[
https://issues.apache.org/jira/browse/HADOOP-18517?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17627888#comment-17627888
]
ASF GitHub Bot commented on HADOOP-18517:
-----------------------------------------
steveloughran opened a new pull request, #5103:
URL: https://github.com/apache/hadoop/pull/5103
Adds new config option fs.azure.enable.readahead to turn off readahead
* also allows it to be passed in through openFile(), which means that if you
set it in a mr/spark/hive job conf with "mapreduce.job.input.file.option." as
the prefix, it will be disabled when used through LineRecordReader,
FixedLengthRecordReader and a few others.
* extends ITestAbfsReadWriteAndSeek to use the option, including one
replicated test...that shows that turning it off is slower.
### How was this patch tested?
1. modified (scale) itest.
2. been in the cloudera builds for a long time.
3.
### For code changes:
- [X] Does the title or this PR starts with the corresponding JIRA issue id
(e.g. 'HADOOP-17799. Your PR title ...')?
- [X] Object storage: have the integration tests been executed and the
endpoint declared according to the connector-specific documentation?
- [ ] If adding new dependencies to the code, are these dependencies
licensed in a way that is compatible for inclusion under [ASF
2.0](http://www.apache.org/legal/resolved.html#category-a)?
- [ ] If applicable, have you updated the `LICENSE`, `LICENSE-binary`,
`NOTICE-binary` files?
> ABFS: Add fs.azure.enable.readahead option to disable readahead
> ---------------------------------------------------------------
>
> Key: HADOOP-18517
> URL: https://issues.apache.org/jira/browse/HADOOP-18517
> Project: Hadoop Common
> Issue Type: Improvement
> Components: fs/azure
> Affects Versions: 3.3.4
> Reporter: Steve Loughran
> Assignee: Steve Loughran
> Priority: Major
>
> Add an option fs.azure.enable.readahead to allow the abfs readahead to be
> disabled.
> This has been shipping in cloudera releases, because workloads with heavy
> random IO don't always benefit from prefetching.
> Pulling this in to ASF hadoop allows for
> * everyone to experiment with/without prefetching
> * ability to disable prefetching if you have problems with it
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
