Hello, everyone! We have recently discovered that Apache Orc support is unfortunately broken in Spark 3.x modules of Apache Hudi, due to the fact that Spark 3.x switched from "nohive <https://repo1.maven.org/maven2/org/apache/orc/orc-core/1.8.0/>" flavor of Apache Orc to conventional one (which brings in some common Hive-originating interfaces). More details could on that be found in HUDI-4496 <https://issues.apache.org/jira/browse/HUDI-4496>
As such, we're trying to understand a little more around usage of Apache Orc and Hudi, and will be very grateful to everyone currently having such setup and responding back on this thread w/ the answers to following questions: - What Spark version are you using? - What Hudi version are you using? - Are you considering upgrading to Spark 3.x? - Are you considering upgrading to the latest Hudi release(s)? As always, appreciate your help and look forward to hearing back from you.
