[
https://issues.apache.org/jira/browse/FLINK-29144?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Arseniy Tashoyan updated FLINK-29144:
-------------------------------------
Description:
The setting _job.jarURI_ accepts a string with the path to the jar-file:
{code:yaml}
job:
jarURI: local:///opt/flink/examples/streaming/StateMachineExample.jar
{code}
This could be improved to accept a list of jars:
{code:yaml}
job:
jarURIs:
- local:///opt/flink/examples/streaming/StateMachineExample.jar
- local:///opt/libext/scala-logging.jar
{code}
This could also be improved to accept a directory with jars:
{code:yaml}
job:
jarDir: ///opt/basic/lib
{code}
Internally, Flink Kubernetes Operator uses the property _pipeline.jars_ - see
[FlinkConfigBuilder.java
|https://github.com/apache/flink-kubernetes-operator/blob/main/flink-kubernetes-operator/src/main/java/org/apache/flink/kubernetes/operator/config/FlinkConfigBuilder.java#L259]:
{code:java}
effectiveConfig.set(PipelineOptions.JARS,
Collections.singletonList(uri.toString()));
{code}
The property _pipeline.jars_ allows to pass more than one jar entry.
This improvement allows to avoid building a fat-jar. Instead we could provide a
directory with normal jars.
was:
The setting _job.jarURI_ accepts a string with the path to the jar-file:
{code:yaml}
job:
jarURI: local:///opt/flink/examples/streaming/StateMachineExample.jar
{code}
This could be improved to accept a list of jars:
{code:yaml}
job:
jarURIs:
- local:///opt/flink/examples/streaming/StateMachineExample.jar
- local:///opt/libext/scala-logging.jar
{code}
This could also be improved to accept a directory with jars:
{code:yaml}
job:
jarDir: ///opt/basic/lib
{code}
Internally, Flink Kubernetes Operator uses the property _pipeline.jars_ - see
[FlinkConfigBuilder.java
|https://github.com/apache/flink-kubernetes-operator/blob/main/flink-kubernetes-operator/src/main/java/org/apache/flink/kubernetes/operator/config/FlinkConfigBuilder.java#L259]:
{code:java}
effectiveConfig.set(PipelineOptions.JARS,
Collections.singletonList(uri.toString()));
{code}
The property _pipeline.jars_ allows to pass more than one jar entry.
This improvement allows to avoid building a fat-jar. Instead we could provide a
directory with normal jars.
> Enable multiple jar entries for jarURI
> --------------------------------------
>
> Key: FLINK-29144
> URL: https://issues.apache.org/jira/browse/FLINK-29144
> Project: Flink
> Issue Type: Improvement
> Components: Kubernetes Operator
> Affects Versions: 1.15.1
> Reporter: Arseniy Tashoyan
> Priority: Major
>
> The setting _job.jarURI_ accepts a string with the path to the jar-file:
> {code:yaml}
> job:
> jarURI: local:///opt/flink/examples/streaming/StateMachineExample.jar
> {code}
> This could be improved to accept a list of jars:
> {code:yaml}
> job:
> jarURIs:
> - local:///opt/flink/examples/streaming/StateMachineExample.jar
> - local:///opt/libext/scala-logging.jar
> {code}
> This could also be improved to accept a directory with jars:
> {code:yaml}
> job:
> jarDir: ///opt/basic/lib
> {code}
> Internally, Flink Kubernetes Operator uses the property _pipeline.jars_ - see
> [FlinkConfigBuilder.java
> |https://github.com/apache/flink-kubernetes-operator/blob/main/flink-kubernetes-operator/src/main/java/org/apache/flink/kubernetes/operator/config/FlinkConfigBuilder.java#L259]:
> {code:java}
> effectiveConfig.set(PipelineOptions.JARS,
> Collections.singletonList(uri.toString()));
> {code}
> The property _pipeline.jars_ allows to pass more than one jar entry.
> This improvement allows to avoid building a fat-jar. Instead we could provide
> a directory with normal jars.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
