[ 
https://issues.apache.org/jira/browse/FLINK-28069?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

SunShun updated FLINK-28069:
----------------------------
    Environment: Flink on Yarn, session mode  (was: I am using the session mode 
to launch the job as below:

```

./bin/yarn-session.sh --detached

./bin/flink run ./job/finance-libra-stats-1.0.jar -yt ./client.keystore.jks 
./client.truststore.jks

```)

> Cannot attach SSL JKS file for Kafka connector
> ----------------------------------------------
>
>                 Key: FLINK-28069
>                 URL: https://issues.apache.org/jira/browse/FLINK-28069
>             Project: Flink
>          Issue Type: Bug
>    Affects Versions: 1.14.4
>         Environment: Flink on Yarn, session mode
>            Reporter: SunShun
>            Priority: Blocker
>
> Hi, I intend to connect to a SSL enabled Kafka, which require to attach JKS 
> file for truststore and keystore, then I am trying to use the keyword {*}-yt, 
> yarnship{*}, to pass the JKS files, and it's expected to find the find under 
> the classpath of job manager or task manager.
> However, the FileNotFound exception is given when the job initiates, saying 
> that the JKS files is not found from job manager.
>  
> val truststore_path = "client.truststore.jks"
> val keystore_path = "client.keystore.jks"
> val source_ddl =
> s"""
> CREATE TABLE source_table(
> `id` BIGINT,
> `time` TIMESTAMP(3)
> ) WITH (
> 'connector' = 'kafka',
> ...
> 'properties.security.protocol' = 'SSL',
> 'properties.ssl.truststore.location' = '$truststore_path',
> 'properties.ssl.truststore.password' = '$truststore_pwd',
> 'properties.ssl.keystore.location' = '$keystore_path',
> 'properties.ssl.keystore.password' = '$keystore_pwd',
> 'scan.startup.mode' = 'latest-offset',
> 'format' = 'json'
> )
> """
> I am using the session mode to launch the job as below:
> ./bin/yarn-session.sh --detached
> ./bin/flink run ./job/finance-libra-stats-1.0.jar -yt ./client.keystore.jks 
> ./client.truststore.jks
>  
> Anyone can help tell what's the best practice for such case, and any mistake 
> I meet during the process.



--
This message was sent by Atlassian Jira
(v8.20.7#820007)

Reply via email to