You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@spark.apache.org by "viirya (via GitHub)" <gi...@apache.org> on 2025/10/27 22:08:59 UTC
[PR] Release buffers [spark]
viirya opened a new pull request, #52747:
URL: https://github.com/apache/spark/pull/52747
<!--
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
'common/utils/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.
-->
### 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.
-->
### Does this PR introduce _any_ user-facing change?
<!--
Note that it means *any* user-facing change including all aspects such as new features, bug fixes, or other behavior changes. Documentation-only updates are not considered user-facing changes.
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'.
-->
### 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.
-->
### Was this patch authored or co-authored using generative AI tooling?
<!--
If generative AI tooling has been used in the process of authoring this patch, please include the
phrase: 'Generated-by: ' followed by the name of the tool and its version.
If no, write 'No'.
Please refer to the [ASF Generative Tooling Guidance](https://www.apache.org/legal/generative-tooling.html) for details.
-->
--
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-54134][SQL] Optimize Arrow memory usage [spark]
Posted by "viirya (via GitHub)" <gi...@apache.org>.
viirya commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2492482097
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -92,8 +95,25 @@ private[sql] object ArrowConverters extends Logging {
ArrowUtils.rootAllocator.newChildAllocator(
s"to${this.getClass.getSimpleName}", 0, Long.MaxValue)
- private val root = VectorSchemaRoot.create(arrowSchema, allocator)
- protected val unloader = new VectorUnloader(root)
+ protected val root = VectorSchemaRoot.create(arrowSchema, allocator)
+
+ // Create compression codec based on config
+ private val compressionCodecName = SQLConf.get.arrowCompressionCodec
+ private val codec = compressionCodecName match {
+ case "none" => NoCompressionCodec.INSTANCE
+ case "zstd" =>
+ val factory = CompressionCodec.Factory.INSTANCE
+ val codecType = new ZstdCompressionCodec().getCodecType()
+ factory.createCodec(codecType)
+ case "lz4" =>
+ val factory = CompressionCodec.Factory.INSTANCE
+ val codecType = new Lz4CompressionCodec().getCodecType()
+ factory.createCodec(codecType)
+ case other =>
+ throw new IllegalArgumentException(
Review Comment:
Yea, that would be better.
--
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] Release buffers [spark]
Posted by "viirya (via GitHub)" <gi...@apache.org>.
viirya commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2475219755
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowWriterWrapper.scala:
##########
@@ -60,6 +60,13 @@ case class ArrowWriterWrapper(
allocator = null
}
}
+
+ /**
+ * Clear the underlying VectorSchemaRoot to release the memory held by the vectors.
Review Comment:
`clear` and `close` are same on `FieldVector`. `VectorSchemaRoot.clear` additionally resets its `rowCount`, so yes, `VectorSchemaRoot` can be reused after `clear`.
```
/** Alternative to clear(). Allows use as an AutoCloseable in try-with-resources. */
@Override
void close();
```
--
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-xxxxx][SQL] Optimize Arrow memory usage [spark]
Posted by "viirya (via GitHub)" <gi...@apache.org>.
viirya commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2484035269
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -92,9 +94,14 @@ private[sql] object ArrowConverters extends Logging {
ArrowUtils.rootAllocator.newChildAllocator(
s"to${this.getClass.getSimpleName}", 0, Long.MaxValue)
- private val root = VectorSchemaRoot.create(arrowSchema, allocator)
- protected val unloader = new VectorUnloader(root)
- protected val arrowWriter = ArrowWriter.create(root)
+ protected val root = VectorSchemaRoot.create(arrowSchema, allocator)
+
+ val factory = CompressionCodec.Factory.INSTANCE
+ val codecType = new ZstdCompressionCodec().getCodecType()
Review Comment:
I will make this configurable.
--
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-54134][SQL] Optimize Arrow memory usage [spark]
Posted by "zhengruifeng (via GitHub)" <gi...@apache.org>.
zhengruifeng commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2493110731
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/internal/SQLConf.scala:
##########
@@ -3958,6 +3958,20 @@ object SQLConf {
"than zero and less than INT_MAX.")
.createWithDefaultString("64MB")
+ val ARROW_EXECUTION_COMPRESSION_CODEC =
+ buildConf("spark.sql.execution.arrow.compressionCodec")
+ .doc("Compression codec used to compress Arrow IPC data when transferring data " +
Review Comment:
does this optimization take effect in pandas udf?
--
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-54134][SQL] Optimize Arrow memory usage [spark]
Posted by "viirya (via GitHub)" <gi...@apache.org>.
viirya commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2493207797
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/internal/SQLConf.scala:
##########
@@ -3958,6 +3958,20 @@ object SQLConf {
"than zero and less than INT_MAX.")
.createWithDefaultString("64MB")
+ val ARROW_EXECUTION_COMPRESSION_CODEC =
+ buildConf("spark.sql.execution.arrow.compressionCodec")
+ .doc("Compression codec used to compress Arrow IPC data when transferring data " +
Review Comment:
I think no, it is currently applied on `toArrow` and `toPandas` which is on the reported issue. It should be also available to arrow udf and pandas udf. I will try to extend this to such cases.
--
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-xxxxx][SQL] Release Arrow buffers early [spark]
Posted by "cloud-fan (via GitHub)" <gi...@apache.org>.
cloud-fan commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2480723224
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -110,6 +110,8 @@ private[sql] object ArrowConverters extends Logging {
val writeChannel = new WriteChannel(Channels.newChannel(out))
Utils.tryWithSafeFinally {
+ arrowWriter = ArrowWriter.create(root)
Review Comment:
then why do we need to create new writers if buffers are already released by `root.clear()`?
--
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-xxxxx][SQL] Release Arrow buffers early [spark]
Posted by "viirya (via GitHub)" <gi...@apache.org>.
viirya commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2482708197
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -110,6 +110,8 @@ private[sql] object ArrowConverters extends Logging {
val writeChannel = new WriteChannel(Channels.newChannel(out))
Utils.tryWithSafeFinally {
+ arrowWriter = ArrowWriter.create(root)
Review Comment:
For next batch, we create a new writer which will allocate new buffers to hold next batch.
--
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-54134][SQL] Optimize Arrow memory usage [spark]
Posted by "viirya (via GitHub)" <gi...@apache.org>.
viirya commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2493207797
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/internal/SQLConf.scala:
##########
@@ -3958,6 +3958,20 @@ object SQLConf {
"than zero and less than INT_MAX.")
.createWithDefaultString("64MB")
+ val ARROW_EXECUTION_COMPRESSION_CODEC =
+ buildConf("spark.sql.execution.arrow.compressionCodec")
+ .doc("Compression codec used to compress Arrow IPC data when transferring data " +
Review Comment:
I think no, it is currently applied on `toArrow` and `toPandas` which is on the reported issue. It should be also available to arrow udf and pandas udf. I will take a look.
--
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-54134][SQL] Optimize Arrow memory usage [spark]
Posted by "allisonwang-db (via GitHub)" <gi...@apache.org>.
allisonwang-db commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2492418907
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -92,8 +95,25 @@ private[sql] object ArrowConverters extends Logging {
ArrowUtils.rootAllocator.newChildAllocator(
s"to${this.getClass.getSimpleName}", 0, Long.MaxValue)
- private val root = VectorSchemaRoot.create(arrowSchema, allocator)
- protected val unloader = new VectorUnloader(root)
+ protected val root = VectorSchemaRoot.create(arrowSchema, allocator)
+
+ // Create compression codec based on config
+ private val compressionCodecName = SQLConf.get.arrowCompressionCodec
+ private val codec = compressionCodecName match {
+ case "none" => NoCompressionCodec.INSTANCE
+ case "zstd" =>
+ val factory = CompressionCodec.Factory.INSTANCE
+ val codecType = new ZstdCompressionCodec().getCodecType()
+ factory.createCodec(codecType)
+ case "lz4" =>
+ val factory = CompressionCodec.Factory.INSTANCE
+ val codecType = new Lz4CompressionCodec().getCodecType()
+ factory.createCodec(codecType)
+ case other =>
+ throw new IllegalArgumentException(
Review Comment:
Should be `SparkException.internalError`
--
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-xxxxx][SQL] Release Arrow buffers early [spark]
Posted by "cloud-fan (via GitHub)" <gi...@apache.org>.
cloud-fan commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2476439503
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -110,6 +110,8 @@ private[sql] object ArrowConverters extends Logging {
val writeChannel = new WriteChannel(Channels.newChannel(out))
Utils.tryWithSafeFinally {
+ arrowWriter = ArrowWriter.create(root)
Review Comment:
is there any arrow doc mentioning this "non-reusable" thing? This is a bit supring to be, as it provides a `reset` method.
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -110,6 +110,8 @@ private[sql] object ArrowConverters extends Logging {
val writeChannel = new WriteChannel(Channels.newChannel(out))
Utils.tryWithSafeFinally {
+ arrowWriter = ArrowWriter.create(root)
Review Comment:
is there any arrow doc mentioning this "non-reusable" thing? This is a bit supring to me, as it provides a `reset` method.
--
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-xxxxx][SQL] Release Arrow buffers early [spark]
Posted by "cloud-fan (via GitHub)" <gi...@apache.org>.
cloud-fan commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2477654654
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -110,6 +110,8 @@ private[sql] object ArrowConverters extends Logging {
val writeChannel = new WriteChannel(Channels.newChannel(out))
Utils.tryWithSafeFinally {
+ arrowWriter = ArrowWriter.create(root)
Review Comment:
how do we release it earlier? We didn't null out the `arrowWriter` anywhere.
--
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-xxxxx][SQL] Release Arrow buffers early [spark]
Posted by "viirya (via GitHub)" <gi...@apache.org>.
viirya commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2476625775
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -110,6 +110,8 @@ private[sql] object ArrowConverters extends Logging {
val writeChannel = new WriteChannel(Channels.newChannel(out))
Utils.tryWithSafeFinally {
+ arrowWriter = ArrowWriter.create(root)
Review Comment:
They "can" be reused. This change intentionally doesn't reuse it to release the buffers early.
--
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-54134][SQL] Optimize Arrow memory usage [spark]
Posted by "viirya (via GitHub)" <gi...@apache.org>.
viirya commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2495982683
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -92,8 +95,25 @@ private[sql] object ArrowConverters extends Logging {
ArrowUtils.rootAllocator.newChildAllocator(
s"to${this.getClass.getSimpleName}", 0, Long.MaxValue)
- private val root = VectorSchemaRoot.create(arrowSchema, allocator)
- protected val unloader = new VectorUnloader(root)
+ protected val root = VectorSchemaRoot.create(arrowSchema, allocator)
+
+ // Create compression codec based on config
+ private val compressionCodecName = SQLConf.get.arrowCompressionCodec
+ private val codec = compressionCodecName match {
+ case "none" => NoCompressionCodec.INSTANCE
+ case "zstd" =>
+ val factory = CompressionCodec.Factory.INSTANCE
+ val codecType = new ZstdCompressionCodec().getCodecType()
+ factory.createCodec(codecType)
Review Comment:
Okay, we can add compression level option together.
--
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-54134][SQL] Optimize Arrow memory usage [spark]
Posted by "viirya (via GitHub)" <gi...@apache.org>.
viirya commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2501293945
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -92,8 +95,25 @@ private[sql] object ArrowConverters extends Logging {
ArrowUtils.rootAllocator.newChildAllocator(
s"to${this.getClass.getSimpleName}", 0, Long.MaxValue)
- private val root = VectorSchemaRoot.create(arrowSchema, allocator)
- protected val unloader = new VectorUnloader(root)
+ protected val root = VectorSchemaRoot.create(arrowSchema, allocator)
+
+ // Create compression codec based on config
+ private val compressionCodecName = SQLConf.get.arrowCompressionCodec
+ private val codec = compressionCodecName match {
+ case "none" => NoCompressionCodec.INSTANCE
+ case "zstd" =>
+ val factory = CompressionCodec.Factory.INSTANCE
+ val codecType = new ZstdCompressionCodec().getCodecType()
+ factory.createCodec(codecType)
Review Comment:
I am going to add the option in https://github.com/apache/spark/pull/52925 along with Pandas UDF support.
--
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-xxxxx][SQL] Release Arrow buffers early [spark]
Posted by "cloud-fan (via GitHub)" <gi...@apache.org>.
cloud-fan commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2476205234
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -110,6 +110,8 @@ private[sql] object ArrowConverters extends Logging {
val writeChannel = new WriteChannel(Channels.newChannel(out))
Utils.tryWithSafeFinally {
+ arrowWriter = ArrowWriter.create(root)
Review Comment:
so we always create a new writer when returning an iterator. is it because the writer is not re-usable?
--
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-54134][SQL] Optimize Arrow memory usage [spark]
Posted by "dongjoon-hyun (via GitHub)" <gi...@apache.org>.
dongjoon-hyun closed pull request #52747: [SPARK-54134][SQL] Optimize Arrow memory usage
URL: https://github.com/apache/spark/pull/52747
--
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-xxxxx][SQL] Release Arrow buffers early [spark]
Posted by "viirya (via GitHub)" <gi...@apache.org>.
viirya commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2478735084
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -110,6 +110,8 @@ private[sql] object ArrowConverters extends Logging {
val writeChannel = new WriteChannel(Channels.newChannel(out))
Utils.tryWithSafeFinally {
+ arrowWriter = ArrowWriter.create(root)
Review Comment:
The buffers are created under the given `root` when doing `ArrowWriter.create`. They are released by `root.clear()`.
--
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-54134][SQL] Optimize Arrow memory usage [spark]
Posted by "dbtsai (via GitHub)" <gi...@apache.org>.
dbtsai commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2495933738
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -92,8 +95,25 @@ private[sql] object ArrowConverters extends Logging {
ArrowUtils.rootAllocator.newChildAllocator(
s"to${this.getClass.getSimpleName}", 0, Long.MaxValue)
- private val root = VectorSchemaRoot.create(arrowSchema, allocator)
- protected val unloader = new VectorUnloader(root)
+ protected val root = VectorSchemaRoot.create(arrowSchema, allocator)
+
+ // Create compression codec based on config
+ private val compressionCodecName = SQLConf.get.arrowCompressionCodec
+ private val codec = compressionCodecName match {
+ case "none" => NoCompressionCodec.INSTANCE
+ case "zstd" =>
+ val factory = CompressionCodec.Factory.INSTANCE
+ val codecType = new ZstdCompressionCodec().getCodecType()
+ factory.createCodec(codecType)
Review Comment:
Would be great that we can have an option to add compression levels
--
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] Release buffers [spark]
Posted by "cloud-fan (via GitHub)" <gi...@apache.org>.
cloud-fan commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2471706604
##########
sql/catalyst/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowWriterWrapper.scala:
##########
@@ -60,6 +60,13 @@ case class ArrowWriterWrapper(
allocator = null
}
}
+
+ /**
+ * Clear the underlying VectorSchemaRoot to release the memory held by the vectors.
Review Comment:
what's the difference between `clear` and `close`? The wrier can be reused after cleared but not closed?
--
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-xxxxx][SQL] Release Arrow buffers early [spark]
Posted by "viirya (via GitHub)" <gi...@apache.org>.
viirya commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2476273726
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -110,6 +110,8 @@ private[sql] object ArrowConverters extends Logging {
val writeChannel = new WriteChannel(Channels.newChannel(out))
Utils.tryWithSafeFinally {
+ arrowWriter = ArrowWriter.create(root)
Review Comment:
Yea, this change made to not reuse the buffers (and the writers) to release the buffers early.
--
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-54134][SQL] Optimize Arrow memory usage [spark]
Posted by "viirya (via GitHub)" <gi...@apache.org>.
viirya commented on PR #52747:
URL: https://github.com/apache/spark/pull/52747#issuecomment-3483298336
Thank you @cloud-fan @dongjoon-hyun
--
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-54134][SQL] Optimize Arrow memory usage [spark]
Posted by "viirya (via GitHub)" <gi...@apache.org>.
viirya commented on code in PR #52747:
URL: https://github.com/apache/spark/pull/52747#discussion_r2493209059
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/arrow/ArrowConverters.scala:
##########
@@ -92,8 +95,25 @@ private[sql] object ArrowConverters extends Logging {
ArrowUtils.rootAllocator.newChildAllocator(
s"to${this.getClass.getSimpleName}", 0, Long.MaxValue)
- private val root = VectorSchemaRoot.create(arrowSchema, allocator)
- protected val unloader = new VectorUnloader(root)
+ protected val root = VectorSchemaRoot.create(arrowSchema, allocator)
+
+ // Create compression codec based on config
+ private val compressionCodecName = SQLConf.get.arrowCompressionCodec
+ private val codec = compressionCodecName match {
+ case "none" => NoCompressionCodec.INSTANCE
+ case "zstd" =>
+ val factory = CompressionCodec.Factory.INSTANCE
+ val codecType = new ZstdCompressionCodec().getCodecType()
+ factory.createCodec(codecType)
+ case "lz4" =>
+ val factory = CompressionCodec.Factory.INSTANCE
+ val codecType = new Lz4CompressionCodec().getCodecType()
+ factory.createCodec(codecType)
+ case other =>
+ throw new IllegalArgumentException(
Review Comment:
I will change to `SparkException` when I extend this to pandas udf.
--
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