[HIVE-27926] Iceberg: Allow restricting Iceberg data file reads to table location

• Type: Sub-task
• Status: Closed
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 4.0.0-alpha-2
• Fix Version/s: 4.0.0
• Component/s: Iceberg integration
• Labels: pull-request-available
• Assignee: Ayush Saxena
• Reporter: Janos Kovacs
• Created: 01/Dec/23 15:14
• Updated: 29/Mar/24 17:50
• Resolved: 05/Dec/23 05:35

Description

As the first quick solution there should be a configuration flag to allow us to restrict Iceberg reads to data files located only inside of the table locations.

e.g. with the following definition

CREATE EXTERNAL TABLE default.iceloc1 (txt string, secret string)
STORED BY ICEBERG 
LOCATION '/data/hive/external/iceloc1/'
TBLPROPERTIES (
  'external.table.purge'='true',
  'write.metadata.path'='/data/ice/meta/iceloc1/',
  'write.data.path'='/data/ice/data/iceloc1/');

The restricted location should be

/data/hive/external/iceloc1/

Note: this configuration should not be enabled by default as this breaks Iceberg's functionality storing data files in different locations and would only be useful when users use iceberg only as standard external tables with meta+data under table location.

Issue Links

• links to: GitHub Pull Request #4921