dongjoon-hyun commented on a change in pull request #25443: [WIP][SPARK-28723][test-hadoop3.2][test-maven] Test JDK 11 with Hadoop-3.2/Hive 2.3.6 on jenkins URL: https://github.com/apache/spark/pull/25443#discussion_r314931221
########## File path: dev/deps/spark-deps-hadoop-2.7 ########## @@ -95,12 +95,15 @@ jackson-module-jaxb-annotations-2.9.9.jar jackson-module-paranamer-2.9.9.jar jackson-module-scala_2.12-2.9.9.jar jackson-xc-1.9.13.jar +jakarta.activation-1.2.1.jar +jakarta.activation-api-1.2.1.jar Review comment: @srowen . It seems that we found that our library have difference dependency on JDK8 and JDK11. On my Mac, I also noticed that this `jakarta.activation-1.2.1` occurs only on JDK11 environment. I'm not sure how this library does this, but in this case, is it possible for Apache Spark to provide a single artifact for JDK8/JDK11? It seems that `Maven` artifacts might be possible, but the pre-built Apache Spark binary should be different due to this from now. Or, we need to use always JDK11 for updating manifests to build the most inclusive jar list. cc @dbtsai , @gatorsmile ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
