C-Loftus commented on issue #1077:
URL: https://github.com/apache/iceberg-go/issues/1077#issuecomment-4445086098
Thank you very much @tanmayrauth and @laskoviymishka that is very useful
information and really helps!
I have two follow ups if you wouldn't mind:
1. Am I correct that iceberg-go exposes one delete API, `tbl.Delete` and
that whether it is an equality delete or a position delete is not necessarily
exposed to the client?
- Or does `tbl.Delete` always produce position delete files if
`write.delete.mode=merge-on-read` is set?
2. Is it correct that the presence of position-delete files?
- Namely, in the example above, I delete all subjects with `foo` with
`tbl.Delete`. If I then run `tbl.Scan` for `foo` is it correct that this could
still result in a full table scan? i.e. since the position delete files just
reduces the amount of rows, not anything about the actual queries. It would be
up to the parquet row group statistics to potentially tell the query engine to
skip entire row group scans?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
