hadoop-common-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Seth Fitzsimmons (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HADOOP-14028) S3A block output streams don't clear temporary files
Date Thu, 26 Jan 2017 22:53:24 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-14028?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15840605#comment-15840605
] 

Seth Fitzsimmons commented on HADOOP-14028:
-------------------------------------------

Reading through the AWS SDK code, it looks like this is the line ultimately responsible for
closing the input stream: https://github.com/aws/aws-sdk-java/blob/master/aws-java-sdk-core/src/main/java/com/amazonaws/internal/ReleasableInputStream.java#L85

I'm using the default settings (other than {{fs.s3a.fast.upload=true}}.

>From watching the atimes, it looks like there's only 1 block going up at a time while
the next one fills up. (My producer is relatively slow and I'm running in EC2, so it makes
sense that the uploader can keep up).

> S3A block output streams don't clear temporary files
> ----------------------------------------------------
>
>                 Key: HADOOP-14028
>                 URL: https://issues.apache.org/jira/browse/HADOOP-14028
>             Project: Hadoop Common
>          Issue Type: Bug
>          Components: fs/s3
>    Affects Versions: 3.0.0-alpha2
>         Environment: JDK 8 + ORC 1.3.0 + hadoop-aws 3.0.0-alpha2
>            Reporter: Seth Fitzsimmons
>
> I have `fs.s3a.fast.upload` enabled with 3.0.0-alpha2 (it's exactly what I was looking
for after running into the same OOM problems) and don't see it cleaning up the disk-cached
blocks.
> I'm generating a ~50GB file on an instance with ~6GB free when the process starts. My
expectation is that local copies of the blocks would be deleted after those parts finish uploading,
but I'm seeing more than 15 blocks in /tmp (and none of them have been deleted thus far).
> I see that DiskBlock deletes temporary files when closed, but is it closed after individual
blocks have finished uploading or when the entire file has been fully written to the FS (full
upload completed, including all parts)?
> As a temporary workaround to avoid running out of space, I'm listing files, sorting by
atime, and deleting anything older than the first 20: `ls -ut | tail -n +21 | xargs rm`
> Steve Loughran says:
> > They should be deleted as soon as the upload completes; the close() call that the
AWS httpclient makes on the input stream triggers the deletion. Though there aren't tests
for it, as I recall.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: common-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: common-issues-help@hadoop.apache.org


Mime
View raw message