[ https://issues.apache.org/jira/browse/HDFS-8019?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14528210#comment-14528210 ]
Vinayakumar B commented on HDFS-8019: ------------------------------------- for client side erasure encoding, chunk buffers are obtained from {{ByteArrayManager}} provided by the DFSClient. This will allow client to re-use the byte arrays (onheap) if the byte arrays are having length more than threshold \[128 default\]. This was introduced in HDFS-7276. [~drankye] and [~szetszwo], Do you think re-using {{ByteArrayManager}} in datanode side erasure recovery would be sufficient for this? > Erasure Coding: erasure coding chunk buffer allocation and management > --------------------------------------------------------------------- > > Key: HDFS-8019 > URL: https://issues.apache.org/jira/browse/HDFS-8019 > Project: Hadoop HDFS > Issue Type: Sub-task > Reporter: Kai Zheng > Assignee: Vinayakumar B > > As a task of HDFS-7344, this is to come up a chunk buffer pool allocating and > managing coding chunk buffers, either based on on-heap or off-heap. Note this > assumes some DataNodes are powerful in computing and performing EC coding > work, so better to have this dedicated buffer pool and management. -- This message was sent by Atlassian JIRA (v6.3.4#6332)