You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Rui Li (Jira)" <ji...@apache.org> on 2020/02/21 09:21:00 UTC
[jira] [Updated] (FLINK-16197) Failed to query partitioned table
when partition folder is removed
[ https://issues.apache.org/jira/browse/FLINK-16197?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Rui Li updated FLINK-16197:
---------------------------
Description:
If a partition exists in HMS but not in HDFS, the query fails with
{noformat}
Caused by: org.apache.hadoop.mapred.InvalidInputException: Input path does not exist: hdfs://...........
at org.apache.hadoop.mapred.LocatedFileStatusFetcher.getFileStatuses(LocatedFileStatusFetcher.java:155)
at org.apache.hadoop.mapred.FileInputFormat.listStatus(FileInputFormat.java:237)
at org.apache.hadoop.mapred.FileInputFormat.getSplits(FileInputFormat.java:315)
at org.apache.flink.connectors.hive.read.HiveTableInputFormat.createInputSplits(HiveTableInputFormat.java:219)
at org.apache.flink.connectors.hive.HiveTableSource.getDataStream(HiveTableSource.java:152)
{noformat}
Expected behavior is to run the query ignoring this partition.
> Failed to query partitioned table when partition folder is removed
> ------------------------------------------------------------------
>
> Key: FLINK-16197
> URL: https://issues.apache.org/jira/browse/FLINK-16197
> Project: Flink
> Issue Type: Bug
> Components: Connectors / Hive
> Reporter: Rui Li
> Priority: Major
>
> If a partition exists in HMS but not in HDFS, the query fails with
> {noformat}
> Caused by: org.apache.hadoop.mapred.InvalidInputException: Input path does not exist: hdfs://...........
> at org.apache.hadoop.mapred.LocatedFileStatusFetcher.getFileStatuses(LocatedFileStatusFetcher.java:155)
> at org.apache.hadoop.mapred.FileInputFormat.listStatus(FileInputFormat.java:237)
> at org.apache.hadoop.mapred.FileInputFormat.getSplits(FileInputFormat.java:315)
> at org.apache.flink.connectors.hive.read.HiveTableInputFormat.createInputSplits(HiveTableInputFormat.java:219)
> at org.apache.flink.connectors.hive.HiveTableSource.getDataStream(HiveTableSource.java:152)
> {noformat}
> Expected behavior is to run the query ignoring this partition.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)