[
https://issues.apache.org/jira/browse/PARQUET-2126?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17570545#comment-17570545
]
ASF GitHub Bot commented on PARQUET-2126:
-----------------------------------------
shangxinli commented on PR #959:
URL: https://github.com/apache/parquet-mr/pull/959#issuecomment-1193390004
@theosib-amazon, I am not concerned if release/close isn't called and I
agree the caller must call release/close after finishing. My question is that
before release/close is called, there could be short-living threads that are
used to create the compressor/decompressor in the cache. Those short-living
threads exit and the cache is not aware of that, then that causes the cache
grows with a lot of dead compressor/decompressors. In the scenario where
short-living threads just come and go as a normal business, this could be a
problem. I know normally it is not a problem because in most of the cases we
use thread pool but I am just not sure there is a corner case like that.
Parquet is a low-level library and is used in so many cases.
I am sorry if I didn't make my previous comment more obvious.
> Thread safety bug in CodecFactory
> ---------------------------------
>
> Key: PARQUET-2126
> URL: https://issues.apache.org/jira/browse/PARQUET-2126
> Project: Parquet
> Issue Type: Bug
> Components: parquet-mr
> Affects Versions: 1.12.2
> Reporter: James Turton
> Priority: Major
>
> The code for returning Compressor objects to the caller goes to some lengths
> to achieve thread safety, including keeping Codec objects in an Apache
> Commons pool that has thread-safe borrow semantics. This is all undone by
> the BytesCompressor and BytesDecompressor Maps in
> org.apache.parquet.hadoop.CodecFactory which end up caching single compressor
> and decompressor instances due to code in CodecFactory@getCompressor and
> CodecFactory@getDecompressor. When the caller runs multiple threads, those
> threads end up sharing compressor and decompressor instances.
> For compressors based on Xerial Snappy this bug has no effect because that
> library is itself thread safe. But when BuiltInGzipCompressor from Hadoop is
> selected for the CompressionCodecName.GZIP case, serious problems ensue.
> That class is not thread safe and sharing one instance of it between threads
> produces both silent data corruption and JVM crashes.
> To fix this situation, parquet-mr should stop caching single compressor and
> decompressor instances.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
