[
https://issues.apache.org/jira/browse/HDFS-7716?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14306789#comment-14306789
]
Jing Zhao commented on HDFS-7716:
---------------------------------
Thanks for the comments, Zhe! I'm currently still working on a patch. For your
first two questions, I still plan to use triplets since currently the block
report processing logic depends on this data structure. The difference between
{{BlockReplicationInfo}} and {{BlockGroupInfo}} mainly lies in that
{{BlockGroupInfo}} requires the triplets to be sorted based on the block
sequence in the group. Thus the {{addStorage}}, {{removeStorage}} and
{{numNodes}} will be different in these two classes. In the meanwhile, to
handle over-replicated blocks in a block group (due to temporary datanode
failure etc.) we may need extra data structure in {{BlockGroupInfo}} to
explicitly record the block index of each triplet.
> Erasure Coding: extend BlockInfo to handle EC info
> --------------------------------------------------
>
> Key: HDFS-7716
> URL: https://issues.apache.org/jira/browse/HDFS-7716
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Jing Zhao
> Assignee: Jing Zhao
>
> The current BlockInfo's implementation only supports the replication
> mechanism. To use the same blocksMap handling block group and its data/parity
> blocks, we need to define a new BlockGroupInfo class.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
