[
https://issues.apache.org/jira/browse/HDFS-17521?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17846310#comment-17846310
]
ASF GitHub Bot commented on HDFS-17521:
---------------------------------------
zhengchenyu commented on PR #6813:
URL: https://github.com/apache/hadoop/pull/6813#issuecomment-2110181645
@zhangshuyan0
Thanks for your review!
I don't means that parity index is smaller. Parity index and data index is
fixed number, we can't update it.
The reproduce case:
When we call RawErasureDecoder::decode, and if the parameter `erasedIndexes`
is in special order. The special order is that the parity index precedes the
data index. For example, if erasedIndexes is [8,0], will reproduce this problem.
And you can run unit tests directly and reproduce this easily.
I printed erasedIndexes for all errors in the single test in
[wrongindex.txt](https://github.com/apache/hadoop/files/15308714/wrongindex.txt),
all meet this characteristic.
> EC: Fix calculation errors caused by special index order
> --------------------------------------------------------
>
> Key: HDFS-17521
> URL: https://issues.apache.org/jira/browse/HDFS-17521
> Project: Hadoop HDFS
> Issue Type: Bug
> Reporter: Chenyu Zheng
> Assignee: Chenyu Zheng
> Priority: Critical
> Labels: pull-request-available
>
> I found that if the erasedIndexes distribution is such that the parity index
> is in front of the data index, ec will produce wrong results when decoding.
> In fact, HDFS-15186 has described this problem, but does not fundamentally
> solve it.
> The reason is that the code assumes that erasedIndexes is preceded by the
> data index and followed by parity index. If there is a parity index placed in
> front of the data index, a calculation error will occur.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
