[
https://issues.apache.org/jira/browse/HADOOP-14090?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15872655#comment-15872655
]
Sean Mackrory commented on HADOOP-14090:
----------------------------------------
Just had a quick offline chat with [~fabbri] about this and changed my thinking
a little bit. Some key points (not all of which directly relate, but that are
important to know):
* Initially in the patch I wasn't clear on if specifying a region and a
specific endpoint together was ever valid. Having reviewed the docs, it does
seem that regions are a convenient abstraction to look up endpoints, so they
should be treated as mutually exclusive, and in the end this change would just
be adding a very similar option and making the code more complex.
* DynamoDB table names are not globally unique, unlike S3 buckets, so there are
a few more pain points if you DON'T specify the endpoint or region.
* Regions are slightly more intuitive for many AWS users, but endpoints are not
very hard to look up, and may be required for more advanced users.
So over all, I'll just withdraw this suggestion / patch, I think. There's
definitely some clean up I'd like to do in this code: in the functions that
call DynamoDBClientFactory.createDynamoDBClient(URI, String), there are several
layers in which the region name might be extracted from the bucket that appear
unnecessary), and I think we need to switch to full-word flags in S3GuardTool
anyway (see HADOOP-14094), but I think we should just allow the endpoint to be
set like we currently do for S3 too.
> Allow users to specify region for DynamoDB table instead of endpoint
> --------------------------------------------------------------------
>
> Key: HADOOP-14090
> URL: https://issues.apache.org/jira/browse/HADOOP-14090
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: fs/s3
> Reporter: Sean Mackrory
> Attachments: HADOOP-14090-HADOOP-13345.001.patch
>
>
> Assuming the AWS SDK allows this, I think this would be a better way to
> configure it for any usage on AWS itself (with endpoint still being an option
> for AWS-compatible non-AWS use cases). Unless users actually care about a
> specific endpoint, this is easier. Perhaps less important, HADOOP-14023 shows
> that inferring the region from the endpoint (which granted, isn't that
> necessary) doesn't work very well at all.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
