CDH 5.2.6 Release Notes
The following lists all Columnar storage format for Hadoop Jiras included in CDH 5.2.6
that are not included in the Columnar storage format for Hadoop base version 1.5.0. The
file lists all changes included in CDH 5.2.6. The patch for each
change can be found in the cloudera/patches directory in the release tarball.
Changes Not In Columnar storage format for Hadoop 1.5.0
- [PARQUET-145] - InternalParquetRecordReader.close() should not throw an exception if initialization has failed
- [PARQUET-107] - Add option to disable summary metadata aggregation after MR jobs
- [PARQUET-82] - ColumnChunkPageWriteStore assumes pages are smaller than Integer.MAX_VALUE
- [PARQUET-63] - Fixed-length columns cannot be dictionary encoded.
- [PARQUET-75] - String decode using 'new String' is slow
- [PARQUET-80] - upgrade semver plugin version to 0.9.27
- [PARQUET-66] - InternalParquetRecordWriter int overflow causes unnecessary memory check warning
- [PARQUET-62] - DictionaryValuesWriter dictionaries are corrupted by user changes.
- [PARQUET-13] - The `-d` option for `parquet-schema` shouldn't have optional argument
- [PARQUET-18] - Cannot read dictionary-encoded pages with all null values
- [PARQUET-4] - Use LRU caching for footers in ParquetInputFormat.
- [PARQUET-9] - InternalParquetRecordReader will not read multiple blocks when filtering
- [PARQUET-59] - Scrooge tests use hadoop-1 API, fail on hadoop-2
- [PARQUET-140] - Allow clients to control the GenericData object that is used to read Avro records
- [PARQUET-92] - Parallel Footer Read Control
- [PARQUET-25] - Pushdown predicates only work with hardcoded arguments