[
https://issues.apache.org/jira/browse/HDFS-9173?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15064658#comment-15064658
]
Jing Zhao commented on HDFS-9173:
---------------------------------
Based on our original hflush design, we will write parity blocks for the
existing data when calling hflush, and we overwrite them later for further
data. Thus if the hflush is in the picture, we need to collect extra
information from {{initReplicaRecovery}} to understand if the last chunk of a
parity block is generated based on the full or just partial stripe. Otherwise
we can use wrong data for later decoding.
Also, hflush will not always update visible length of the block. Only the first
hflush call does, or we pass the UPDATE_LENGTH flag.
> Erasure Coding: Lease recovery for striped file
> -----------------------------------------------
>
> Key: HDFS-9173
> URL: https://issues.apache.org/jira/browse/HDFS-9173
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Walter Su
> Assignee: Walter Su
> Attachments: HDFS-9173.00.wip.patch, HDFS-9173.01.patch,
> HDFS-9173.02.step125.patch, HDFS-9173.03.patch, HDFS-9173.04.patch,
> HDFS-9173.05.patch, HDFS-9173.06.patch, HDFS-9173.07.patch,
> HDFS-9173.08.patch, HDFS-9173.09.patch, HDFS-9173.09.patch
>
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
