[
https://issues.apache.org/jira/browse/HDFS-8033?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Zhe Zhang updated HDFS-8033:
----------------------------
Attachment: HDFS-8033.002.patch
Thanks Yi for the discussion.
This new patch makes the following main changes based on the previous one:
# Maintains a {{currentNodes}} array, recording the list of DNs hosting the
current block group
# Overrides {{readWithStrategy}} instead of {{readBuffer}}.
# Adds logic in extended {{readWithStrategy}} to cross cell boundary in a
single read() call. Note that it still tries to drain existing buffers in
{{blockReaders}} without trying to {{readFully}} from them.
# Handles retries for encryption key and token
# Updates the test to read through entire file, which has 2 block groups. This
verifies that the extended {{blockSeekTo}} takes care of proceeding to the next
block group
# Changed {{DFSInputStream#closeCurrentBlockReader}} to
{{closeCurrentBlockReaders}}
> Erasure coding: stateful (non-positional) read from files in striped layout
> ---------------------------------------------------------------------------
>
> Key: HDFS-8033
> URL: https://issues.apache.org/jira/browse/HDFS-8033
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Zhe Zhang
> Assignee: Zhe Zhang
> Attachments: HDFS-8033.000.patch, HDFS-8033.001.patch,
> HDFS-8033.002.patch
>
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
