[FLINK-13068][hive] HiveTableSink should implement PartitionableTable… #8965

lirui-apache · 2019-07-03T09:33:20Z

…Sink

What is the purpose of the change

Make HiveTableSink implement PartitionableTableSink, so that HiveTableSink supports static partitioning.

Brief change log

Implemented PartitionableTableSink.
Moved another test case from HiveTableOutputFormatTest to HiveTableSinkTest.

Verifying this change

Existing test case.

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

Dependencies (does it add or upgrade a dependency): no
The public API, i.e., is any changed class annotated with @Public(Evolving): no
The serializers: no
The runtime per-record code paths (performance sensitive): no
Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Yarn/Mesos, ZooKeeper: no
The S3 file system connector: no

Documentation

Does this pull request introduce a new feature? no
If yes, how is the feature documented? NA

flinkbot · 2019-07-03T09:35:09Z

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 8e18c2d (Wed Aug 07 08:15:00 UTC 2019)

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.
- Approved by @bowenli86
✅ 2. There is [consensus] that the contribution should go into to Flink.
- Approved by @bowenli86
❓ 3. Needs [attention] from.
✅ 4. The change fits into the overall [architecture].
- Approved by @bowenli86
✅ 5. Overall code [quality] is good.
- Approved by @bowenli86

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

lirui-apache · 2019-07-03T09:36:19Z

cc @zjuwangg @bowenli86 @xuefuz

...flink-connector-hive/src/main/java/org/apache/flink/batch/connectors/hive/HiveTableSink.java

xuefuz · 2019-07-03T18:01:38Z

...flink-connector-hive/src/main/java/org/apache/flink/batch/connectors/hive/HiveTableSink.java

+					Preconditions.checkArgument(numStaticPart == 0,
+							"Dynamic partition cannot appear before static partition");
+				} else {
+					numStaticPart--;


This line doesn't seem having any effect.

It actually does. We'll use it to make sure we don't see any dynamic partition columns before we have seen all the static ones.

I see. After re-read the code, the logic becomes clear. Looks good.

xuefuz · 2019-07-03T18:06:25Z

...flink-connector-hive/src/main/java/org/apache/flink/batch/connectors/hive/HiveTableSink.java

+		int numStaticPart = staticPartitionSpec.size();
+		if (numStaticPart < partitionCols.size()) {
+			for (String partitionCol : partitionCols) {
+				if (!staticPartitionSpec.containsKey(partitionCol)) {


It seems that this if condition will be true (so the assertion will fail) at a certain point as long as it's dynamic partition.

Yes. And when it's true, we will verify that we have checked all the static partition columns.

xuefuz

Thanks for the contribution. I left a couple of minor comments for consideration.

bowenli86 · 2019-07-03T18:26:28Z

...flink-connector-hive/src/main/java/org/apache/flink/batch/connectors/hive/HiveTableSink.java

+		// make it a LinkedHashMap to maintain partition column order
+		staticPartitionSpec = new LinkedHashMap<>();
+		for (String partitionCol : getPartitionFieldNames()) {
+			if (partitions.containsKey(partitionCol)) {


what if it doesn't contain the key?

It means that partition column is not contained in the static spec, and therefore it's a dynamic partition column.

bowenli86 · 2019-07-03T18:28:17Z

...flink-connector-hive/src/main/java/org/apache/flink/batch/connectors/hive/HiveTableSink.java

+
+	private void validatePartitionSpec() {
+		List<String> partitionCols = getPartitionFieldNames();
+		Preconditions.checkArgument(new HashSet<>(partitionCols).containsAll(


A few comments here:

can we output which column is unknown to the error message?

the check of new HashSet<>(partitionCols).containsAll(staticPartitionSpec.keySet()) will lose order of the partitions, which seems to be not enough for a legit validation. Shall it be partitionCols.equals(new ArrayList(staticPartitionSpec.keySet()))? since staticPartitionSpec is a linked hashmap, keys from its keySet() should be ordered

nit: can we reformat it to be more readable?

Preconditions.checkArgument( ... "...";

Yeah I'll print the specific columns and reformat the code.
We don't need the check the order here. Partition columns order is defined by getPartitionFieldNames(). We can always reorder a partition spec (which is a map) as long as it only contains valid partition columns.

lirui-apache · 2019-07-04T05:34:23Z

Thanks @xuefuz and @bowenli86 for the review. Please take another look.

lirui-apache · 2019-07-04T07:37:20Z

CI passed on my personal repo:
https://travis-ci.org/lirui-apache/flink/builds/554079872

xuefuz

LGTM.

lirui-apache · 2019-07-05T02:28:55Z

The CI failure is due to flink-table-planner-blink failed checkstyle and thus not related to this PR.

bowenli86 · 2019-07-05T18:06:46Z

LGTM, thanks for your contribution!

@flinkbot approve all

I've restarted the CI since it's failing at the compile stage. Will merge it once it passes

zentol · 2019-07-05T22:50:57Z

CI report for commit 24f08d4: FAILURE Travis

bowenli86 · 2019-07-08T18:15:13Z

@lirui-apache can you rebase the PR?

flinkbot · 2019-07-08T21:23:01Z

CI report for commit 24f08d4: FAILURE Build

…Sink

flinkbot · 2019-07-09T05:41:51Z

CI report for commit 8e18c2d: FAILURE Build

lirui-apache · 2019-07-09T06:04:59Z

The test failure cannot be reproduced locally.
Travis build of my own repo passed at: https://travis-ci.org/lirui-apache/flink/builds/556113232

lirui-apache · 2019-07-09T06:05:41Z

PR rebased. @bowenli86 please take another look, thanks.

bowenli86 · 2019-07-09T16:31:57Z

LGTM. Merging

Resolve bad merge between #8965 and #8987 though they passed CI separately.

rmetzger added the review=description? label Jul 3, 2019

rmetzger added the component=Connectors/Hive label Jul 3, 2019

xuefuz reviewed Jul 3, 2019

View reviewed changes

...flink-connector-hive/src/main/java/org/apache/flink/batch/connectors/hive/HiveTableSink.java Outdated Show resolved Hide resolved

xuefuz reviewed Jul 3, 2019

View reviewed changes

bowenli86 reviewed Jul 3, 2019

View reviewed changes

lirui-apache force-pushed the FLINK-13068 branch from 249bb26 to 24f08d4 Compare July 4, 2019 05:23

xuefuz approved these changes Jul 4, 2019

View reviewed changes

lirui-apache added 3 commits July 9, 2019 10:52

[FLINK-13068][hive] HiveTableSink should implement PartitionableTable…

b2e0843

…Sink

address comments

388fc9b

rebase

8e18c2d

lirui-apache force-pushed the FLINK-13068 branch from 24f08d4 to 8e18c2d Compare July 9, 2019 04:16

asfgit closed this in d5f8078 Jul 9, 2019

asfgit pushed a commit that referenced this pull request Jul 9, 2019

[hotfix][hive] resolve bad merge in HiveTableSinkTest

fac54e9

Resolve bad merge between #8965 and #8987 though they passed CI separately.

lirui-apache deleted the FLINK-13068 branch July 10, 2019 03:10

[FLINK-13068][hive] HiveTableSink should implement PartitionableTable… #8965

[FLINK-13068][hive] HiveTableSink should implement PartitionableTable… #8965

Uh oh!

Conversation

lirui-apache commented Jul 3, 2019

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 Jul 3, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Automated Checks

Review Progress

Uh oh!

lirui-apache commented Jul 3, 2019

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

xuefuz left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lirui-apache commented Jul 4, 2019

Uh oh!

lirui-apache commented Jul 4, 2019

Uh oh!

xuefuz left a comment

Choose a reason for hiding this comment

Uh oh!

lirui-apache commented Jul 5, 2019

Uh oh!

bowenli86 commented Jul 5, 2019

Uh oh!

zentol commented Jul 5, 2019

Uh oh!

bowenli86 commented Jul 8, 2019

Uh oh!

flinkbot commented Jul 8, 2019

Uh oh!

flinkbot commented Jul 9, 2019

Uh oh!

lirui-apache commented Jul 9, 2019

Uh oh!

lirui-apache commented Jul 9, 2019

Uh oh!

bowenli86 commented Jul 9, 2019

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

flinkbot commented Jul 3, 2019 •

edited

Loading