[
https://issues.apache.org/jira/browse/HDDS-9903?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17795493#comment-17795493
]
Tsz-wo Sze commented on HDDS-9903:
----------------------------------
EC could use Streaming Write Pipeline (HDDS-4454) to write data.
> EC data write as data streams
> -----------------------------
>
> Key: HDDS-9903
> URL: https://issues.apache.org/jira/browse/HDDS-9903
> Project: Apache Ozone
> Issue Type: Improvement
> Components: EC, EC Client, Ozone Datanode
> Reporter: Duong
> Priority: Major
> Labels: ozone-performance
>
> Today, EC write is done in a discrete fashion via GRPC where each chunk is
> modeled as a WriteChunk request. Because of that, we have to implement a big
> chunk size to archive stream-like performance, and other unnatural
> improvements like ZeroCopy on GRPC.
> GRPC and discrete data transfer protocol is not the perfect fit for data
> transfer. We should move to use native stream, leverage Netty, to implement
> EC writes.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
