You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "gustavodemorais (via GitHub)" <gi...@apache.org> on 2025/08/07 12:29:44 UTC
[PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management [flink]
gustavodemorais opened a new pull request, #26880:
URL: https://github.com/apache/flink/pull/26880
<!--
*Thank you very much for contributing to Apache Flink - we are happy that you want to help us improve Flink. To help the community review your contribution in the best possible way, please go through the checklist below, which will get the contribution into a shape in which it can be best reviewed.*
*Please understand that we do not do this to make contributions to Flink a hassle. In order to uphold a high standard of quality for code contributions, while at the same time managing a large number of contributions, we need contributors to prepare the contributions well, and give reviewers enough contextual information for the review. Please also understand that contributions that do not follow this guide will take longer to review and thus typically be picked up with lower priority by the community.*
## Contribution Checklist
- Make sure that the pull request corresponds to a [JIRA issue](https://issues.apache.org/jira/projects/FLINK/issues). Exceptions are made for typos in JavaDoc or documentation files, which need no JIRA issue.
- Name the pull request in the form "[FLINK-XXXX] [component] Title of the pull request", where *FLINK-XXXX* should be replaced by the actual issue number. Skip *component* if you are unsure about which is the best component.
Typo fixes that have no associated JIRA issue should be named following this pattern: `[hotfix] [docs] Fix typo in event time introduction` or `[hotfix] [javadocs] Expand JavaDoc for PuncuatedWatermarkGenerator`.
- Fill out the template below to describe the changes contributed by the pull request. That will give reviewers the context they need to do the review.
- Make sure that the change passes the automated tests, i.e., `mvn clean verify` passes. You can set up Azure Pipelines CI to do that following [this guide](https://cwiki.apache.org/confluence/display/FLINK/Azure+Pipelines#AzurePipelines-Tutorial:SettingupAzurePipelinesforaforkoftheFlinkrepository).
- Each pull request should address only one issue, not mix up code from multiple issues.
- Each commit in the pull request has a meaningful commit message (including the JIRA id)
- Once all items of the checklist are addressed, remove the above text and this checklist, leaving only the filled out template below.
**(The sections below can be removed for hotfixes of typos)**
-->
## What is the purpose of the change
When creating MultiJoins, we currently use UpsertKeys as the UniqueKeys for the inputSideSpecs. Even though that is not wrong since upsertKeys are always unique keys, that makes us only use UniqueKeys optimizations for a subset of cases when we have unique keys.
In other words, for joins, we'll often use a state with hasNoUniqueKey even though there's an unique key. This is a considerable optimization and we want to change that.
## Brief change log
- Pass down unique keys instead of upsert keys
- Update restore tests **(this is ok because the operator was released in an experimental state)**
## Verifying this change
- Updated restore tests: You can see now that in restore tests for cases where we have a unique key and no upsertKeys, we will pass that the unique key, which will be used when picking up the state view to use.
## 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, Kubernetes/Yarn, 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? (not applicable)
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management for MultiJoin [flink]
Posted by "snuyanzin (via GitHub)" <gi...@apache.org>.
snuyanzin commented on code in PR #26880:
URL: https://github.com/apache/flink/pull/26880#discussion_r2260194734
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/nodes/exec/stream/StreamExecMultiJoin.java:
##########
@@ -81,7 +81,7 @@ public class StreamExecMultiJoin extends ExecNodeBase<RowData>
private static final String FIELD_NAME_JOIN_TYPES = "joinTypes";
private static final String FIELD_NAME_JOIN_CONDITIONS = "joinConditions";
private static final String FIELD_NAME_JOIN_ATTRIBUTE_MAP = "joinAttributeMap";
- private static final String FIELD_NAME_INPUT_UPSERT_KEYS = "inputUpsertKeys";
+ private static final String FIELD_NAME_INPUT_UNIQUE_KEYS = "inputUniqueKeys";
Review Comment:
in that case it probably should be mentioned in release notes of jira issue...
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management for MultiJoin [flink]
Posted by "gustavodemorais (via GitHub)" <gi...@apache.org>.
gustavodemorais commented on code in PR #26880:
URL: https://github.com/apache/flink/pull/26880#discussion_r2261674148
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/nodes/physical/stream/StreamPhysicalMultiJoin.java:
##########
@@ -194,26 +194,27 @@ private RexNode createMultiJoinCondition() {
return RexUtil.composeConjunction(getCluster().getRexBuilder(), conjunctions, true);
}
- private List<List<int[]>> getUpsertKeysForInputs() {
+ private List<List<int[]>> getUniqueKeysForInputs() {
return inputs.stream()
.map(
input -> {
- final Set<ImmutableBitSet> upsertKeys = getUpsertKeys(input);
+ final Set<ImmutableBitSet> uniqueKeys = getUniqueKeys(input);
- if (upsertKeys == null) {
+ if (uniqueKeys == null) {
return Collections.<int[]>emptyList();
}
- return upsertKeys.stream()
+
+ return uniqueKeys.stream()
.map(ImmutableBitSet::toArray)
.collect(Collectors.toList());
})
.collect(Collectors.toList());
}
- private @Nullable Set<ImmutableBitSet> getUpsertKeys(RelNode input) {
+ private @Nullable Set<ImmutableBitSet> getUniqueKeys(RelNode input) {
final FlinkRelMetadataQuery fmq =
FlinkRelMetadataQuery.reuseOrCreate(input.getCluster().getMetadataQuery());
- return fmq.getUpsertKeys(input);
+ return fmq.getUniqueKeys(input);
Review Comment:
Added MULTI_JOIN_MIXED_CHANGELOG_MODES
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management for MultiJoin [flink]
Posted by "twalthr (via GitHub)" <gi...@apache.org>.
twalthr merged PR #26880:
URL: https://github.com/apache/flink/pull/26880
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management for MultiJoin [flink]
Posted by "gustavodemorais (via GitHub)" <gi...@apache.org>.
gustavodemorais commented on code in PR #26880:
URL: https://github.com/apache/flink/pull/26880#discussion_r2260222010
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/nodes/exec/stream/StreamExecMultiJoin.java:
##########
@@ -81,7 +81,7 @@ public class StreamExecMultiJoin extends ExecNodeBase<RowData>
private static final String FIELD_NAME_JOIN_TYPES = "joinTypes";
private static final String FIELD_NAME_JOIN_CONDITIONS = "joinConditions";
private static final String FIELD_NAME_JOIN_ATTRIBUTE_MAP = "joinAttributeMap";
- private static final String FIELD_NAME_INPUT_UPSERT_KEYS = "inputUpsertKeys";
+ private static final String FIELD_NAME_INPUT_UNIQUE_KEYS = "inputUniqueKeys";
Review Comment:
Does that look alright? https://issues.apache.org/jira/browse/FLINK-38209 If not, you can point me to other tickets were similar release notes were added
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management for MultiJoin [flink]
Posted by "gustavodemorais (via GitHub)" <gi...@apache.org>.
gustavodemorais commented on code in PR #26880:
URL: https://github.com/apache/flink/pull/26880#discussion_r2260219981
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/nodes/exec/stream/StreamExecMultiJoin.java:
##########
@@ -81,7 +81,7 @@ public class StreamExecMultiJoin extends ExecNodeBase<RowData>
private static final String FIELD_NAME_JOIN_TYPES = "joinTypes";
private static final String FIELD_NAME_JOIN_CONDITIONS = "joinConditions";
private static final String FIELD_NAME_JOIN_ATTRIBUTE_MAP = "joinAttributeMap";
- private static final String FIELD_NAME_INPUT_UPSERT_KEYS = "inputUpsertKeys";
+ private static final String FIELD_NAME_INPUT_UNIQUE_KEYS = "inputUniqueKeys";
Review Comment:
Good observation! Done 👍
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management for MultiJoin [flink]
Posted by "gustavodemorais (via GitHub)" <gi...@apache.org>.
gustavodemorais commented on code in PR #26880:
URL: https://github.com/apache/flink/pull/26880#discussion_r2260222010
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/nodes/exec/stream/StreamExecMultiJoin.java:
##########
@@ -81,7 +81,7 @@ public class StreamExecMultiJoin extends ExecNodeBase<RowData>
private static final String FIELD_NAME_JOIN_TYPES = "joinTypes";
private static final String FIELD_NAME_JOIN_CONDITIONS = "joinConditions";
private static final String FIELD_NAME_JOIN_ATTRIBUTE_MAP = "joinAttributeMap";
- private static final String FIELD_NAME_INPUT_UPSERT_KEYS = "inputUpsertKeys";
+ private static final String FIELD_NAME_INPUT_UNIQUE_KEYS = "inputUniqueKeys";
Review Comment:
Does that look alright? https://issues.apache.org/jira/browse/FLINK-38209 If not, you can point me to other tickets where similar release notes were added
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management for MultiJoin [flink]
Posted by "gustavodemorais (via GitHub)" <gi...@apache.org>.
gustavodemorais commented on code in PR #26880:
URL: https://github.com/apache/flink/pull/26880#discussion_r2260191260
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/nodes/exec/stream/StreamExecMultiJoin.java:
##########
@@ -81,7 +81,7 @@ public class StreamExecMultiJoin extends ExecNodeBase<RowData>
private static final String FIELD_NAME_JOIN_TYPES = "joinTypes";
private static final String FIELD_NAME_JOIN_CONDITIONS = "joinConditions";
private static final String FIELD_NAME_JOIN_ATTRIBUTE_MAP = "joinAttributeMap";
- private static final String FIELD_NAME_INPUT_UPSERT_KEYS = "inputUpsertKeys";
+ private static final String FIELD_NAME_INPUT_UNIQUE_KEYS = "inputUniqueKeys";
Review Comment:
Yes, it is. But that is ok here.
We launched the operator in an experimental state and explicitly stated in the docs and release notes that that's the case and optimizations with breaking changes are being worked on.
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management for MultiJoin [flink]
Posted by "flinkbot (via GitHub)" <gi...@apache.org>.
flinkbot commented on PR #26880:
URL: https://github.com/apache/flink/pull/26880#issuecomment-3164000980
<!--
Meta data
{
"version" : 1,
"metaDataEntries" : [ {
"hash" : "e534d9d7987a3a848cfcb6c6c893c913be2194dc",
"status" : "UNKNOWN",
"url" : "TBD",
"triggerID" : "e534d9d7987a3a848cfcb6c6c893c913be2194dc",
"triggerType" : "PUSH"
} ]
}-->
## CI report:
* e534d9d7987a3a848cfcb6c6c893c913be2194dc UNKNOWN
<details>
<summary>Bot commands</summary>
The @flinkbot bot supports the following commands:
- `@flinkbot run azure` re-run the last Azure build
</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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management for MultiJoin [flink]
Posted by "gustavodemorais (via GitHub)" <gi...@apache.org>.
gustavodemorais commented on code in PR #26880:
URL: https://github.com/apache/flink/pull/26880#discussion_r2262660425
##########
flink-table/flink-table-planner/src/test/java/org/apache/flink/table/planner/plan/nodes/exec/stream/MultiJoinTestPrograms.java:
##########
@@ -907,4 +907,65 @@ public class MultiJoinTestPrograms {
+ "INNER JOIN OrdersWithRowtime o ON u.user_id_0 = o.user_id_1 "
+ "INNER JOIN Payments p ON u.user_id_0 = p.user_id_2")
.build();
+
+ public static final TableTestProgram MULTI_JOIN_MIXED_CHANGELOG_MODES =
+ TableTestProgram.of(
+ "three-way-mixed-changelog-modes",
+ "three way join with mixed changelog modes and primary key configurations")
+ .setupConfig(OptimizerConfigOptions.TABLE_OPTIMIZER_MULTI_JOIN_ENABLED, true)
+ .setupTableSource(
+ SourceTestStep.newBuilder("AppendTable")
+ .addSchema("id STRING PRIMARY KEY NOT ENFORCED, val STRING")
+ .addOption("changelog-mode", "I")
+ .producedValues(
+ Row.ofKind(RowKind.INSERT, "1", "append1"),
+ Row.ofKind(RowKind.INSERT, "2", "append2"),
+ Row.ofKind(RowKind.INSERT, "3", "append3"))
+ .build())
+ .setupTableSource(
+ SourceTestStep.newBuilder("RetractTable")
+ .addSchema("ref_id STRING, data STRING")
+ .addOption("changelog-mode", "I,UA,UB,D")
+ .producedValues(
+ Row.ofKind(RowKind.INSERT, "1", "retract1"),
+ Row.ofKind(RowKind.INSERT, "2", "retract2"),
+ Row.ofKind(RowKind.INSERT, "3", "retract3"),
+ Row.ofKind(RowKind.DELETE, "3", "retract3"),
+ Row.ofKind(RowKind.INSERT, "1", "retract1_new"))
+ .build())
+ .setupTableSource(
+ SourceTestStep.newBuilder("UpsertTable")
+ .addSchema(
+ "key_id STRING PRIMARY KEY NOT ENFORCED, status STRING")
+ .addOption("changelog-mode", "I,UA,D")
+ .producedValues(
+ Row.ofKind(RowKind.INSERT, "1", "active"),
+ Row.ofKind(RowKind.INSERT, "2", "pending"),
+ Row.ofKind(RowKind.UPDATE_BEFORE, "2", "pending"),
Review Comment:
Oops, 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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management for MultiJoin [flink]
Posted by "twalthr (via GitHub)" <gi...@apache.org>.
twalthr commented on code in PR #26880:
URL: https://github.com/apache/flink/pull/26880#discussion_r2262588859
##########
flink-table/flink-table-planner/src/test/java/org/apache/flink/table/planner/plan/nodes/exec/stream/MultiJoinTestPrograms.java:
##########
@@ -907,4 +907,65 @@ public class MultiJoinTestPrograms {
+ "INNER JOIN OrdersWithRowtime o ON u.user_id_0 = o.user_id_1 "
+ "INNER JOIN Payments p ON u.user_id_0 = p.user_id_2")
.build();
+
+ public static final TableTestProgram MULTI_JOIN_MIXED_CHANGELOG_MODES =
+ TableTestProgram.of(
+ "three-way-mixed-changelog-modes",
+ "three way join with mixed changelog modes and primary key configurations")
+ .setupConfig(OptimizerConfigOptions.TABLE_OPTIMIZER_MULTI_JOIN_ENABLED, true)
+ .setupTableSource(
+ SourceTestStep.newBuilder("AppendTable")
+ .addSchema("id STRING PRIMARY KEY NOT ENFORCED, val STRING")
+ .addOption("changelog-mode", "I")
+ .producedValues(
+ Row.ofKind(RowKind.INSERT, "1", "append1"),
+ Row.ofKind(RowKind.INSERT, "2", "append2"),
+ Row.ofKind(RowKind.INSERT, "3", "append3"))
+ .build())
+ .setupTableSource(
+ SourceTestStep.newBuilder("RetractTable")
+ .addSchema("ref_id STRING, data STRING")
+ .addOption("changelog-mode", "I,UA,UB,D")
+ .producedValues(
+ Row.ofKind(RowKind.INSERT, "1", "retract1"),
+ Row.ofKind(RowKind.INSERT, "2", "retract2"),
+ Row.ofKind(RowKind.INSERT, "3", "retract3"),
+ Row.ofKind(RowKind.DELETE, "3", "retract3"),
+ Row.ofKind(RowKind.INSERT, "1", "retract1_new"))
+ .build())
+ .setupTableSource(
+ SourceTestStep.newBuilder("UpsertTable")
+ .addSchema(
+ "key_id STRING PRIMARY KEY NOT ENFORCED, status STRING")
+ .addOption("changelog-mode", "I,UA,D")
+ .producedValues(
+ Row.ofKind(RowKind.INSERT, "1", "active"),
+ Row.ofKind(RowKind.INSERT, "2", "pending"),
+ Row.ofKind(RowKind.UPDATE_BEFORE, "2", "pending"),
Review Comment:
Upsert should not contain an update before.
##########
flink-table/flink-table-planner/src/test/java/org/apache/flink/table/planner/plan/nodes/exec/stream/MultiJoinTestPrograms.java:
##########
@@ -907,4 +907,65 @@ public class MultiJoinTestPrograms {
+ "INNER JOIN OrdersWithRowtime o ON u.user_id_0 = o.user_id_1 "
+ "INNER JOIN Payments p ON u.user_id_0 = p.user_id_2")
.build();
+
+ public static final TableTestProgram MULTI_JOIN_MIXED_CHANGELOG_MODES =
+ TableTestProgram.of(
+ "three-way-mixed-changelog-modes",
+ "three way join with mixed changelog modes and primary key configurations")
+ .setupConfig(OptimizerConfigOptions.TABLE_OPTIMIZER_MULTI_JOIN_ENABLED, true)
+ .setupTableSource(
+ SourceTestStep.newBuilder("AppendTable")
+ .addSchema("id STRING PRIMARY KEY NOT ENFORCED, val STRING")
+ .addOption("changelog-mode", "I")
+ .producedValues(
+ Row.ofKind(RowKind.INSERT, "1", "append1"),
+ Row.ofKind(RowKind.INSERT, "2", "append2"),
+ Row.ofKind(RowKind.INSERT, "3", "append3"))
+ .build())
+ .setupTableSource(
+ SourceTestStep.newBuilder("RetractTable")
+ .addSchema("ref_id STRING, data STRING")
+ .addOption("changelog-mode", "I,UA,UB,D")
+ .producedValues(
+ Row.ofKind(RowKind.INSERT, "1", "retract1"),
+ Row.ofKind(RowKind.INSERT, "2", "retract2"),
+ Row.ofKind(RowKind.INSERT, "3", "retract3"),
+ Row.ofKind(RowKind.DELETE, "3", "retract3"),
+ Row.ofKind(RowKind.INSERT, "1", "retract1_new"))
+ .build())
+ .setupTableSource(
+ SourceTestStep.newBuilder("UpsertTable")
+ .addSchema(
+ "key_id STRING PRIMARY KEY NOT ENFORCED, status STRING")
+ .addOption("changelog-mode", "I,UA,D")
+ .producedValues(
+ Row.ofKind(RowKind.INSERT, "1", "active"),
+ Row.ofKind(RowKind.INSERT, "2", "pending"),
+ Row.ofKind(RowKind.UPDATE_BEFORE, "2", "pending"),
+ Row.ofKind(RowKind.UPDATE_AFTER, "2", "active"),
+ Row.ofKind(RowKind.INSERT, "3", "inactive"),
+ Row.ofKind(RowKind.DELETE, "3", "inactive"))
+ .build())
+ .setupTableSink(
+ SinkTestStep.newBuilder("sink")
+ .addSchema(
+ "id STRING",
+ "val STRING",
+ "data STRING",
+ "status STRING")
+ .addOption("changelog-mode", "I,UA,UB,D")
Review Comment:
sinks use a different flag, otherwise the mode is not considered: `sink-changelog-mode-enforced`
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management for MultiJoin [flink]
Posted by "gustavodemorais (via GitHub)" <gi...@apache.org>.
gustavodemorais commented on code in PR #26880:
URL: https://github.com/apache/flink/pull/26880#discussion_r2262666010
##########
flink-table/flink-table-planner/src/test/java/org/apache/flink/table/planner/plan/nodes/exec/stream/MultiJoinTestPrograms.java:
##########
@@ -907,4 +907,65 @@ public class MultiJoinTestPrograms {
+ "INNER JOIN OrdersWithRowtime o ON u.user_id_0 = o.user_id_1 "
+ "INNER JOIN Payments p ON u.user_id_0 = p.user_id_2")
.build();
+
+ public static final TableTestProgram MULTI_JOIN_MIXED_CHANGELOG_MODES =
+ TableTestProgram.of(
+ "three-way-mixed-changelog-modes",
+ "three way join with mixed changelog modes and primary key configurations")
+ .setupConfig(OptimizerConfigOptions.TABLE_OPTIMIZER_MULTI_JOIN_ENABLED, true)
+ .setupTableSource(
+ SourceTestStep.newBuilder("AppendTable")
+ .addSchema("id STRING PRIMARY KEY NOT ENFORCED, val STRING")
+ .addOption("changelog-mode", "I")
+ .producedValues(
+ Row.ofKind(RowKind.INSERT, "1", "append1"),
+ Row.ofKind(RowKind.INSERT, "2", "append2"),
+ Row.ofKind(RowKind.INSERT, "3", "append3"))
+ .build())
+ .setupTableSource(
+ SourceTestStep.newBuilder("RetractTable")
+ .addSchema("ref_id STRING, data STRING")
+ .addOption("changelog-mode", "I,UA,UB,D")
+ .producedValues(
+ Row.ofKind(RowKind.INSERT, "1", "retract1"),
+ Row.ofKind(RowKind.INSERT, "2", "retract2"),
+ Row.ofKind(RowKind.INSERT, "3", "retract3"),
+ Row.ofKind(RowKind.DELETE, "3", "retract3"),
+ Row.ofKind(RowKind.INSERT, "1", "retract1_new"))
+ .build())
+ .setupTableSource(
+ SourceTestStep.newBuilder("UpsertTable")
+ .addSchema(
+ "key_id STRING PRIMARY KEY NOT ENFORCED, status STRING")
+ .addOption("changelog-mode", "I,UA,D")
+ .producedValues(
+ Row.ofKind(RowKind.INSERT, "1", "active"),
+ Row.ofKind(RowKind.INSERT, "2", "pending"),
+ Row.ofKind(RowKind.UPDATE_BEFORE, "2", "pending"),
+ Row.ofKind(RowKind.UPDATE_AFTER, "2", "active"),
+ Row.ofKind(RowKind.INSERT, "3", "inactive"),
+ Row.ofKind(RowKind.DELETE, "3", "inactive"))
+ .build())
+ .setupTableSink(
+ SinkTestStep.newBuilder("sink")
+ .addSchema(
+ "id STRING",
+ "val STRING",
+ "data STRING",
+ "status STRING")
+ .addOption("changelog-mode", "I,UA,UB,D")
Review Comment:
Ah, thanks for the hint. I also updated the other test programs in the file. Everything still green.
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management for MultiJoin [flink]
Posted by "twalthr (via GitHub)" <gi...@apache.org>.
twalthr commented on code in PR #26880:
URL: https://github.com/apache/flink/pull/26880#discussion_r2260702156
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/nodes/exec/stream/StreamExecMultiJoin.java:
##########
@@ -99,12 +99,14 @@ public class StreamExecMultiJoin extends ExecNodeBase<RowData>
@JsonInclude(JsonInclude.Include.NON_EMPTY)
private final Map<Integer, List<ConditionAttributeRef>> joinAttributeMap;
- @JsonProperty(FIELD_NAME_INPUT_UPSERT_KEYS)
+ @JsonProperty(FIELD_NAME_INPUT_UNIQUE_KEYS)
@JsonInclude(JsonInclude.Include.NON_EMPTY)
- // List of upsert keys for each input, where each inner list corresponds to an input
- // The reason it's a List<List<int[]>> is that SQL allows only one primary key but
- // multiple upsert (unique) keys per input
- private final List<List<int[]>> inputUpsertKeys;
+
Review Comment:
nit: remove empty line and put the annotations over the code, the comment can be on top
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/nodes/physical/stream/StreamPhysicalMultiJoin.java:
##########
@@ -194,26 +194,27 @@ private RexNode createMultiJoinCondition() {
return RexUtil.composeConjunction(getCluster().getRexBuilder(), conjunctions, true);
}
- private List<List<int[]>> getUpsertKeysForInputs() {
+ private List<List<int[]>> getUniqueKeysForInputs() {
return inputs.stream()
.map(
input -> {
- final Set<ImmutableBitSet> upsertKeys = getUpsertKeys(input);
+ final Set<ImmutableBitSet> uniqueKeys = getUniqueKeys(input);
- if (upsertKeys == null) {
+ if (uniqueKeys == null) {
return Collections.<int[]>emptyList();
}
- return upsertKeys.stream()
+
+ return uniqueKeys.stream()
.map(ImmutableBitSet::toArray)
.collect(Collectors.toList());
})
.collect(Collectors.toList());
}
- private @Nullable Set<ImmutableBitSet> getUpsertKeys(RelNode input) {
+ private @Nullable Set<ImmutableBitSet> getUniqueKeys(RelNode input) {
final FlinkRelMetadataQuery fmq =
FlinkRelMetadataQuery.reuseOrCreate(input.getCluster().getMetadataQuery());
- return fmq.getUpsertKeys(input);
+ return fmq.getUniqueKeys(input);
Review Comment:
Could we add one final test that uses 1 append table with primary key, 1 retract table without primary key, and one upsert table with primary key? Just as a final validation that unique keys and changelog mode nicely fit 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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] [FLINK-38209][table-planner] Use UniqueKeys instead of Upsertkeys for inputSpec and state management for MultiJoin [flink]
Posted by "snuyanzin (via GitHub)" <gi...@apache.org>.
snuyanzin commented on code in PR #26880:
URL: https://github.com/apache/flink/pull/26880#discussion_r2260181927
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/nodes/exec/stream/StreamExecMultiJoin.java:
##########
@@ -81,7 +81,7 @@ public class StreamExecMultiJoin extends ExecNodeBase<RowData>
private static final String FIELD_NAME_JOIN_TYPES = "joinTypes";
private static final String FIELD_NAME_JOIN_CONDITIONS = "joinConditions";
private static final String FIELD_NAME_JOIN_ATTRIBUTE_MAP = "joinAttributeMap";
- private static final String FIELD_NAME_INPUT_UPSERT_KEYS = "inputUpsertKeys";
+ private static final String FIELD_NAME_INPUT_UNIQUE_KEYS = "inputUniqueKeys";
Review Comment:
looks like a breaking change, no?
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org