[
https://issues.apache.org/jira/browse/HDDS-4553?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Prashant Pogde updated HDDS-4553:
---------------------------------
Target Version/s: 1.2.0 (was: 1.1.0)
> ChunkInputStream should release buffer as soon as last byte in the buffer is
> read
> ---------------------------------------------------------------------------------
>
> Key: HDDS-4553
> URL: https://issues.apache.org/jira/browse/HDDS-4553
> Project: Hadoop Distributed Data Store
> Issue Type: Improvement
> Reporter: Shashikant Banerjee
> Assignee: Hanisha Koneru
> Priority: Major
>
> We currently wait for reading up till the Chunk EOF before releasing the
> buffers in ChunkInputStream (or when the stream is closed). Let's say a
> client reads first 3 MB of a chunk of size 4MB and does not close the stream
> immediately. This would lead to the 3MB of data being cached in the
> ChunkInputStream buffers till the stream is closed.
> Once HDDS-4552 is resolved, a chunk read from DN would return the chunk data
> as an array of ByteBuffers. After each ByteBuffer is read, it can be
> released. This would greatly help with optimizing memory usage of
> ChunkInputStream.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
