[
https://issues.apache.org/jira/browse/HDDS-792?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16744868#comment-16744868
]
Steve Loughran commented on HDDS-792:
-------------------------------------
on multipart uploads, the etag on AWS S3 is not the MD5 of the whole file; it's
some string based on a concatenation of the checksums (or something similar).
If you set a small fs.s3a.block.size value and upload a larger file you can see
this. If the test suite you are using absolutely requires MD5 checksums, it's
not going to work at larger scale
> Use md5 hash as ETag for Ozone S3 objects
> -----------------------------------------
>
> Key: HDDS-792
> URL: https://issues.apache.org/jira/browse/HDDS-792
> Project: Hadoop Distributed Data Store
> Issue Type: Sub-task
> Components: S3
> Reporter: Elek, Marton
> Priority: Major
>
> AWS S3 uses md5 hash of the files as ETag.
> Not a strict requirement, but s3 tests (https://github.com/gaul/s3-tests/)
> can not been executed without that.
> It requires to support custom key/value annotations on key objects.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
