You are viewing a plain text version of this content. The canonical link for it is here.
Posted to jira@kafka.apache.org by "Brandon Moe (Jira)" <ji...@apache.org> on 2020/04/08 03:31:00 UTC

[jira] [Updated] (KAFKA-9833) Kafka Connect: Allow environment variable to set output directory of Kafka Connect logs

     [ https://issues.apache.org/jira/browse/KAFKA-9833?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Brandon Moe updated KAFKA-9833:
-------------------------------
    Summary: Kafka Connect: Allow environment variable to set output directory of Kafka Connect logs  (was: Allow environment variable to set output directory of Kafka Connect logs)

> Kafka Connect: Allow environment variable to set output directory of Kafka Connect logs
> ---------------------------------------------------------------------------------------
>
>                 Key: KAFKA-9833
>                 URL: https://issues.apache.org/jira/browse/KAFKA-9833
>             Project: Kafka
>          Issue Type: Improvement
>          Components: KafkaConnect
>            Reporter: Brandon Moe
>            Priority: Minor
>   Original Estimate: 2h
>  Remaining Estimate: 2h
>
> Currently in Kafka Connect I believe that if you want to change what file logs are written to you need to include a new log4j .properties file and then override the *KAFKA_LOG4J_OPTS* environment variable. This can be seen here [https://github.com/apache/kafka/blob/trunk/bin/connect-distributed.sh] I believe there is a valid use case for allowing a separate environment variable such as *CONNECT_LOG4J_DIR* or similar to allow a user simply to chose where the logs are written. This makes sense to me in a case where the default .properties file works great but we simply want to change the output location.  An example could be if we are deploying to Kubernetes or similar and want to write logs to a persistent volume. With an environment variable dedicated to this use it saves busy work and extra steps while providing the desired functionality succinctly. In the case when the variable is not set we simply revert to the default settings. I do not believe I am alone on wanting this use case as I have come across several other Stack Overflow and similar posts. 
>  
> I believe this change would be pretty simple to implement and I would like to take a shot at it if this seems legitimate, would love some feedback
>  
> Best,
> Brandon



--
This message was sent by Atlassian Jira
(v8.3.4#803005)