[
https://issues.apache.org/jira/browse/PARQUET-2159?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Fang-Xie updated PARQUET-2159:
------------------------------
Description:
Current Spark use Parquet-mr as parquet reader/writer library, but the built-in
bit-packing en/decode is not efficient enough.
Our optimization for Parquet bit-packing en/decode with jdk.incubator.vector in
Open JDK18 brings prominent performance improvement.
Due to Vector API is added to OpenJDK since 16, So this optimization request
JDK16 or higher.
*Below are our test results*
Functional test is based on open-source parquet-mr Bit-pack decoding function:
*_public final void unpack8Values(final byte[] in, final int inPos, final int[]
out, final int outPos)_* __
compared with our implementation with vector API *_public final void
unpack8Values_vec(final byte[] in, final int inPos, final int[] out, final int
outPos)_*
We tested 10 pairs (open source parquet bit unpacking vs ours optimized
vectorized SIMD implementation) decode function with bit
width=\{1,2,3,4,5,6,7,8,9,10}, below are test results:
!image-2022-06-15-22-56-08-396.png|width=437,height=223!
We integrated our bit-packing decode implementation into parquet-mr, tested the
parquet batch reader ability from Spark VectorizedParquetRecordReader which get
parquet column data by the batch way. We construct parquet file with different
row count and column count, the column data type is Int32, the maximum int
value is 127 which satisfies bit pack encode with bit width=7, the count of
the row is from 10k to 100 million and the count of the column is from 1 to 4.
!image-2022-06-15-22-57-15-964.png|width=453,height=229!
!image-2022-06-15-22-58-01-442.png|width=439,height=217!
!image-2022-06-15-22-58-40-704.png|width=415,height=208!
was:Here provide a new way of optimized bit-packing en/decode algorithm
> Parquet bit-packing de/encode optimization
> ------------------------------------------
>
> Key: PARQUET-2159
> URL: https://issues.apache.org/jira/browse/PARQUET-2159
> Project: Parquet
> Issue Type: Improvement
> Components: parquet-mr
> Affects Versions: 1.13.0
> Reporter: Fang-Xie
> Priority: Major
> Fix For: 1.13.0
>
> Attachments: image-2022-06-15-22-56-08-396.png,
> image-2022-06-15-22-57-15-964.png, image-2022-06-15-22-58-01-442.png,
> image-2022-06-15-22-58-40-704.png
>
>
> Current Spark use Parquet-mr as parquet reader/writer library, but the
> built-in bit-packing en/decode is not efficient enough.
> Our optimization for Parquet bit-packing en/decode with jdk.incubator.vector
> in Open JDK18 brings prominent performance improvement.
> Due to Vector API is added to OpenJDK since 16, So this optimization request
> JDK16 or higher.
> *Below are our test results*
> Functional test is based on open-source parquet-mr Bit-pack decoding
> function: *_public final void unpack8Values(final byte[] in, final int inPos,
> final int[] out, final int outPos)_* __
> compared with our implementation with vector API *_public final void
> unpack8Values_vec(final byte[] in, final int inPos, final int[] out, final
> int outPos)_*
> We tested 10 pairs (open source parquet bit unpacking vs ours optimized
> vectorized SIMD implementation) decode function with bit
> width=\{1,2,3,4,5,6,7,8,9,10}, below are test results:
> !image-2022-06-15-22-56-08-396.png|width=437,height=223!
> We integrated our bit-packing decode implementation into parquet-mr, tested
> the parquet batch reader ability from Spark VectorizedParquetRecordReader
> which get parquet column data by the batch way. We construct parquet file
> with different row count and column count, the column data type is Int32, the
> maximum int value is 127 which satisfies bit pack encode with bit width=7,
> the count of the row is from 10k to 100 million and the count of the column
> is from 1 to 4.
> !image-2022-06-15-22-57-15-964.png|width=453,height=229!
> !image-2022-06-15-22-58-01-442.png|width=439,height=217!
> !image-2022-06-15-22-58-40-704.png|width=415,height=208!
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
