[GitHub] flink issue #4397: [FLINK-7265] [FLINK-7266] Introduce FileSystemKind and Co...

2017-11-17 Thread aljoscha
Github user aljoscha commented on the issue: https://github.com/apache/flink/pull/4397 @StephanEwen Could you please close this PR now that it's also merged for master/1.4? ---

[GitHub] flink issue #4397: [FLINK-7265] [FLINK-7266] Introduce FileSystemKind and Co...

2017-08-02 Thread StephanEwen
Github user StephanEwen commented on the issue: https://github.com/apache/flink/pull/4397 @steveloughran Thanks for the comment. We'll take this input into account for the more elaborate handling in the next version. --- If your project is set up for it, you can reply to this email

[GitHub] flink issue #4397: [FLINK-7265] [FLINK-7266] Introduce FileSystemKind and Co...

2017-08-01 Thread steveloughran
Github user steveloughran commented on the issue: https://github.com/apache/flink/pull/4397 We've long experimented with the best way to do this in Hadoop, and I think we're converting on moving off any form of enum to some `hasFeature(String)` predicate. Why? Lets you handle a more

[GitHub] flink issue #4397: [FLINK-7265] [FLINK-7266] Introduce FileSystemKind and Co...

2017-07-28 Thread StephanEwen
Github user StephanEwen commented on the issue: https://github.com/apache/flink/pull/4397 Want to see how we proceed with master. Same minimal version, or the more extensive version here? --- If your project is set up for it, you can reply to this email and have your reply appear on

[GitHub] flink issue #4397: [FLINK-7265] [FLINK-7266] Introduce FileSystemKind and Co...

2017-07-28 Thread StephanEwen
Github user StephanEwen commented on the issue: https://github.com/apache/flink/pull/4397 Merged a minimal version of this into `release-1.3` in 854b05376a459a6197e41e141bb28a9befe481ad --- If your project is set up for it, you can reply to this email and have your reply appear on

[GitHub] flink issue #4397: [FLINK-7265] [FLINK-7266] Introduce FileSystemKind and Co...

2017-07-28 Thread aljoscha
Github user aljoscha commented on the issue: https://github.com/apache/flink/pull/4397 Yes, I think either is fine because it's not user facing. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not

[GitHub] flink issue #4397: [FLINK-7265] [FLINK-7266] Introduce FileSystemKind and Co...

2017-07-28 Thread StephanEwen
Github user StephanEwen commented on the issue: https://github.com/apache/flink/pull/4397 How about I just remove the consistency level completely and reduce the `FileSystemKind` to "file system" and "object store"? I tried to think through the ecosystem of file systems quite

[GitHub] flink issue #4397: [FLINK-7265] [FLINK-7266] Introduce FileSystemKind and Co...

2017-07-26 Thread StephanEwen
Github user StephanEwen commented on the issue: https://github.com/apache/flink/pull/4397 I made the tests stable across Hadoop versions and addressed the comments/annotations/formatting. --- If your project is set up for it, you can reply to this email and have your reply appear on

[GitHub] flink issue #4397: [FLINK-7265] [FLINK-7266] Introduce FileSystemKind and Co...

2017-07-26 Thread zentol
Github user zentol commented on the issue: https://github.com/apache/flink/pull/4397 Test failure: ```testS3Kind(org.apache.flink.runtime.fs.hdfs.HdfsKindTest) Time elapsed: 0.186 sec <<< ERROR! java.io.IOException: No file system found with scheme s3, referenced in file URI