[
https://issues.apache.org/jira/browse/HDFS-6934?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14158477#comment-14158477
]
Tsz Wo Nicholas Sze commented on HDFS-6934:
-------------------------------------------
Currently, client computes checksum and sends it to the datanodes. Datanode
verifies the checksum and then persists it with the block data.
In this JIRA, we propose that client does not compute checksum anymore for
in-memory replica in order to give a better performance. The checksum will be
computed when the datanode lazy persists the replica to disk.
> Move checksum computation off the hot path when writing to RAM disk
> -------------------------------------------------------------------
>
> Key: HDFS-6934
> URL: https://issues.apache.org/jira/browse/HDFS-6934
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: datanode
> Reporter: Arpit Agarwal
> Assignee: Tsz Wo Nicholas Sze
>
> Since local RAM is considered reliable we can avoid writing checksums on the
> hot path when replicas are being written to a local RAM disk.
> The checksum can be computed by the lazy writer when moving replicas to disk.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
