[
https://issues.apache.org/jira/browse/HDFS-17811?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
farmmamba updated HDFS-17811:
-----------------------------
Description:
This is an very useful improvment for EC read. It can reduce the the frequency
of read failures due to a rolling restart of the DataNodes.
We have met below use case:
{code:java}
InputStream in = open();
while (not reach EOF) {
in.read(buf, offset, len);
do some logic which is long time costed.
} {code}
If we start rolling restart of the datanodes, above case will finally fail
because all blockReaders may be marked as skip status by StripeReader. So, we
should make DFSStripedInputStream support retrying just like what
DFSInputStream does.
was:EC: DFSStripedInputStream supports retrying just like DFSInputStream
> EC: DFSStripedInputStream supports retrying just like DFSInputStream
> --------------------------------------------------------------------
>
> Key: HDFS-17811
> URL: https://issues.apache.org/jira/browse/HDFS-17811
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: ec, hdfs-client
> Reporter: farmmamba
> Assignee: farmmamba
> Priority: Major
> Labels: pull-request-available
>
> This is an very useful improvment for EC read. It can reduce the the
> frequency of read failures due to a rolling restart of the DataNodes.
> We have met below use case:
> {code:java}
> InputStream in = open();
> while (not reach EOF) {
> in.read(buf, offset, len);
> do some logic which is long time costed.
> } {code}
> If we start rolling restart of the datanodes, above case will finally fail
> because all blockReaders may be marked as skip status by StripeReader. So, we
> should make DFSStripedInputStream support retrying just like what
> DFSInputStream does.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
