You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@drill.apache.org by "cdmikechen (Jira)" <ji...@apache.org> on 2022/01/20 00:40:00 UTC

[jira] [Created] (DRILL-8109) Drill can not read parquet timestamp type with logical type in hive storage plugin

cdmikechen created DRILL-8109:
---------------------------------

             Summary: Drill can not read parquet timestamp type with logical type in hive storage plugin
                 Key: DRILL-8109
                 URL: https://issues.apache.org/jira/browse/DRILL-8109
             Project: Apache Drill
          Issue Type: Bug
          Components: Storage - Hive
    Affects Versions: 1.19.0
         Environment: drill 1.19.0 
hive 3.12
            Reporter: cdmikechen
             Fix For: 1.20.0


Parquet timestamp int96 had been deprecated.
https://github.com/apache/parquet-format/pull/86/files

Many computing engines that support parquet reading (such as Hive4, Hudi, Spark and so on) can use Int64 as the time storage type. When using these computing engines to read parquet files, we should also support long data type reading.

We should support int64 with logical type timestamp by hive storage.



--
This message was sent by Atlassian Jira
(v8.20.1#820001)