That sounds strange. When my Hadoop 3.3.4 uses Java 8 then m/r jobs don't work so Hive throw an exception. After I change to Java 11 for Hadoop they work right now. That's to say, I am using Java 11 for Hadoop 3.3.4, and Java 8 for Hive 3.1, on the same node. Regards -----Original-Nachricht----- Betreff: Re: Make Hive 3.1 and Hadoop 3.3 work in the same node Datum: 2022-12-24T15:54:32+0100 Von: "Ayush Saxena" <ayush...@gmail.com> An: "user@hive.apache.org" <user@hive.apache.org> While most other Hadoop ecosystem including Hadoop 3.3, Spark 3.3, Kafka 3.3 seem require java 11. Hadoop 3.3.x line just supports Java-11 runtime support, the default supported JDK is still 8, atleast for hadoop, if it wasn't working with JDK-8, some issue with your hadoop installation or some bug there in the hadoop code, which you need to fix or reach out to hadoop ML/Jira -Ayush
On Sat, 24 Dec 2022 at 11:26, yp...@t-online.de <mailto:yp...@t-online.de> <yp...@t-online.de <mailto:yp...@t-online.de> > wrote: Hello, Hive 3.1 requires java 8 to work. While most other Hadoop ecosystem including Hadoop 3.3, Spark 3.3, Kafka 3.3 seem require java 11. I have all of them installed in the same system to work together. This is how I did, <https://blog.crypt.pw/How-to-run-Hadoop-3.3-and-Hive-3.1-in-same-node> Hope it also helps to you too. Thanks.
