You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hive.apache.org by "Raghav Aggarwal (Jira)" <ji...@apache.org> on 2026/09/24 17:49:00 UTC

[jira] [Work started] (HIVE-30094) Use microsecond timestamp precision for Iceberg Z-ordering

     [ https://issues.apache.org/jira/browse/HIVE-30094?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Work on HIVE-30094 started by Raghav Aggarwal.
----------------------------------------------
> Use microsecond timestamp precision for Iceberg Z-ordering
> ----------------------------------------------------------
>
>                 Key: HIVE-30094
>                 URL: https://issues.apache.org/jira/browse/HIVE-30094
>             Project: Hive
>          Issue Type: Bug
>            Reporter: Raghav Aggarwal
>            Assignee: Raghav Aggarwal
>            Priority: Major
>
> As per the [Iceberg Spec|https://iceberg.apache.org/spec/#primitive-types], the {{timestamp}} and {{timestamptz}} primitive types are strictly represented as integers using *microseconds* from the unix epoch.
> However, Hive's ZOrder UDF ({{{}GenericUDFIcebergZorder{}}}) is truncating timestamps to *milliseconds* when computing the Z-Order curve.
> This can cause wrong grouping/clustering.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)