[
https://issues.apache.org/jira/browse/HIVE-28818?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Shohei Okumiya updated HIVE-28818:
----------------------------------
Summary: Snapshot-level Partition-Aware Optimization (was: Partition-Aware
Optimization with Iceberg's time travel)
> Snapshot-level Partition-Aware Optimization
> -------------------------------------------
>
> Key: HIVE-28818
> URL: https://issues.apache.org/jira/browse/HIVE-28818
> Project: Hive
> Issue Type: Sub-task
> Components: Iceberg integration
> Reporter: Shohei Okumiya
> Priority: Major
>
> The current implementation is triggered when the "metadata.json" of an
> Iceberg table retains exactly one partition spec. We introduced the
> limitation because it is not trivial to apply Bucket Map Join when partition
> specifications have evolved.
> However, the restriction is too strict because the "metadata.json" is
> top-level metadata that includes the partition specs of all snapshots. This
> ticket would gather partition specs tied to only a related snapshot.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
