linyanghao commented on issue #7618:
URL: https://github.com/apache/iceberg/issues/7618#issuecomment-1582078900
Currently, it's necessary to include partition fields as a part of the pk.
If not, iceberg won't be able to ensure the uniqueness of pk across partitions.
For example: both ('partition1', 'key1'), ('partition2', 'key1') can exist in a
table in the same time, which breaks the uniqueness constraint.
To support not including partition fields as pk. We need to implement a
global index like hudi's. Given a record's pk, the global index routes the
record to the partition containing it. In this way, we can ensure that the pk
is unique.
--
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]
