[
https://issues.apache.org/jira/browse/HADOOP-18351?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17654508#comment-17654508
]
ASF GitHub Bot commented on HADOOP-18351:
-----------------------------------------
sauraank opened a new pull request, #5274:
URL: https://github.com/apache/hadoop/pull/5274
<!--
Thanks for sending a pull request!
1. If this is your first time, please read our contributor guidelines:
https://cwiki.apache.org/confluence/display/HADOOP/How+To+Contribute
2. Make sure your PR title starts with JIRA issue id, e.g.,
'HADOOP-17799. Your PR title ...'.
-->
### Description of PR
Logged the errors at the info level with just the message. Added the stack
at the Debug level. Not made any changes where error stack was not present and
had only the message and parameters at Warn/INFO level.
Also, removed the case of multiple logging in case of prefetch. Moved the
log one method above ReadBlock() (which was common for sync read thread and
async prefetch thread) and logged it at ERROR level for the case of read.
### How was this patch tested?
Logging levels of Error/info was tested locally by throwing errors in case
of both cases of async prefetch and sync get calls for read separately. Then
the unit test of ITestS3APrefetchingInputStream integration test (
testReadLargeFileFully) was used.
Finally done the integration test by `mvn -Dparallel-tests clean verify` in
the `eu-west-1` region.
Following is the result of the tests -
[INFO] Results:
[INFO]
[WARNING] Tests run: 1154, Failures: 0, Errors: 0, Skipped: 182
[INFO]
[INFO]
[INFO] -
> S3A prefetching: Error logging during reads
> -------------------------------------------
>
> Key: HADOOP-18351
> URL: https://issues.apache.org/jira/browse/HADOOP-18351
> Project: Hadoop Common
> Issue Type: Sub-task
> Affects Versions: 3.4.0
> Reporter: Ahmar Suhail
> Assignee: Ankit Saurabh
> Priority: Minor
>
> Look at how errors during read are logged, current implementation could flood
> logs with stack traces on failures.
> proposed
> * errors in prefetch only logged at info with error text but not stack
> * full stack logged at debug
> but: we do want the most recent failure to be raised on the next read() on
> the stream when there is no data in the cache.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
