[
https://issues.apache.org/jira/browse/SPARK-6016?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Yin Huai updated SPARK-6016:
----------------------------
Description:
saveAsTable is fine and seems we have successfully deleted the old data and
written the new data. However, when reading the newly created table, an error
will be thrown.
{code}
Error in SQL statement: java.lang.RuntimeException: java.lang.RuntimeException:
could not merge metadata: key org.apache.spark.sql.parquet.row.metadata has
conflicting values:
at parquet.hadoop.api.InitContext.getMergedKeyValueMetaData(InitContext.java:67)
at parquet.hadoop.api.ReadSupport.init(ReadSupport.java:84)
at
org.apache.spark.sql.parquet.FilteringParquetRowInputFormat.getSplits(ParquetTableOperations.scala:469)
at
parquet.hadoop.ParquetInputFormat.getSplits(ParquetInputFormat.java:245)
at
org.apache.spark.sql.parquet.ParquetRelation2$$anon$1.getPartitions(newParquet.scala:461)
...
{code}
If I set spark.sql.parquet.cacheMetadata to false, it's fine to query the data.
Note: the newly created table needs to have more than one file to trigger the
bug (if there is only a single file, we will not need to merge metadata).
was:
saveAsTable is fine and seems we have successfully deleted the old data and
written the new data. However, when reading the newly created table, an error
will be thrown.
{code}
Error in SQL statement: java.lang.RuntimeException: java.lang.RuntimeException:
could not merge metadata: key org.apache.spark.sql.parquet.row.metadata has
conflicting values:
at parquet.hadoop.api.InitContext.getMergedKeyValueMetaData(InitContext.java:67)
at parquet.hadoop.api.ReadSupport.init(ReadSupport.java:84)
at
org.apache.spark.sql.parquet.FilteringParquetRowInputFormat.getSplits(ParquetTableOperations.scala:469)
at
parquet.hadoop.ParquetInputFormat.getSplits(ParquetInputFormat.java:245)
at
org.apache.spark.sql.parquet.ParquetRelation2$$anon$1.getPartitions(newParquet.scala:461)
...
{code}
If I set spark.sql.parquet.cacheMetadata to false, it's fine to query the data.
> Cannot read the parquet table after overwriting the existing table when
> spark.sql.parquet.cacheMetadata=true
> ------------------------------------------------------------------------------------------------------------
>
> Key: SPARK-6016
> URL: https://issues.apache.org/jira/browse/SPARK-6016
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Reporter: Yin Huai
> Priority: Blocker
>
> saveAsTable is fine and seems we have successfully deleted the old data and
> written the new data. However, when reading the newly created table, an error
> will be thrown.
> {code}
> Error in SQL statement: java.lang.RuntimeException:
> java.lang.RuntimeException: could not merge metadata: key
> org.apache.spark.sql.parquet.row.metadata has conflicting values:
> at
> parquet.hadoop.api.InitContext.getMergedKeyValueMetaData(InitContext.java:67)
> at parquet.hadoop.api.ReadSupport.init(ReadSupport.java:84)
> at
> org.apache.spark.sql.parquet.FilteringParquetRowInputFormat.getSplits(ParquetTableOperations.scala:469)
> at
> parquet.hadoop.ParquetInputFormat.getSplits(ParquetInputFormat.java:245)
> at
> org.apache.spark.sql.parquet.ParquetRelation2$$anon$1.getPartitions(newParquet.scala:461)
> ...
> {code}
> If I set spark.sql.parquet.cacheMetadata to false, it's fine to query the
> data.
> Note: the newly created table needs to have more than one file to trigger the
> bug (if there is only a single file, we will not need to merge metadata).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
