pvary

cc: @zentol: FYI - backporting your fix to release-1.15 branch

@flinkbot run azure

@szlta, @lcspinter any thoughts?

Incremental read is fine when you have append only tables, but starts to get tricky for tables with deletes. @marton-bod did some experimentation with it in Hive, but we did...

@igorcalabria: Let’s wait for the Spark team members to chime in then.

Thanks for the PR @sumeetgajjar !

Thanks for the fix @sumeetgajjar and @wypoon and @samredai for the review!

@linfey90: I think this is by design. The split size should be considered even if the target size is defined. Thanks, Peter

@rdblue: We would like to use relative path for replication of the Iceberg tables between on-prem Hive instances, or moving on-prem Iceberg tables to the cloud and sometimes even moving...

> @pvary, we can explore that, but it is fairly easy to write a new metadata file and use that for a table. At least that's a small operation, compared...