[
https://issues.apache.org/jira/browse/HDFS-9040?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14738730#comment-14738730
]
Walter Su commented on HDFS-9040:
---------------------------------
bq. The role of BlockGroupDataStreamer looks similar (or closely related) to
Coordinator. Should we make it part of Coordinator?
it's not like that. The old way requires swtiching streamers. With
BGDataStreamer, the data flow becomes
DFSStripedOutputStream --> BGDataStreamer --> StripedDataStreamer(s)
bq. I think a challenge is whether / how to reuse the existing DFSOutputStream
/ DataStreamer code. It doesn't seem easy to avoid refactoring them.
bq. Implementing a block group writer is a better solution but you have to copy
a lot of existing code to the new class, which may be hard to be...
Thanks for sharing the information. I'll try my best to avoid that. I'm still
working on this. Will update a demo soon.
> Erasure coding: A BlockGroupDataStreamer to rule all internal blocks streamers
> ------------------------------------------------------------------------------
>
> Key: HDFS-9040
> URL: https://issues.apache.org/jira/browse/HDFS-9040
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Walter Su
> Assignee: Walter Su
>
> A {{BlockGroupDataStreamer}} to communicate with NN to allocate/update block,
> and {{StripedDataStreamer}} s only have to stream blocks to DNs.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
