Rajesh Balamohan created HIVE-26913:
---------------------------------------

             Summary: HiveVectorizedReader::parquetRecordReader should reuse 
footer information
                 Key: HIVE-26913
                 URL: https://issues.apache.org/jira/browse/HIVE-26913
             Project: Hive
          Issue Type: Improvement
          Components: Iceberg integration
            Reporter: Rajesh Balamohan


HiveVectorizedReader::parquetRecordReader should reuse details of parquet 
footer, instead of reading it again.

 

It reads parquet footer here:

[https://github.com/apache/hive/blob/master/iceberg/iceberg-handler/src/main/java/org/apache/iceberg/mr/hive/vector/HiveVectorizedReader.java#L230-L232]

Again it reads the footer here for constructing vectorized recordreader

[https://github.com/apache/hive/blob/master/iceberg/iceberg-handler/src/main/java/org/apache/iceberg/mr/hive/vector/HiveVectorizedReader.java#L249]

 

[https://github.com/apache/hive/blob/master/ql/src/java/org/apache/hadoop/hive/ql/io/parquet/VectorizedParquetInputFormat.java#L50]

 

Check the codepath of 
VectorizedParquetRecordReader::setupMetadataAndParquetSplit

[https://github.com/apache/hive/blob/6b0139188aba6a95808c8d1bec63a651ec9e4bdc/ql/src/java/org/apache/hadoop/hive/ql/io/parquet/vector/VectorizedParquetRecordReader.java#L180]

 

It should be possible to share "ParquetMetadata" in 
VectorizedParuqetRecordReader.

 



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to