You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@spark.apache.org by "pan3793 (via GitHub)" <gi...@apache.org> on 2023/02/20 05:48:08 UTC

[GitHub] [spark] pan3793 opened a new pull request, #40091: [SPARK-41952] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160

pan3793 opened a new pull request, #40091:
URL: https://github.com/apache/spark/pull/40091

   <!--
   Thanks for sending a pull request!  Here are some tips for you:
     1. If this is your first time, please read our contributor guidelines: https://spark.apache.org/contributing.html
     2. Ensure you have added or run the appropriate tests for your PR: https://spark.apache.org/developer-tools.html
     3. If the PR is unfinished, add '[WIP]' in your PR title, e.g., '[WIP][SPARK-XXXX] Your PR title ...'.
     4. Be sure to keep the PR description updated to reflect all changes.
     5. Please write your PR title to summarize what this PR proposes.
     6. If possible, provide a concise example to reproduce the issue for a faster review.
     7. If you want to add a new configuration, please read the guideline first for naming configurations in
        'core/src/main/scala/org/apache/spark/internal/config/ConfigEntry.scala'.
     8. If you want to add or modify an error type or message, please read the guideline first in
        'core/src/main/resources/error/README.md'.
   -->
   
   ### What changes were proposed in this pull request?
   <!--
   Please clarify what changes you are proposing. The purpose of this section is to outline the changes and how this PR fixes the issue. 
   If possible, please consider writing useful notes for better and faster reviews in your PR. See the examples below.
     1. If you refactor some codes with changing classes, showing the class hierarchy will help reviewers.
     2. If you fix some SQL features, you can provide some references of other DBMSes.
     3. If there is design documentation, please add the link.
     4. If there is a discussion in the mailing list, please add the link.
   -->
   SPARK-41952 was raised for a while, but unfortunately, the Parquet community has not published the patched version yet, as a workaround, we can fix the issue on the Spark side first.
   
   We encountered this memory issue when migrating data from parquet/snappy to parquet/zstd, Spark executors always occupy unreasonable off-heap memory and have a high risk of being killed by NM.
   
   See more discussions at https://github.com/apache/parquet-mr/pull/982 and https://github.com/apache/iceberg/pull/5681
   
   ### Why are the changes needed?
   <!--
   Please clarify why the changes are needed. For instance,
     1. If you propose a new API, clarify the use case for a new API.
     2. If you fix a bug, you can clarify why it is a bug.
   -->
   The issue is fixed in the parquet community [PARQUET-2160](https://issues.apache.org/jira/browse/PARQUET-2160), but the patched version is not available yet.
   
   ### Does this PR introduce _any_ user-facing change?
   <!--
   Note that it means *any* user-facing change including all aspects such as the documentation fix.
   If yes, please clarify the previous behavior and the change this PR proposes - provide the console output, description and/or an example to show the behavior difference if possible.
   If possible, please also clarify if this is a user-facing change compared to the released Spark versions or within the unreleased branches such as master.
   If no, write 'No'.
   -->
   Yes, it's bug fix.
   
   ### How was this patch tested?
   <!--
   If tests were added, say they were added here. Please make sure to add some test cases that check the changes thoroughly including negative and positive cases if possible.
   If it was tested in a way different from regular unit tests, please clarify how you tested step by step, ideally copy and paste-able, so that other reviewers can test and check, and descendants can verify in the future.
   If tests were not added, please describe why they were not added and/or why it was difficult to add.
   If benchmark tests were added, please run the benchmarks in GitHub Actions for the consistent environment, and the instructions could accord to: https://spark.apache.org/developer-tools.html#github-workflow-benchmarks.
   -->
   The existing UT should cover the correctness check, I'm going to run it w/ a large table scan to valid the memory leak get fixed.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

Re: [PR] [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160 [spark]

Posted by "pan3793 (via GitHub)" <gi...@apache.org>.
pan3793 commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1786555127

   @yujhe Oops, I totally forgot this one...
   
   Your analysis makes sense, I took a look at the parquet non-vectorized reading code path, injecting such a workaround does not clean as we did in this PR. As this is not a regression, and the new Spark version adopts new Parquet version which already fixed this issue, I prefer to keep things as-is in branch-3.3.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

[GitHub] [spark] sunchao commented on pull request #40091: [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160

Posted by "sunchao (via GitHub)" <gi...@apache.org>.
sunchao commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1437363350

   Thanks! merged to master/branch-3.3/branch-3.2


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

Re: [PR] [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160 [spark]

Posted by "pan3793 (via GitHub)" <gi...@apache.org>.
pan3793 commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1756848790

   @yujhe possible to provide a text thread stack? I will find some time to take a look later


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

[GitHub] [spark] LuciferYang commented on pull request #40091: [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160

Posted by "LuciferYang (via GitHub)" <gi...@apache.org>.
LuciferYang commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1436389833

   banch-3.3/3.2 use parquet 1.12.2, if this fix is accepted, would you mind submitting pr for these two branches? @pan3793 
   
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

[GitHub] [spark] pan3793 commented on pull request #40091: [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160

Posted by "pan3793 (via GitHub)" <gi...@apache.org>.
pan3793 commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1436414874

   I verified this patch by scanning a large parquet/zstd table and updated the PR description.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

[GitHub] [spark] LuciferYang commented on pull request #40091: [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160

Posted by "LuciferYang (via GitHub)" <gi...@apache.org>.
LuciferYang commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1437364337

   @sunchao also need to merge to branch-3.4 ...


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

[GitHub] [spark] pan3793 commented on pull request #40091: [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160

Posted by "pan3793 (via GitHub)" <gi...@apache.org>.
pan3793 commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1436411798

   > banch-3.3/3.2 use parquet 1.12.2, if this fix is accepted, would you mind submitting pr for these two branches? @pan3793
   
   sure.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

Re: [PR] [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160 [spark]

Posted by "yujhe (via GitHub)" <gi...@apache.org>.
yujhe commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1756863804

   Sure, here is the thread stack.
   
   ```text
   "Executor task launch worker for task 0.0 in stage 1.0 (TID 1)@7377" daemon prio=5 tid=0x36 nid=NA runnable
     java.lang.Thread.State: RUNNABLE
   	  at org.apache.parquet.hadoop.ParquetFileReader.<init>(ParquetFileReader.java:770)
   	  at org.apache.parquet.hadoop.ParquetFileReader.open(ParquetFileReader.java:657)
   	  at org.apache.parquet.hadoop.ParquetRecordReader.initializeInternalReader(ParquetRecordReader.java:162)
   	  at org.apache.parquet.hadoop.ParquetRecordReader.initialize(ParquetRecordReader.java:140)
   	  at org.apache.spark.sql.execution.datasources.parquet.ParquetFileFormat.$anonfun$buildReaderWithPartitionValues$2(ParquetFileFormat.scala:405)
   	  at org.apache.spark.sql.execution.datasources.parquet.ParquetFileFormat$$Lambda$2412.203214861.apply(Unknown Source:-1)
   	  at org.apache.spark.sql.execution.datasources.FileScanRDD$$anon$1.org$apache$spark$sql$execution$datasources$FileScanRDD$$anon$$readCurrentFile(FileScanRDD.scala:209)
   	  at org.apache.spark.sql.execution.datasources.FileScanRDD$$anon$1.nextIterator(FileScanRDD.scala:270)
   	  at org.apache.spark.sql.execution.datasources.FileScanRDD$$anon$1.hasNext(FileScanRDD.scala:116)
   	  at scala.collection.Iterator$$anon$9.hasNext(Iterator.scala:583)
   	  at org.apache.spark.sql.execution.datasources.FileFormatDataWriter.writeWithIterator(FileFormatDataWriter.scala:91)
   	  at org.apache.spark.sql.execution.datasources.FileFormatWriter$.$anonfun$executeTask$1(FileFormatWriter.scala:341)
   	  at org.apache.spark.sql.execution.datasources.FileFormatWriter$$$Lambda$2521.582578288.apply(Unknown Source:-1)
   	  at org.apache.spark.util.Utils$.tryWithSafeFinallyAndFailureCallbacks(Utils.scala:1538)
   	  at org.apache.spark.sql.execution.datasources.FileFormatWriter$.executeTask(FileFormatWriter.scala:348)
   	  at org.apache.spark.sql.execution.datasources.FileFormatWriter$.$anonfun$write$22(FileFormatWriter.scala:266)
   	  at org.apache.spark.sql.execution.datasources.FileFormatWriter$$$Lambda$2415.242150313.apply(Unknown Source:-1)
   	  at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:90)
   	  at org.apache.spark.scheduler.Task.run(Task.scala:136)
   	  at org.apache.spark.executor.Executor$TaskRunner.$anonfun$run$3(Executor.scala:548)
   	  at org.apache.spark.executor.Executor$TaskRunner$$Lambda$1345.354344687.apply(Unknown Source:-1)
   	  at org.apache.spark.util.Utils$.tryWithSafeFinally(Utils.scala:1504)
   	  at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:551)
   	  at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
   	  at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
   	  at java.lang.Thread.run(Thread.java:748)
   
   ```


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

[GitHub] [spark] sunchao commented on pull request #40091: [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160

Posted by "sunchao (via GitHub)" <gi...@apache.org>.
sunchao commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1437366108

   Yes, it's in branch-3.4 as well 


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

Re: [PR] [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160 [spark]

Posted by "LuciferYang (via GitHub)" <gi...@apache.org>.
LuciferYang commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1786558496

   > @yujhe Oops, I totally forgot this one...
   > 
   > Your analysis makes sense, I took a look at the parquet non-vectorized reading code path, injecting such a workaround does not clean as we did in this PR. As this is not a regression, and the new Spark version adopts new Parquet version which already fixed this issue, I prefer to keep things as-is in branch-3.3.
   
   +1, It seems there's no elegant way to fix the path where the native parquet-mr initializes the reader. 
   
   
   
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

[GitHub] [spark] pan3793 commented on a diff in pull request #40091: [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160

Posted by "pan3793 (via GitHub)" <gi...@apache.org>.
pan3793 commented on code in PR #40091:
URL: https://github.com/apache/spark/pull/40091#discussion_r1111598380


##########
sql/core/src/main/java/org/apache/spark/sql/execution/datasources/parquet/ParquetCodecFactory.java:
##########
@@ -0,0 +1,110 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.sql.execution.datasources.parquet;
+
+import java.io.IOException;
+import java.io.InputStream;
+import java.nio.ByteBuffer;
+
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.io.compress.CodecPool;
+import org.apache.hadoop.io.compress.CompressionCodec;
+import org.apache.hadoop.io.compress.Decompressor;
+import org.apache.parquet.bytes.BytesInput;
+import org.apache.parquet.hadoop.CodecFactory;
+import org.apache.parquet.hadoop.codec.ZstandardCodec;
+import org.apache.parquet.hadoop.metadata.CompressionCodecName;
+
+/**
+ * This class implements a codec factory that is used when reading from Parquet. It adds a
+ * workaround for memory issues encountered when reading from zstd-compressed files. For
+ * details, see <a href="https://issues.apache.org/jira/browse/PARQUET-2160">PARQUET-2160</a>
+ */
+public class ParquetCodecFactory extends CodecFactory {

Review Comment:
   updated



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

[GitHub] [spark] pan3793 commented on pull request #40091: [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160

Posted by "pan3793 (via GitHub)" <gi...@apache.org>.
pan3793 commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1436374948

   Based on https://github.com/apache/parquet-mr/pull/982#issuecomment-1376750703, I guess that the Parquet community may think it's not a critical issue, but in my case, it's critical.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

[GitHub] [spark] dongjoon-hyun commented on pull request #40091: [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160

Posted by "dongjoon-hyun (via GitHub)" <gi...@apache.org>.
dongjoon-hyun commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1437353508

   Thank you all. And, feel free to merge to land this to the release branches, @sunchao .
   
   cc @cloud-fan , @HyukjinKwon , @mridulm , @tgravescs 


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

[GitHub] [spark] sunchao closed pull request #40091: [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160

Posted by "sunchao (via GitHub)" <gi...@apache.org>.
sunchao closed pull request #40091: [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160
URL: https://github.com/apache/spark/pull/40091


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

[GitHub] [spark] LuciferYang commented on a diff in pull request #40091: [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160

Posted by "LuciferYang (via GitHub)" <gi...@apache.org>.
LuciferYang commented on code in PR #40091:
URL: https://github.com/apache/spark/pull/40091#discussion_r1111576062


##########
sql/core/src/main/java/org/apache/spark/sql/execution/datasources/parquet/ParquetCodecFactory.java:
##########
@@ -0,0 +1,110 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.sql.execution.datasources.parquet;
+
+import java.io.IOException;
+import java.io.InputStream;
+import java.nio.ByteBuffer;
+
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.io.compress.CodecPool;
+import org.apache.hadoop.io.compress.CompressionCodec;
+import org.apache.hadoop.io.compress.Decompressor;
+import org.apache.parquet.bytes.BytesInput;
+import org.apache.parquet.hadoop.CodecFactory;
+import org.apache.parquet.hadoop.codec.ZstandardCodec;
+import org.apache.parquet.hadoop.metadata.CompressionCodecName;
+
+/**
+ * This class implements a codec factory that is used when reading from Parquet. It adds a
+ * workaround for memory issues encountered when reading from zstd-compressed files. For
+ * details, see <a href="https://issues.apache.org/jira/browse/PARQUET-2160">PARQUET-2160</a>
+ */
+public class ParquetCodecFactory extends CodecFactory {

Review Comment:
   I think we should add a `TODO` to remind us to revert this file when updating Parquet version
   
   



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

[GitHub] [spark] LuciferYang commented on a diff in pull request #40091: [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160

Posted by "LuciferYang (via GitHub)" <gi...@apache.org>.
LuciferYang commented on code in PR #40091:
URL: https://github.com/apache/spark/pull/40091#discussion_r1111602428


##########
sql/core/src/main/java/org/apache/spark/sql/execution/datasources/parquet/ParquetCodecFactory.java:
##########
@@ -34,6 +34,8 @@
  * This class implements a codec factory that is used when reading from Parquet. It adds a
  * workaround for memory issues encountered when reading from zstd-compressed files. For
  * details, see <a href="https://issues.apache.org/jira/browse/PARQUET-2160">PARQUET-2160</a>
+ *
+ * TODO: Remove this workaround after upgrading Parquet which include PARQUET-2160.

Review Comment:
   should be `TODO(SPARK-xxx):`



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

Re: [PR] [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160 [spark]

Posted by "yujhe (via GitHub)" <gi...@apache.org>.
yujhe commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1786538476

   We found that this happens if we are reading Parquet file with nested columns in schema.
   
   ```scala
   val path = "/tmp/parquet_zstd"
   (1 to 100).map(i => (i, Seq(i)))
     .toDF("id", "value")
     .repartition(1)
     .write
     .mode("overwrite")
     .parquet(path)
   
   val df = spark.read.parquet(path)
   df.write.mode("overwrite").parquet("/tmp/dummy")
   ```
   
   After tracing the code, [ParquetCodecFactory](https://github.com/apache/spark/blob/branch-3.3/sql/core/src/main/java/org/apache/spark/sql/execution/datasources/parquet/ParquetCodecFactory.java#L40) only applies to [VectorizedParquetRecordReader](https://github.com/apache/spark/blob/branch-3.3/sql/core/src/main/java/org/apache/spark/sql/execution/datasources/parquet/VectorizedParquetRecordReader.java#L171) with [vectorized reader enabled](https://github.com/apache/spark/blob/branch-3.3/sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/parquet/ParquetFileFormat.scala#L355). 
   
   However, for schema with nested columns, the vectorized reader is disabled by default in Spark 3.3 (`spark.sql.parquet.enableNestedColumnVectorizedReader=false`). Therefore, this workaround does not work in this case. 


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

Re: [PR] [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160 [spark]

Posted by "yujhe (via GitHub)" <gi...@apache.org>.
yujhe commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1786583003

   Totally understand, we will try to apply patch [PARQUET-2160](https://issues.apache.org/jira/browse/PARQUET-2160) to parquet-hadoop-1.12.2 to our cluster.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org

Re: [PR] [SPARK-41952][SQL] Fix Parquet zstd off-heap memory leak as a workaround for PARQUET-2160 [spark]

Posted by "yujhe (via GitHub)" <gi...@apache.org>.
yujhe commented on PR #40091:
URL: https://github.com/apache/spark/pull/40091#issuecomment-1756720258

   @pan3793 We found that in some cases this workaround did not work when reading Parquet files in zstd codec and outputting it to a new path.
   
   ```scala
   val df = spark.read.parquet("zstd_parquet_file")
   df.repartition(3).write.mode("overwrite").parquet("/tmp/test")
   ```
   
   ![Screenshot 2023-10-11 at 10 48 19 AM](https://github.com/apache/spark/assets/6325926/577ecfe1-e9c0-41ea-ab63-bddb2142a3be)
   
   In debug mode, you can see it won't set ParquetCodecFactory when initializing `org.apache.parquet.hadoop.ParquetFileReader` which is called from [ParquetFileFormat.buildReaderWithPartitionValues](https://github.com/apache/spark/blob/branch-3.3/sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/parquet/ParquetFileFormat.scala#L405C32-L405C32).
   
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org