[
https://issues.apache.org/jira/browse/HIVE-18323?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16317297#comment-16317297
]
Vihang Karajgaonkar commented on HIVE-18323:
--------------------------------------------
Thanks [~aihuaxu] for taking this up. Can you please add some test cases?
Vectorizer expects all the primitive to be vectorized if a input format
implements VectorizedInputFormatInterface. This means any parquet table which
has timestamp or {{INTERVAL_DAY_TIME}} will fail to execute in vectorized mode.
I think we should fix this soon. I will create another JIRA for supporting
interval_day_time.
I took a quick look. The isRepeating flag is always set to false in the patch.
Can we change line 330 to something like below?
{noformat}
c.isRepeating = c.isRepeating && (c.vector[0] == c.vector[rowId]);
{noformat}
Would be good if [~Ferd] also takes a look at this one.
> Vectorization: add the support of timestamp in
> VectorizedPrimitiveColumnReader for parquet
> ------------------------------------------------------------------------------------------
>
> Key: HIVE-18323
> URL: https://issues.apache.org/jira/browse/HIVE-18323
> Project: Hive
> Issue Type: Improvement
> Components: Vectorization
> Affects Versions: 3.0.0
> Reporter: Aihua Xu
> Assignee: Aihua Xu
> Attachments: HIVE-18323.1.patch
>
>
> {noformat}
> CREATE TABLE `t1`(
> `ts` timestamp,
> `s1` string)
> STORED AS PARQUET;
> set hive.vectorized.execution.enabled=true;
> SELECT * from t1 SORT BY s1;
> {noformat}
> This query will throw exception since timestamp is not supported here yet.
> {noformat}
> Caused by: java.io.IOException: java.io.IOException: Unsupported type:
> optional int96 ts
> at
> org.apache.hadoop.hive.io.HiveIOExceptionHandlerChain.handleRecordReaderNextException(HiveIOExceptionHandlerChain.java:121)
> at
> org.apache.hadoop.hive.io.HiveIOExceptionHandlerUtil.handleRecordReaderNextException(HiveIOExceptionHandlerUtil.java:77)
> at
> org.apache.hadoop.hive.ql.io.HiveContextAwareRecordReader.doNext(HiveContextAwareRecordReader.java:365)
> at
> org.apache.hadoop.hive.ql.io.CombineHiveRecordReader.doNext(CombineHiveRecordReader.java:116)
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
