Aäron created SPARK-26663:
-----------------------------
Summary: Cannot query a Hive table with orc files in subdirectories
Key: SPARK-26663
URL: https://issues.apache.org/jira/browse/SPARK-26663
Project: Spark
Issue Type: Bug
Components: SQL
Affects Versions: 2.4.0
Reporter: Aäron
Hello,
I want to report the following issue (my first one :) )
When I create a table in Hive based on a union all then Spark 2.4 is unable to
query this table.
To reproduce:
*Hive 1.2.1*
{code:java}
hive> creat table a(id int);
insert into a values(1);
hive> creat table b(id int);
insert into b values(2);
hive> create table c(id int) as select id from a union all select id from b;
{code}
*Spark 2.3.1*
{code:java}
scala> spark.table("c").show
+---+
| id|
+---+
| 1|
| 2|
+---+
scala> spark.table("c").count
res5: Long = 2
{code}
*Spark 2.4.0*
{code:java}
scala> spark.table("c").show
19/01/18 17:00:49 WARN HiveMetastoreCatalog: Unable to infer schema for table
perftest_be.c from file format ORC (inference mode: INFER_AND_SAVE). Using
metastore schema.
+---+
| id|
+---+
+---+
scala> spark.table("c").count
res3: Long = 0
{code}
I did not find an existing issue for this. Might be important to investigate.
Kind regards.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
