[
https://issues.apache.org/jira/browse/APEXCORE-494?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15388560#comment-15388560
]
ASF GitHub Bot commented on APEXCORE-494:
-----------------------------------------
Github user vrozov commented on a diff in the pull request:
https://github.com/apache/apex-core/pull/359#discussion_r71801328
--- Diff:
engine/src/main/java/com/datatorrent/stram/plan/physical/PhysicalPlan.java ---
@@ -895,8 +895,10 @@ private void redoPartitions(PMapping currentMapping,
String note)
addedPartitions.add(newPartition);
} else {
// check whether mapping was changed
+ int currentPartitionsSize = mainPC.currentPartitions.size();
for (DefaultPartition<Operator> pi : mainPC.currentPartitions) {
- if (pi == newPartition && pi.isModified()) {
+ if (pi == newPartition && (pi.isModified() ||
--- End diff --
@tweise Should not redeploy of downstream being done only when downstream
pipeline has a unifier and only unifier needs to be redeployed for the pipeline
connected to the current partition? For example, if logical dag is defined as:
```
A->B-C->Aggregator
```
and it is partitioned to 2 pipelines starting with A:
```
A1->B1->C1->
U->Aggregator
A2->B2->C2->
```
and A is repartitioned to 3:
```
A1->B1->C1->
A2->B2->C2->U->Aggregator
A3->B3->C3->
```
A1->B1->C1 and A2->B2->C2 don't need to be redeploed?
> Window id of downstream operator is not moving after dynamic partition of
> upstream operator.
> --------------------------------------------------------------------------------------------
>
> Key: APEXCORE-494
> URL: https://issues.apache.org/jira/browse/APEXCORE-494
> Project: Apache Apex Core
> Issue Type: Bug
> Reporter: Chaitanya
> Assignee: Chaitanya
> Attachments: PhysicalDAGAfterKafkaRepartition.png,
> PhysicalDAGBeforeKafkaRepartition.png,
> PhysicalOperatorAfterKafkaRepartition.png
>
>
> Created the following application to reproduce this issue.
> Application has the following DAG:
> KafkaSinglePortStringInputOperator(Input) -> ConsoleOutputOperator(Output)
>
> I launched the application with below configuration:
> Kafka topic created with single partition and replication factor as 1.
> Partition Strategy: ONE_TO_ONE
> Launched the application successfully. After some time, I increased the
> topic partitions to 2. After re-partition, the window of down stream operator
> is not moving. By looking into the app Physical DAG, it looks like there is
> an issue in construction of Physical DAG after re-partition.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
