[
https://issues.apache.org/jira/browse/HADOOP-18399?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17712517#comment-17712517
]
ASF GitHub Bot commented on HADOOP-18399:
-----------------------------------------
virajjasani commented on code in PR #5054:
URL: https://github.com/apache/hadoop/pull/5054#discussion_r1167231126
##########
hadoop-tools/hadoop-aws/src/main/java/org/apache/hadoop/fs/s3a/prefetch/S3APrefetchingInputStream.java:
##########
@@ -110,11 +117,17 @@ public S3APrefetchingInputStream(
streamStatistics);
} else {
LOG.debug("Creating in caching input stream for {}", context.getPath());
+ final String bufferDir =
+ conf.get(BUFFER_DIR) != null ? BUFFER_DIR : HADOOP_TMP_DIR;
+ final LocalDirAllocator localDirAllocator =
Review Comment:
i think i missed this one in the first place, should have passed down S3AFS
dir allocator only.
Done.
> SingleFilePerBlockCache to use LocalDirAllocator for file allocation
> --------------------------------------------------------------------
>
> Key: HADOOP-18399
> URL: https://issues.apache.org/jira/browse/HADOOP-18399
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: fs/s3
> Affects Versions: 3.4.0
> Reporter: Steve Loughran
> Assignee: Viraj Jasani
> Priority: Major
> Labels: pull-request-available
>
> prefetching stream's SingleFilePerBlockCache uses Files.tempFile() to
> allocate a temp file.
> it should be using LocalDirAllocator to allocate space from a list of dirs,
> taking a config key to use. for s3a we will use the Constants.BUFFER_DIR
> option, which on yarn deployments is fixed under the env.LOCAL_DIR path, so
> automatically cleaned up on container exit
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
