[
https://issues.apache.org/jira/browse/HDFS-15795?focusedWorklogId=543589&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-543589
]
ASF GitHub Bot logged work on HDFS-15795:
-----------------------------------------
Author: ASF GitHub Bot
Created on: 28/Jan/21 11:39
Start Date: 28/Jan/21 11:39
Worklog Time Spent: 10m
Work Description: sodonnel commented on a change in pull request #2657:
URL: https://github.com/apache/hadoop/pull/2657#discussion_r566026704
##########
File path:
hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/BlockChecksumHelper.java
##########
@@ -503,6 +503,7 @@ void compute() throws IOException {
}
} catch (IOException e) {
Review comment:
Yes, we could just remove the try-catch block entirely and let the
exception propagate upwards. However I wonder if it would make sense to catch
it as we do now, but add some more information to the log for later debugging,
eg the block causing the problem and the block index:
```
LOG.warn("Failed to get the checksum for block index {} and block {} in
block group {}", idx, block, blockGroup, e);
throw(e);
```
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
Issue Time Tracking
-------------------
Worklog Id: (was: 543589)
Time Spent: 1h (was: 50m)
> EC: Returned wrong checksum when reconstruction was failed by exception
> -----------------------------------------------------------------------
>
> Key: HDFS-15795
> URL: https://issues.apache.org/jira/browse/HDFS-15795
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: datanode, ec, erasure-coding
> Reporter: Yushi Hayasaka
> Assignee: Yushi Hayasaka
> Priority: Major
> Labels: pull-request-available
> Time Spent: 1h
> Remaining Estimate: 0h
>
> If the reconstruction task is failed on StripedBlockChecksumReconstructor by
> exception, the checksum becomes wrong one because it is calculated with
> blocks except a failure one.
> It is caused by catching exception with not appropriate way. As a result, the
> failed block is not fetched again.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
