[HDDS-3748] Consider reusing bytebuffer in FilePerBlockStrategy::readChunk

• Type: Improvement
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Labels: performance
• Assignee: Unassigned
• Reporter: Rajesh Balamohan
• Created: 08/Jun/20 04:52
• Updated: 07/Sep/22 00:25

Description

https://github.com/apache/hadoop-ozone/blob/master/hadoop-hdds/container-service/src/main/java/org/apache/hadoop/ozone/container/keyvalue/impl/FilePerBlockStrategy.java#L148

 

            long len = info.getLen();            
            long offset = info.getOffset();
            ByteBuffer data = ByteBuffer.allocate((int) len);
            ChunkUtils.readData(chunkFile, data, offset, len, volumeIOStats); 

 

Instead of allocating buffer in every readChunk, it may be possible to reuse via threadLocal ByteBuffer. Sizes/Limits could be adjusted on need basis.

This is to reduce memory pressure on DN.

Issue Links

• is related to: HDDS-7117 Consider reading chunk files using MappedByteBuffer