[
https://issues.apache.org/jira/browse/FLINK-38509?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18029423#comment-18029423
]
Royston Tauro edited comment on FLINK-38509 at 10/13/25 8:18 AM:
-----------------------------------------------------------------
[~gaborgsomogyi] Had found something along these lines in a few articles and
stack overflow answers
conf = SparkConf() \
.setMaster('local[*]') \
.setAppName('test') \
.set("spark.jars", "./lib/gcs-connector-hadoop3-latest.jar") \
.set("spark.hadoop.google.cloud.auth.service.account.enable", "true") \
.set("spark.hadoop.google.cloud.auth.service.account.json.keyfile",
credentials_location)
was (Author: JIRAUSER311187):
conf = SparkConf() \
.setMaster('local[*]') \
.setAppName('test') \
.set("spark.jars", "./lib/gcs-connector-hadoop3-latest.jar") \
.set("spark.hadoop.google.cloud.auth.service.account.enable", "true") \
.set("spark.hadoop.google.cloud.auth.service.account.json.keyfile",
credentials_location)
> Dynamic Credentials in Flink for Google Cloud Storage
> -----------------------------------------------------
>
> Key: FLINK-38509
> URL: https://issues.apache.org/jira/browse/FLINK-38509
> Project: Flink
> Issue Type: Improvement
> Components: Connectors / FileSystem
> Affects Versions: 2.1.0
> Reporter: Royston Tauro
> Priority: Major
>
> Currently in session cluster mode, the only way to provide credentials for
> Google Cloud Storage is using the env or the core-site.xml in hadoop config
> which is read during the creation of the cluster only.
> Are there plans to make it dynamic to Jobs i.e each job can have have its own
> credentials via flinkConfig something similar to how spark allows it.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
