bogthe opened a new pull request #3260: URL: https://github.com/apache/hadoop/pull/3260
[HADOOP-17198](https://issues.apache.org/jira/browse/HADOOP-17198) This change aims to add support for S3 AccessPoints. To use S3 object level APIs for an AccessPoint, one has to use the AccessPoint (AP) ARN. Hence the following have been added: - a new property to set the AccessPoint ARN; - S3a parsing and using the new property with appropriate exceptions; - initial documentation update for AccessPoints; What this PR enables: - If `apname` is the name of an AccessPoint you have for created bucket then S3a now allows you to use paths like `s3a://apname/` IF the new `s3a.accesspoint.arn` is set to the AccessPoint ARN e.g. `arn:aws:s3:eu-west-1:123456789101:accesspoint/apname`; There's one thing I'm not sure about with this initial implementation so am looking for feedback if and how I should tackle it: `S3a` bucket now has 2 "meanings" it can be a bucket name or an Access Point ARN. From the point of view of interacting with the SDK, they are interchangeable and internal string parsing logic is used to create the request for the right endpoint. However, I think it would be nicer to have a clearer abstraction for bucket names or access point ARNs that S3a operations can work with. This abstraction comes with the cost of doing a refactor which I'm not sure it's worth it right now. Even by doing a quick search on `.getHost()` there's quite a few places where the bucket name is deduced from the `host`. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
