[
https://issues.apache.org/jira/browse/SPARK-12262?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15070503#comment-15070503
]
Xiu (Joe) Guo commented on SPARK-12262:
---------------------------------------
The property {code}spark.sql.hive.convertMetastoreParquet{code} promotes
parquet tables with built-in access instead of going through Hive metastore
route. I am thinking maybe the `describe extended` behavior here is intended?
A workaround (or intended usage) would be:
{code}
set spark.sql.hive.convertMetastoreParquet=false;
{code}
before describing a parquet table.
> describe extended doesn't return table on detail info tabled stored as
> PARQUET format
> -------------------------------------------------------------------------------------
>
> Key: SPARK-12262
> URL: https://issues.apache.org/jira/browse/SPARK-12262
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Affects Versions: 1.5.2
> Reporter: pin_zhang
>
> 1. start hive server with start-thriftserver.sh
> 2. create table table1 (id int) ;
> create table table2(id int) STORED AS PARQUET;
> 3. describe extended table1 ;
> return detailed info
> 4. describe extended table2 ;
> result has no detailed info
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
