[ https://issues.apache.org/jira/browse/HADOOP-18887?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17764788#comment-17764788 ]
Steve Loughran commented on HADOOP-18887: ----------------------------------------- # happy to watch this, but not got time to review anything # changes like jar updates are best isolated into their own jiras -maybe create them under the big "java17" jira but declare as a requirement for here. just knowing what needs fixing is a good first step; launching mr jobs with the right --add-opens stuff for every process launched. # some of my colleagues have been busy with java17; they may have some input here. > Java 17 Runtime Support > ----------------------- > > Key: HADOOP-18887 > URL: https://issues.apache.org/jira/browse/HADOOP-18887 > Project: Hadoop Common > Issue Type: New Feature > Components: build > Reporter: Bruno Ramirez > Priority: Major > Original Estimate: 336h > Remaining Estimate: 336h > > This JIRA feature aims to extend the Java runtime support for Hadoop to > include Java 17 in addition to Java 8 and Java 11. Currently Hadoop version > 3.3.6 supports Java 8 compile and Java 8/11 runtime. The goal is to make > Hadoop compatible with Java 17 runtime as well. > The plan for this release is to allow Hadoop to default to Java 11/Java 17, > while still providing the flexibility for customers to configure Hadoop to > use Java 8, Java 11, or Java 17 based on their specific needs. This project's > objectives include: > # Certifying that Hadoop works seamlessly on Java 8/11/17 for common use > cases. > # Ensuring that running Hadoop on Java 11/17 does not disrupt other > applications and libraries such as Spark, Hive, Flink, Presto/Trino, and > HBase. > The decision to support Java 17 runtime is motivated by customer requests and > significant performance improvements observed in downstream applications like > Apache Hive and Apache Spark. The testing process encompasses unit tests, > integration tests, and performance tests, as well as verifying the proper > functioning of all Hadoop daemons with Java 17. > The project will address compile time issues across various Hadoop > components, ensuring that Hadoop remains compatible with Java 17 throughout > the entire codebase. > This ticket serves as a vital step in enhancing Hadoop's capabilities, > providing customers with more choices and improved performance for their big > data processing needs. -- This message was sent by Atlassian Jira (v8.20.10#820010) --------------------------------------------------------------------- To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org