HDDS-14938. Implement Iceberg's RewriteTablePath action with input validation

[sreejasahithi]

What changes were proposed in this pull request?

This patch implements Iceberg's RewriteTablePath action and validates the inputs provided for the path rewrite.

Prefix Rewrite Configuration
Allow specifying:

• sourcePrefix (existing path prefix)
• targetPrefix (replacement prefix)
• Ensure source and target prefixes are non-empty and not identical

Optional startVersion and endVersion inputs.

• Validate that provided version exist in table.
• Defaults:
  • endVersion → current metadata version if not provided
  • startVersion → if not set, rewrite applies up to end version from beginning version.

Staging Location Handling

• Supports user-defined staging directory.
• If not provided, automatically generates a staging path within the tables's metadata directory.

This change is required for upcoming patches in the epic: : HDDS-14937

What is the link to the Apache JIRA

HDDS-14938

How was this patch tested?

Green CI : https://github.com/sreejasahithi/ozone/actions/runs/24018893628

[adoroszlai]

Thanks @sreejasahithi for the patch.

[adoroszlai]

Thanks @sreejasahithi for updating the patch.

[adoroszlai]

Thanks @sreejasahithi for updating the patch, LGTM.

[ashishkumar50]

@sreejasahithi Thanks for working on this, Can we put this into new module something like ozone-iceberg/tools instead of placing inside hadoop-ozone.

[adoroszlai]

Can we put this into new module something like ozone-iceberg/tools instead of placing inside hadoop-ozone.

That would not fit existing build setup, why would we do that?

[ashishkumar50]

Can we put this into new module something like ozone-iceberg/tools instead of placing inside hadoop-ozone.

That would not fit existing build setup, why would we do that?

Since this is specific just for iceberg ozone integration and not directly related to ozone module, so i was thinking not to put inside hadoop-ozone. If it doesn't fit in existing build setup then it is fine to put inside hadoop-ozone then.

[sreejasahithi]

Can we put this into new module something like ozone-iceberg/tools instead of placing inside hadoop-ozone.

That would not fit existing build setup, why would we do that?

Since this is specific just for iceberg ozone integration and not directly related to ozone module, so i was thinking not to put inside hadoop-ozone. If it doesn't fit in existing build setup then it is fine to put inside hadoop-ozone then.

If we want to move the module outside hadoop-ozone then we would have to change the existing hadoop-ozone and root pom.xml and also add in 'hadoop-ozone/dist/src/main/assemblies/ozone-src.xml'.
Do we want to proceed with this or keep it in hadoop-ozone?
@adoroszlai and @ashishkumar50

[adoroszlai]

we would have to change the existing hadoop-ozone and root pom.xml and also add in 'hadoop-ozone/dist/src/main/assemblies/ozone-src.xml'.

At least the following scripts would also need to be tweaked:

• dev-support/ci/selective_ci_checks.sh and its test
• hadoop-ozone/dev-support/checks/coverage.sh
• hadoop-ozone/dev-support/checks/findbugs.sh

I don't see any reason to do increase the scope of this change with that.