manual testing. Closes apache#38982 from pan3793/SPARK-41376-3.2. Authored-by: Cheng Pan <[email protected]> Signed-off-by: Dongjoon Hyun <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What changes were proposed in this pull request?
Backport #38901 to branch-3.2.
Fix the condition for judging Netty prefer direct memory on executor start, the logic should match
org.apache.spark.network.client.TransportClientFactory.Why are the changes needed?
The check logical was added in SPARK-27991, the original intention is to avoid potential Netty OOM issue when Netty uses direct memory to consume shuffle data, but the condition is not sufficient, this PR updates the logic to match
org.apache.spark.network.client.TransportClientFactoryDoes this PR introduce any user-facing change?
No.
How was this patch tested?
Manual testing.