[FLINK-20738][table-planner-blink] Separate the implementation of batch group aggregate nodes #14562

godfreyhe · 2021-01-05T12:23:37Z

What is the purpose of the change

Separate the implementation of batch group aggregate nodes, including BatchExecHashAggregate, BatchExecLocalHashAggregate, BatchExecSortAggregate, BatchExecLocalSortAggregate, BatchExecPythonGroupAggregate

Brief change log

Rename BatchExecGroupAggregateBase to BatchPhysicalGroupAggregateBase and do some refactoring
Introduce BatchPhysicalHashAggregate, and make BatchExecHashAggregate only extended from ExecNode
Introduce BatchPhysicalLocalHashAggregate, and make BatchPhysicalLocalHashAggregate only extended from FlinkPhysicalRel
Introduce BatchPhysicalSortAggregate, and make BatchExecSortAggregate only extended from ExecNode
Introduce BatchPhysicalLocalSortAggregate, and make BatchPhysicalLocalSortAggregate only extended from FlinkPhysicalRel
Introduce BatchPhysicalPythonGroupAggregate, and make BatchExecPythonGroupAggregate only extended from ExecNode

Verifying this change

This change is a refactoring rework covered by existing tests.

Does this pull request potentially affect one of the following parts:

Dependencies (does it add or upgrade a dependency): (yes / no)
The public API, i.e., is any changed class annotated with @Public(Evolving): (yes / no)
The serializers: (yes / no / don't know)
The runtime per-record code paths (performance sensitive): (yes / no / don't know)
Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Kubernetes/Yarn/Mesos, ZooKeeper: (yes / no / don't know)
The S3 file system connector: (yes / no / don't know)

Documentation

Does this pull request introduce a new feature? (yes / no)
If yes, how is the feature documented? (not applicable / docs / JavaDocs / not documented)

flinkbot · 2021-01-05T12:26:15Z

Thanks a lot for your contribution to the Apache Flink project. I'm the @flinkbot. I help the community
to review your pull request. We will use this comment to track the progress of the review.

Automated Checks

Last check on commit b882bd3 (Fri May 28 07:02:00 UTC 2021)

Warnings:

No documentation files were touched! Remember to keep the Flink docs up to date!

_{Mention the bot in a comment to re-run the automated checks.}

Review Progress

❓ 1. The [description] looks good.
❓ 2. There is [consensus] that the contribution should go into to Flink.
❓ 3. Needs [attention] from.
❓ 4. The change fits into the overall [architecture].
❓ 5. Overall code [quality] is good.

Please see the Pull Request Review Guide for a full explanation of the review process.

Details

The Bot is tracking the review progress through labels. Labels are applied according to the order of the review items. For consensus, approval by a Flink committer of PMC member is required

Bot commands

The @flinkbot bot supports the following commands:

@flinkbot approve description to approve one or more aspects (aspects: description, consensus, architecture and quality)
@flinkbot approve all to approve all aspects
@flinkbot approve-until architecture to approve everything until architecture
@flinkbot attention @username1 [@username2 ..] to require somebody's attention
@flinkbot disapprove architecture to remove an approval you gave earlier

flinkbot · 2021-01-05T12:35:59Z

CI report:

b882bd3 Azure: SUCCESS

Bot commands

The @flinkbot bot supports the following commands:

@flinkbot run travis re-run the last Travis build
@flinkbot run azure re-run the last Azure build

wenlong88

The overall change LGTM, just left some minor comments.

wenlong88 · 2021-01-06T06:17:43Z

...r-blink/src/main/java/org/apache/flink/table/planner/plan/nodes/exec/utils/ExecNodeUtil.java


+    /** Return bytes size for given option in {@link TableConfig}. */
+    public static long getMemorySize(TableConfig tableConfig, ConfigOption<String> option) {
+        return MemorySize.parse(tableConfig.getConfiguration().getString(option)).getBytes();


ConfigOption supports type of MemorySize, can we just use it?,ref:FRAMEWORK_HEAP_MEMORY

It will broke the current API, because the compile error will occur when user's before code is configuration.set( ExecutionConfigOptions.TABLE_EXEC_RESOURCE_HASH_AGG_MEMORY, "128 mb")

I create an issue to do the improvement, see https://issues.apache.org/jira/browse/FLINK-20879

wenlong88 · 2021-01-06T06:20:34Z

...main/scala/org/apache/flink/table/planner/plan/metadata/FlinkRelMdModifiedMonotonicity.scala

    // if group by an update field or group by a field mono is null, just return null
    if (inputMonotonicity == null ||
-        grouping.exists(e => inputMonotonicity.fieldMonotonicities(e) != CONSTANT)) {
+      grouping.exists(e => inputMonotonicity.fieldMonotonicities(e) != CONSTANT)) {


the old format is better

wenlong88 · 2021-01-06T06:20:51Z

...main/scala/org/apache/flink/table/planner/plan/metadata/FlinkRelMdModifiedMonotonicity.scala

    // if group by a update field or group by a field mono is null, just return null
    if (inputMonotonicity == null ||
-        grouping.exists(e => inputMonotonicity.fieldMonotonicities(e) != CONSTANT)) {
+      grouping.exists(e => inputMonotonicity.fieldMonotonicities(e) != CONSTANT)) {


wenlong88 · 2021-01-06T06:21:00Z

...main/scala/org/apache/flink/table/planner/plan/metadata/FlinkRelMdModifiedMonotonicity.scala

          val currentMono = fieldMonotonicities(index)
          if (childMono != currentMono &&
-              !aggCall.getAggregation.isInstanceOf[SqlCountAggFunction]) {
+            !aggCall.getAggregation.isInstanceOf[SqlCountAggFunction]) {


… to BatchPhysicalGroupAggregateBase and do some refactoring

…te, and make BatchExecHashAggregate only extended from ExecNode

…gregate, and make BatchPhysicalLocalHashAggregate only extended from FlinkPhysicalRel

…te, and make BatchExecSortAggregate only extended from ExecNode

…gregate, and make BatchPhysicalLocalSortAggregate only extended from FlinkPhysicalRel

…Aggregate, and make BatchExecPythonGroupAggregate only extended from ExecNode

wenlong88

LGTM

…te, and make BatchExecHashAggregate only extended from ExecNode This closes apache#14562

…gregate, and make BatchPhysicalLocalHashAggregate only extended from FlinkPhysicalRel This closes apache#14562

…te, and make BatchExecSortAggregate only extended from ExecNode This closes apache#14562

…gregate, and make BatchPhysicalLocalSortAggregate only extended from FlinkPhysicalRel This closes apache#14562

…Aggregate, and make BatchExecPythonGroupAggregate only extended from ExecNode This closes apache#14562

rmetzger added the review=description? label Jan 5, 2021

rmetzger added the component=TableSQL/Planner label Jan 5, 2021

godfreyhe force-pushed the FLINK-20738 branch 2 times, most recently from 662db1d to f08790b Compare January 6, 2021 05:52

wenlong88 reviewed Jan 6, 2021

View reviewed changes

godfreyhe added 7 commits January 7, 2021 16:07

[FLINK-20738][table-planner-blink] Rename BatchExecGroupAggregateBase…

07dc37f

… to BatchPhysicalGroupAggregateBase and do some refactoring

[FLINK-20738][table-planner-blink] Introduce BatchPhysicalHashAggrega…

00f7dd7

…te, and make BatchExecHashAggregate only extended from ExecNode

[FLINK-20738][table-planner-blink] Introduce BatchPhysicalLocalHashAg…

89efd06

…gregate, and make BatchPhysicalLocalHashAggregate only extended from FlinkPhysicalRel

[FLINK-20738][table-planner-blink] Introduce BatchPhysicalSortAggrega…

1467ca8

…te, and make BatchExecSortAggregate only extended from ExecNode

[FLINK-20738][table-planner-blink] Introduce BatchPhysicalLocalSortAg…

66b52d0

…gregate, and make BatchPhysicalLocalSortAggregate only extended from FlinkPhysicalRel

[FLINK-20738][table-planner-blink] Introduce BatchPhysicalPythonGroup…

1828121

…Aggregate, and make BatchExecPythonGroupAggregate only extended from ExecNode

address comments

b882bd3

godfreyhe force-pushed the FLINK-20738 branch from 34812b7 to b882bd3 Compare January 7, 2021 08:38

wenlong88 approved these changes Jan 7, 2021

View reviewed changes

godfreyhe added a commit to godfreyhe/flink that referenced this pull request Jan 7, 2021

[FLINK-20738][table-planner-blink] Introduce BatchPhysicalHashAggrega…

d0d8377

…te, and make BatchExecHashAggregate only extended from ExecNode This closes apache#14562

godfreyhe added a commit to godfreyhe/flink that referenced this pull request Jan 7, 2021

[FLINK-20738][table-planner-blink] Introduce BatchPhysicalLocalHashAg…

9f8f5cd

…gregate, and make BatchPhysicalLocalHashAggregate only extended from FlinkPhysicalRel This closes apache#14562

godfreyhe added a commit to godfreyhe/flink that referenced this pull request Jan 7, 2021

[FLINK-20738][table-planner-blink] Introduce BatchPhysicalSortAggrega…

29c81fe

…te, and make BatchExecSortAggregate only extended from ExecNode This closes apache#14562

godfreyhe added a commit to godfreyhe/flink that referenced this pull request Jan 7, 2021

[FLINK-20738][table-planner-blink] Introduce BatchPhysicalLocalSortAg…

63875b3

…gregate, and make BatchPhysicalLocalSortAggregate only extended from FlinkPhysicalRel This closes apache#14562

godfreyhe added a commit to godfreyhe/flink that referenced this pull request Jan 7, 2021

[FLINK-20738][table-planner-blink] Introduce BatchPhysicalPythonGroup…

049a61a

…Aggregate, and make BatchExecPythonGroupAggregate only extended from ExecNode This closes apache#14562

godfreyhe closed this in 146c68d Jan 7, 2021

godfreyhe deleted the FLINK-20738 branch January 7, 2021 14:09

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[FLINK-20738][table-planner-blink] Separate the implementation of batch group aggregate nodes #14562

[FLINK-20738][table-planner-blink] Separate the implementation of batch group aggregate nodes #14562

Uh oh!

godfreyhe commented Jan 5, 2021

Uh oh!

flinkbot commented Jan 5, 2021 •

edited

Loading

Uh oh!

flinkbot commented Jan 5, 2021 •

edited

Loading

Uh oh!

wenlong88 left a comment

Uh oh!

wenlong88 Jan 6, 2021

Uh oh!

godfreyhe Jan 6, 2021 •

edited

Loading

Uh oh!

godfreyhe Jan 7, 2021

Uh oh!

wenlong88 Jan 6, 2021

Uh oh!

wenlong88 Jan 6, 2021

Uh oh!

wenlong88 Jan 6, 2021

Uh oh!

wenlong88 left a comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

[FLINK-20738][table-planner-blink] Separate the implementation of batch group aggregate nodes #14562

[FLINK-20738][table-planner-blink] Separate the implementation of batch group aggregate nodes #14562

Uh oh!

Conversation

godfreyhe commented Jan 5, 2021

What is the purpose of the change

Brief change log

Verifying this change

Does this pull request potentially affect one of the following parts:

Documentation

Uh oh!

flinkbot commented Jan 5, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Automated Checks

Review Progress

Uh oh!

flinkbot commented Jan 5, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

CI report:

Uh oh!

wenlong88 left a comment

Choose a reason for hiding this comment

Uh oh!

wenlong88 Jan 6, 2021

Choose a reason for hiding this comment

Uh oh!

godfreyhe Jan 6, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

godfreyhe Jan 7, 2021

Choose a reason for hiding this comment

Uh oh!

wenlong88 Jan 6, 2021

Choose a reason for hiding this comment

Uh oh!

wenlong88 Jan 6, 2021

Choose a reason for hiding this comment

Uh oh!

wenlong88 Jan 6, 2021

Choose a reason for hiding this comment

Uh oh!

wenlong88 left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

flinkbot commented Jan 5, 2021 •

edited

Loading

flinkbot commented Jan 5, 2021 •

edited

Loading

godfreyhe Jan 6, 2021 •

edited

Loading