[
https://issues.apache.org/jira/browse/HDFS-8319?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14569927#comment-14569927
]
Kai Zheng commented on HDFS-8319:
---------------------------------
Thanks Jing and Zhe for the good work and discussion here!
I really like {{StripeReader}} and the family. Could it be good to move on a
little bit in the direction? I thought eventually we will need a separate
standalone high level construct like {{StripeDecoder}} which mainly takes care
of collecting inputs, perform decoding, and processing outputs and etc. For
now, maybe StrippeDecoder could be StripeDecoder, move the relevant members to
it, or have something like below
{code}
public class StripeDecoder {
private final int cellSize;
private final short dataBlkNum;
private final short parityBlkNum;
private final short groupSize;
private final ECSchema schema;
private final RawErasureDecoder decoder;
StripeDecoder (ECSchema schema, int cellSize) {
assert schema != null;
this.schema = schema;
this.cellSize = cellSize;
dataBlkNum = (short) schema.getNumDataUnits();
parityBlkNum = (short) schema.getNumParityUnits();
groupSize = dataBlkNum;
decoder = new RSRawDecoder(dataBlkNum, parityBlkNum);
}
{code}
In future higher level {{ErasureCoder}} can be used to replace the current
{{RawErasureCoder}} based on such work, we won't need to involve much change
then in {{DFSStripedInputStream}}.
In addition, PositionStripeReader could be PositionalStripeReader, similar to
StatefulStripeReader.
> Erasure Coding: support decoding for stateful read
> --------------------------------------------------
>
> Key: HDFS-8319
> URL: https://issues.apache.org/jira/browse/HDFS-8319
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Jing Zhao
> Assignee: Jing Zhao
> Attachments: HDFS-8319.001.patch, HDFS-8319.002.patch
>
>
> HDFS-7678 adds the decoding functionality for pread. This jira plans to add
> decoding to stateful read.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
