[SPARK-34862][SQL] Support nested column in ORC vectorized reader #31958

c21 · 2021-03-25T09:11:39Z

What changes were proposed in this pull request?

This PR is to support nested column type in Spark ORC vectorized reader. Currently ORC vectorized reader does not support nested column type (struct, array and map). We implemented nested column vectorized reader for FB-ORC in our internal fork of Spark. We are seeing performance improvement compared to non-vectorized reader when reading nested columns. In addition, this can also help improve the non-nested column performance when reading non-nested and nested columns together in one query.

Before this PR:

OrcColumnVector is the implementation class for Spark's ColumnVector to wrap Hive's/ORC's ColumnVector to read AtomicType data.

After this PR:

OrcColumnVector is an abstract class to keep interface being shared between multiple implementation class of orc column vectors, namely OrcAtomicColumnVector (for AtomicType), OrcArrayColumnVector (for ArrayType), OrcMapColumnVector (for MapType), OrcStructColumnVector (for StructType). So the original logic to read AtomicType data is moved from OrcColumnVector to OrcAtomicColumnVector. The abstract class of OrcColumnVector is needed here because of supporting nested column (i.e. nested column vectors).
A utility method OrcColumnVectorUtils.toOrcColumnVector is added to create Spark's OrcColumnVector from Hive's/ORC's ColumnVector.
A new user-facing config spark.sql.orc.enableNestedColumnVectorizedReader is added to control enabling/disabling vectorized reader for nested columns. The default value is false (i.e. disabling by default). For certain tables having deep nested columns, vectorized reader might take too much memory for each sub-column vectors, compared to non-vectorized reader. So providing a config here to work around OOM for query reading wide and deep nested columns if any. We plan to enable it by default on 3.3. Leave it disable in 3.2 in case for any unknown bugs.

Why are the changes needed?

Improve query performance when reading nested columns from ORC file format.
Tested with locally adding a small benchmark in OrcReadBenchmark.scala. Seeing more than 1x run time improvement.

Running benchmark: SQL Nested Column Scan
  Running case: Native ORC MR
  Stopped after 2 iterations, 37850 ms
  Running case: Native ORC Vectorized (Enabled Nested Column)
  Stopped after 2 iterations, 15892 ms
  Running case: Native ORC Vectorized (Disabled Nested Column)
  Stopped after 2 iterations, 37954 ms
  Running case: Hive built-in ORC
  Stopped after 2 iterations, 35118 ms

Java HotSpot(TM) 64-Bit Server VM 1.8.0_181-b13 on Mac OS X 10.15.7
Intel(R) Core(TM) i9-9980HK CPU @ 2.40GHz
SQL Nested Column Scan:                         Best Time(ms)   Avg Time(ms)   Stdev(ms)    Rate(M/s)   Per Row(ns)   Relative
------------------------------------------------------------------------------------------------------------------------------
Native ORC MR                                           18706          18925         310          0.1       17839.6       1.0X
Native ORC Vectorized (Enabled Nested Column)            7625           7946         455          0.1        7271.6       2.5X
Native ORC Vectorized (Disabled Nested Column)          18415          18977         796          0.1       17561.5       1.0X
Hive built-in ORC                                       17469          17559         127          0.1       16660.1       1.1X

Benchmark:

nestedColumnScanBenchmark(1024 * 1024)
def nestedColumnScanBenchmark(values: Int): Unit = {
    val benchmark = new Benchmark(s"SQL Nested Column Scan", values, output = output)

    withTempPath { dir =>
      withTempTable("t1", "nativeOrcTable", "hiveOrcTable") {
        import spark.implicits._
        spark.range(values).map(_ => Random.nextLong).map { x =>
          val arrayOfStructColumn = (0 until 5).map(i => (x + i, s"$x" * 5))
          val mapOfStructColumn = Map(
            s"$x" -> (x * 0.1, (x, s"$x" * 100)),
            (s"$x" * 2) -> (x * 0.2, (x, s"$x" * 200)),
            (s"$x" * 3) -> (x * 0.3, (x, s"$x" * 300)))
          (arrayOfStructColumn, mapOfStructColumn)
        }.toDF("col1", "col2")
          .createOrReplaceTempView("t1")

        prepareTable(dir, spark.sql(s"SELECT * FROM t1"))

        benchmark.addCase("Native ORC MR") { _ =>
          withSQLConf(SQLConf.ORC_VECTORIZED_READER_ENABLED.key -> "false") {
            spark.sql("SELECT SUM(SIZE(col1)), SUM(SIZE(col2)) FROM nativeOrcTable").noop()
          }
        }

        benchmark.addCase("Native ORC Vectorized (Enabled Nested Column)") { _ =>
          spark.sql("SELECT SUM(SIZE(col1)), SUM(SIZE(col2)) FROM nativeOrcTable").noop()
        }

        benchmark.addCase("Native ORC Vectorized (Disabled Nested Column)") { _ =>
          withSQLConf(SQLConf.ORC_VECTORIZED_READER_NESTED_COLUMN_ENABLED.key -> "false") {
            spark.sql("SELECT SUM(SIZE(col1)), SUM(SIZE(col2)) FROM nativeOrcTable").noop()
          }
        }

        benchmark.addCase("Hive built-in ORC") { _ =>
          spark.sql("SELECT SUM(SIZE(col1)), SUM(SIZE(col2)) FROM hiveOrcTable").noop()
        }

        benchmark.run()
      }
    }
  }

Does this PR introduce any user-facing change?

No.

How was this patch tested?

Added one simple test in OrcSourceSuite.scala to verify correctness.
Definitely need more unit tests and add benchmark here, but I want to first collect feedback before crafting more tests.

c21 · 2021-03-25T09:13:49Z

cc @cloud-fan, @maropu and @dongjoon-hyun could you help take a look when you have time, thanks.

SparkQA · 2021-03-25T09:19:07Z

Test build #136514 has finished for PR 31958 at commit 7037893.

This patch fails RAT tests.
This patch merges cleanly.
This patch adds the following public classes (experimental):
public class OrcArrayColumnVector extends OrcColumnVector
public class OrcAtomicColumnVector extends OrcColumnVector
public abstract class OrcColumnVector extends org.apache.spark.sql.vectorized.ColumnVector
class OrcColumnVectorUtils
public class OrcMapColumnVector extends OrcColumnVector
public class OrcStructColumnVector extends OrcColumnVector

SparkQA · 2021-03-25T12:11:38Z

Kubernetes integration test starting
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41099/

SparkQA · 2021-03-25T12:19:37Z

Kubernetes integration test status failure
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41099/

dongjoon-hyun

Thank you for working on this, @c21 .

c21 · 2021-03-25T23:27:45Z

Thank you @dongjoon-hyun and look forward to getting your feedback on this.
In the meanwhile I am checking the binary compatibility failures in unit test.

HyukjinKwon · 2021-03-26T01:08:32Z

cc @viirya too

SparkQA · 2021-03-26T01:42:17Z

Test build #136538 has finished for PR 31958 at commit 94df62c.

This patch fails MiMa tests.
This patch merges cleanly.
This patch adds no public classes.

SparkQA · 2021-03-26T02:18:33Z

Kubernetes integration test starting
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41122/

SparkQA · 2021-03-26T02:53:29Z

Kubernetes integration test status failure
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41122/

c21 · 2021-03-26T05:49:44Z

The unit test failed with MiMa tests

abstract method getBoolean(Int)Boolean in class org.apache.spark.sql.vectorized.ColumnVector does not have a correspondent in current version

However in this PR, the class org.apache.spark.sql.vectorized.ColumnVector is not changed at all. I am still checking why this test is failing.

SparkQA · 2021-03-26T22:36:36Z

Test build #136577 has finished for PR 31958 at commit d669815.

This patch fails MiMa tests.
This patch merges cleanly.
This patch adds no public classes.

SparkQA · 2021-03-26T23:07:45Z

Kubernetes integration test starting
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41161/

SparkQA · 2021-03-26T23:14:50Z

Kubernetes integration test status failure
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41161/

SparkQA · 2021-03-27T00:07:30Z

Kubernetes integration test starting
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41162/

SparkQA · 2021-03-27T00:37:12Z

Kubernetes integration test status failure
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41162/

SparkQA · 2021-03-27T01:13:34Z

Test build #136578 has finished for PR 31958 at commit 5e4eb0f.

This patch fails Spark unit tests.
This patch merges cleanly.
This patch adds no public classes.

SparkQA · 2021-03-27T02:07:57Z

Kubernetes integration test starting
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41164/

SparkQA · 2021-03-27T02:55:33Z

Kubernetes integration test status failure
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41164/

SparkQA · 2021-03-27T03:10:16Z

Test build #136580 has finished for PR 31958 at commit 7cb533c.

This patch fails Spark unit tests.
This patch merges cleanly.
This patch adds no public classes.

….scala

SparkQA · 2021-03-27T09:23:05Z

Kubernetes integration test starting
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41171/

SparkQA · 2021-03-27T10:17:54Z

Kubernetes integration test status failure
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41171/

dongjoon-hyun · 2021-03-28T19:06:53Z

sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/orc/OrcFileFormat.scala

-      schema.forall(_.dataType.isInstanceOf[AtomicType])
+      schema.forall(s => supportDataType(s.dataType) &&
+        !s.dataType.isInstanceOf[UserDefinedType[_]]) &&
+      supportBatchForNestedColumn(sparkSession, schema)


Can we do the same thing for Parquet, @c21 ?

@dongjoon-hyun - do you mean implementing Parquet vectorized reader for nested column? I created https://issues.apache.org/jira/browse/SPARK-34863 and plan to do it after this one, thanks.

Yes, thank you for creating SPARK-34863.

dongjoon-hyun · 2021-03-28T19:26:03Z

Also, cc @viirya since this is related to the nested columns.

SparkQA · 2021-03-29T05:34:51Z

Test build #136635 has started for PR 31958 at commit 9cd3bc5.

c21 · 2021-03-30T05:13:37Z

@cloud-fan and @viirya could you help take a look? Thanks.

cloud-fan · 2021-03-30T12:39:51Z

sql/core/src/main/java/org/apache/spark/sql/execution/datasources/orc/OrcArrayColumnVector.java

+  private final long[] lengths;
+
+  OrcArrayColumnVector(
+    DataType type,


nit: 4 spaces indentation for parameters.

@cloud-fan - updated.

cloud-fan · 2021-03-30T12:43:30Z

project/MimaExcludes.scala

-    ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.sql.streaming.SourceProgress.this")
+    ProblemFilters.exclude[DirectMissingMethodProblem]("org.apache.spark.sql.streaming.SourceProgress.this"),
+
+    // [SPARK-34862][SQL] Support nested column in ORC vectorized reader


This is weird, where do we change org.apache.spark.sql.vectorized.ColumnVector in this PR?

@cloud-fan - yeah it's weird. We don't change ColumnVector class at all. Do you have any idea for how to debug on this? I am still checking why, thanks.

maybe it's some bugs in Mima, not a bit deal as we know this PR doesn't break binary compatibility.

@cloud-fan - spent some time checking, but still not sure where the issue is, so I agree with you that might be some bug in Mima.

cloud-fan · 2021-03-30T13:08:51Z

sql/core/src/main/java/org/apache/spark/sql/execution/datasources/orc/OrcMapColumnVector.java

+  private final long[] lengths;
+
+  OrcMapColumnVector(
+    DataType type,


ditto, indentation

@cloud-fan - updated.

cloud-fan · 2021-03-30T13:16:59Z

For orc files without nested schema, do we observe perf regression after this PR?

c21 · 2021-03-31T10:50:57Z

For orc files without nested schema, do we observe perf regression after this PR?

@cloud-fan - in theory here the difference for non-nested schema:

Before this PR:

OrcColumnVector -> org.apache.spark.sql.vectorized.ColumnVector

After this PR:

OrcAtomicColumnVector -> OrcColumnVector -> org.apache.spark.sql.vectorized.ColumnVector

The only overhead introduced here is one more layer in classes, and might be some overhead for virtual function call and class loading.

Tested on same type of machine in AWS EC2 for all ORC reader benchmarks: OrcReadBenchmark-results.txt and DataSourceReadBenchmark-results.txt (java 8 here). Do not see regression compared ORC vectorized reader and other ORC readers.

Results:

OrcReadBenchmark-results.txt
DataSourceReadBenchmark-results.txt

Machine:

Amazon AWS EC2
type: r3.xlarge
region: us-west-2 (Oregon)
OS: Linux

SparkQA · 2021-03-31T11:41:22Z

Kubernetes integration test starting
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41345/

SparkQA · 2021-03-31T12:17:34Z

Kubernetes integration test status failure
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41345/

SparkQA · 2021-03-31T13:47:37Z

Test build #136764 has finished for PR 31958 at commit fda6b12.

This patch passes all tests.
This patch merges cleanly.
This patch adds no public classes.

viirya · 2021-03-31T19:56:11Z

sql/core/src/main/java/org/apache/spark/sql/execution/datasources/orc/OrcArrayColumnVector.java

+import org.apache.hadoop.hive.ql.exec.vector.ColumnVector;
+import org.apache.spark.sql.types.ArrayType;


We have a blank line between third party import and Spark's import.

@viirya - yes, I missed somehow, thanks for the careful review, updated.

viirya

One comment about style, otherwise lgtm

SparkQA · 2021-03-31T21:39:07Z

Kubernetes integration test starting
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41364/

SparkQA · 2021-03-31T21:47:09Z

Kubernetes integration test status failure
URL: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder-K8s/41364/

SparkQA · 2021-03-31T23:03:44Z

Test build #136781 has finished for PR 31958 at commit 44feacc.

This patch passes all tests.
This patch merges cleanly.
This patch adds no public classes.

viirya · 2021-04-02T06:09:47Z

Thanks. Merging to master.

c21 · 2021-04-02T06:13:53Z

Thank you @viirya, @cloud-fan and @dongjoon-hyun for review!

dongjoon-hyun · 2021-04-02T17:56:53Z

Thank you, @c21 and all!

### What changes were proposed in this pull request? This PR is to support nested column type in Spark ORC vectorized reader. Currently ORC vectorized reader [does not support nested column type (struct, array and map)](https://github.com/apache/spark/blob/master/sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/orc/OrcFileFormat.scala#L138). We implemented nested column vectorized reader for FB-ORC in our internal fork of Spark. We are seeing performance improvement compared to non-vectorized reader when reading nested columns. In addition, this can also help improve the non-nested column performance when reading non-nested and nested columns together in one query. Before this PR: * `OrcColumnVector` is the implementation class for Spark's `ColumnVector` to wrap Hive's/ORC's `ColumnVector` to read `AtomicType` data. After this PR: * `OrcColumnVector` is an abstract class to keep interface being shared between multiple implementation class of orc column vectors, namely `OrcAtomicColumnVector` (for `AtomicType`), `OrcArrayColumnVector` (for `ArrayType`), `OrcMapColumnVector` (for `MapType`), `OrcStructColumnVector` (for `StructType`). So the original logic to read `AtomicType` data is moved from `OrcColumnVector` to `OrcAtomicColumnVector`. The abstract class of `OrcColumnVector` is needed here because of supporting nested column (i.e. nested column vectors). * A utility method `OrcColumnVectorUtils.toOrcColumnVector` is added to create Spark's `OrcColumnVector` from Hive's/ORC's `ColumnVector`. * A new user-facing config `spark.sql.orc.enableNestedColumnVectorizedReader` is added to control enabling/disabling vectorized reader for nested columns. The default value is false (i.e. disabling by default). For certain tables having deep nested columns, vectorized reader might take too much memory for each sub-column vectors, compared to non-vectorized reader. So providing a config here to work around OOM for query reading wide and deep nested columns if any. We plan to enable it by default on 3.3. Leave it disable in 3.2 in case for any unknown bugs. ### Why are the changes needed? Improve query performance when reading nested columns from ORC file format. Tested with locally adding a small benchmark in `OrcReadBenchmark.scala`. Seeing more than 1x run time improvement. ``` Running benchmark: SQL Nested Column Scan Running case: Native ORC MR Stopped after 2 iterations, 37850 ms Running case: Native ORC Vectorized (Enabled Nested Column) Stopped after 2 iterations, 15892 ms Running case: Native ORC Vectorized (Disabled Nested Column) Stopped after 2 iterations, 37954 ms Running case: Hive built-in ORC Stopped after 2 iterations, 35118 ms Java HotSpot(TM) 64-Bit Server VM 1.8.0_181-b13 on Mac OS X 10.15.7 Intel(R) Core(TM) i9-9980HK CPU 2.40GHz SQL Nested Column Scan: Best Time(ms) Avg Time(ms) Stdev(ms) Rate(M/s) Per Row(ns) Relative ------------------------------------------------------------------------------------------------------------------------------ Native ORC MR 18706 18925 310 0.1 17839.6 1.0X Native ORC Vectorized (Enabled Nested Column) 7625 7946 455 0.1 7271.6 2.5X Native ORC Vectorized (Disabled Nested Column) 18415 18977 796 0.1 17561.5 1.0X Hive built-in ORC 17469 17559 127 0.1 16660.1 1.1X ``` Benchmark: ``` nestedColumnScanBenchmark(1024 * 1024) def nestedColumnScanBenchmark(values: Int): Unit = { val benchmark = new Benchmark(s"SQL Nested Column Scan", values, output = output) withTempPath { dir => withTempTable("t1", "nativeOrcTable", "hiveOrcTable") { import spark.implicits._ spark.range(values).map(_ => Random.nextLong).map { x => val arrayOfStructColumn = (0 until 5).map(i => (x + i, s"$x" * 5)) val mapOfStructColumn = Map( s"$x" -> (x * 0.1, (x, s"$x" * 100)), (s"$x" * 2) -> (x * 0.2, (x, s"$x" * 200)), (s"$x" * 3) -> (x * 0.3, (x, s"$x" * 300))) (arrayOfStructColumn, mapOfStructColumn) }.toDF("col1", "col2") .createOrReplaceTempView("t1") prepareTable(dir, spark.sql(s"SELECT * FROM t1")) benchmark.addCase("Native ORC MR") { _ => withSQLConf(SQLConf.ORC_VECTORIZED_READER_ENABLED.key -> "false") { spark.sql("SELECT SUM(SIZE(col1)), SUM(SIZE(col2)) FROM nativeOrcTable").noop() } } benchmark.addCase("Native ORC Vectorized (Enabled Nested Column)") { _ => spark.sql("SELECT SUM(SIZE(col1)), SUM(SIZE(col2)) FROM nativeOrcTable").noop() } benchmark.addCase("Native ORC Vectorized (Disabled Nested Column)") { _ => withSQLConf(SQLConf.ORC_VECTORIZED_READER_NESTED_COLUMN_ENABLED.key -> "false") { spark.sql("SELECT SUM(SIZE(col1)), SUM(SIZE(col2)) FROM nativeOrcTable").noop() } } benchmark.addCase("Hive built-in ORC") { _ => spark.sql("SELECT SUM(SIZE(col1)), SUM(SIZE(col2)) FROM hiveOrcTable").noop() } benchmark.run() } } } ``` ### Does this PR introduce _any_ user-facing change? No. ### How was this patch tested? Added one simple test in `OrcSourceSuite.scala` to verify correctness. Definitely need more unit tests and add benchmark here, but I want to first collect feedback before crafting more tests. Closes apache#31958 from c21/orc-vector. Authored-by: Cheng Su <chengsu@fb.com> Signed-off-by: Liang-Chi Hsieh <viirya@gmail.com> (cherry picked from commit 1fc66f6)

github-actions bot added the SQL label Mar 25, 2021

dongjoon-hyun reviewed Mar 25, 2021

View reviewed changes

github-actions bot added the BUILD label Mar 26, 2021

c21 added 5 commits March 27, 2021 01:22

Support nested column in ORC vectorized reader

d1d6585

Add missing license header

fbc8c6c

Add DirectAbstractMethodProblem of ColumnVector class in MimaExcludes…

c911e89

….scala

Try to update MimaExcludes to fix MiMa test

3bfc03a

Do not allow UserDefinedType for vectorization to fix unit test failure

f30cc88

c21 force-pushed the orc-vector branch from 7cb533c to f30cc88 Compare March 27, 2021 08:23

dongjoon-hyun reviewed Mar 28, 2021

View reviewed changes

Address all comments

9cd3bc5

cloud-fan reviewed Mar 30, 2021

View reviewed changes

Address indentation comments

fda6b12

cloud-fan approved these changes Mar 31, 2021

View reviewed changes

viirya reviewed Mar 31, 2021

View reviewed changes

viirya approved these changes Mar 31, 2021

View reviewed changes

Address comment for style

44feacc

viirya closed this in 1fc66f6 Apr 2, 2021

c21 deleted the orc-vector branch April 2, 2021 06:14

c21 mentioned this pull request Aug 26, 2021

[SPARK-36594][SQL] ORC vectorized reader should properly check maximal number of fields #33842

Closed

sunchao mentioned this pull request Aug 31, 2021

[SPARK-34863][SQL] Support complex types for Parquet vectorized reader #33695

Closed

		import org.apache.hadoop.hive.ql.exec.vector.ColumnVector;
		import org.apache.spark.sql.types.ArrayType;

[SPARK-34862][SQL] Support nested column in ORC vectorized reader #31958

[SPARK-34862][SQL] Support nested column in ORC vectorized reader #31958

Conversation

c21 commented Mar 25, 2021 • edited Loading

What changes were proposed in this pull request?

Why are the changes needed?

Does this PR introduce any user-facing change?

How was this patch tested?

c21 commented Mar 25, 2021

SparkQA commented Mar 25, 2021

SparkQA commented Mar 25, 2021

SparkQA commented Mar 25, 2021

dongjoon-hyun left a comment

Choose a reason for hiding this comment

c21 commented Mar 25, 2021

HyukjinKwon commented Mar 26, 2021

SparkQA commented Mar 26, 2021

SparkQA commented Mar 26, 2021

SparkQA commented Mar 26, 2021

c21 commented Mar 26, 2021

SparkQA commented Mar 26, 2021

SparkQA commented Mar 26, 2021

SparkQA commented Mar 26, 2021

SparkQA commented Mar 27, 2021

SparkQA commented Mar 27, 2021

SparkQA commented Mar 27, 2021

SparkQA commented Mar 27, 2021

SparkQA commented Mar 27, 2021

SparkQA commented Mar 27, 2021

SparkQA commented Mar 27, 2021

SparkQA commented Mar 27, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

dongjoon-hyun commented Mar 28, 2021

SparkQA commented Mar 29, 2021

c21 commented Mar 30, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

cloud-fan Mar 30, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

cloud-fan commented Mar 30, 2021

c21 commented Mar 31, 2021

SparkQA commented Mar 31, 2021

SparkQA commented Mar 31, 2021

SparkQA commented Mar 31, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

viirya left a comment

Choose a reason for hiding this comment

SparkQA commented Mar 31, 2021

SparkQA commented Mar 31, 2021

SparkQA commented Mar 31, 2021

viirya commented Apr 2, 2021

c21 commented Apr 2, 2021

dongjoon-hyun commented Apr 2, 2021

c21 commented Mar 25, 2021 •

edited

Loading

cloud-fan Mar 30, 2021 •

edited

Loading