[
https://issues.apache.org/jira/browse/PARQUET-2196?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17610639#comment-17610639
]
ASF GitHub Bot commented on PARQUET-2196:
-----------------------------------------
wgtmac commented on PR #1000:
URL: https://github.com/apache/parquet-mr/pull/1000#issuecomment-1261133491
@shangxinli Thanks for the review.
I haven't be able to run the benchmark since the Hadoop Lz4Codec is
implemented via JNI and I cannot simply run it on my Mac M1 laptop.
Generally this is the same LZ4 algorithm with the one shipped by Hadoop. The
key difference is that LZ4_RAW has removed the redundant block header
introduced by Hadoop Lz4Codec and its implementation is in pure Java.
I will add the interop test in a follow-up commit.
> Support LZ4_RAW codec
> ---------------------
>
> Key: PARQUET-2196
> URL: https://issues.apache.org/jira/browse/PARQUET-2196
> Project: Parquet
> Issue Type: Improvement
> Components: parquet-mr
> Reporter: Gang Wu
> Priority: Major
>
> There is a long history about the LZ4 interoperability of parquet files
> between parquet-mr and parquet-cpp (which is now in the Apache Arrow).
> Attached links are the evidence. In short, a new LZ4_RAW codec type has been
> introduced since parquet format v2.9.0. However, only parquet-cpp supports
> LZ4_RAW. The parquet-mr library still uses the old Hadoop-provided LZ4 codec
> and cannot read parquet files with LZ4_RAW.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
