[HIVE-26161] Use Hive's ORC dependency version when producing file footer for Iceberg

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 4.0.0-alpha-2
• Component/s: Iceberg integration
• Labels: pull-request-available
• Assignee: Ádám Szita
• Reporter: Ádám Szita
• Created: 21/Apr/22 09:22
• Updated: 16/Nov/22 13:50
• Resolved: 21/Apr/22 18:03

Description

For schema evolution and projection purposes we produce an ORC file footer byte buffer in VectorizedReadUtils.

Currently Iceberg's bundled/shaded ORC is used to produce these file footer bytes when dealing with Iceberg/ORC tables. This version of ORC is newer (1.7.3) than what Hive uses (1.6.9).

Later on we could face compatibility issues when trying to reconstruct an OrcTail object with a 1.6.9 reader from the bytes that the 1.7.3 reader serialized. We need to invert the direction as we can rely more on backward compatibility than on forward compatibility of ORC.

Issue Links

• links to: GitHub Pull Request #3230