Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[FLINK-1266] Generalize DistributedFileSystem implementation #268

Closed
wants to merge 4 commits into from

Conversation

rmetzger
Copy link
Contributor

to HadoopFileSystem wrapper, which supports all subclasses of org.apache.hadoop.fs.FileSystem.
This allows us to let users use all file systems with support for HDFS.

The change has been tested with Tachyon, Google Cloud Storage Hadoop Adapter and HDFS. (I also contributed again to Tachyon)

The change also cleans up the Hadoop dependency exclusions.
I also found a nice little "bug" in the Instance class.


WordCount.main(new String[]{input, output});

// List<Integer> files = client.listFiles("/", true);
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Could we remove the commented lines here?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes. Its not too hard to come up with the code again if one wants to enhance / debug the test case.

I'll remove the comments.

@StephanEwen
Copy link
Contributor

Looks nice!

Since you touched the documentation of the connectors anyways, can you replace the section on MongDB with a reference to this code: https://github.com/okkam-it/flink-mongodb-test

Maybe add a small (manual) TOC at the top (or put MongoDB before Azure so both are visible initially)

Otherwise, +1 to merge.

<dependency>
<groupId>org.eclipse.jetty</groupId>
<artifactId>jetty-util</artifactId>
<version>7.6.8.v20121106</version>
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This may conflict with the jetty version used by flink-runtime for the web info server

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I will double check what maven makes out of these two versions.
But the Jetty version is test scoped and should not influence the regular build.

it also seems that I forgot to scope the regular tachyon dependency to test.

@StephanEwen
Copy link
Contributor

This pull request moves the hadoop transitive dependency exclusions to the dependency management. Have you checked that this works properly? I ran into an issue there a while back where it did not properly work. Cannot quite recall why, though.

@rmetzger
Copy link
Contributor Author

rmetzger commented Jan 7, 2015

Yes, I found it very inconvenient to copy paste the exclusions everywhere. I will check if the exclusions still work correctly.

to HadoopFileSystem wrapper, which supports all subclasses of org.apache.hadoop.fs.FileSystem.
This allows us to let users use all file systems with support for HDFS.
The change has been tested with Tachyon, Google Cloud Storage Hadoop Adapter and HDFS.

The change also cleans up the Hadoop dependency exclusions.
@rmetzger
Copy link
Contributor Author

rmetzger commented Jan 8, 2015

Update: I fixed a major bug with the "2.0.0-alpha" version on the master.
Currently, the master does not do any dependency exclusions for hadoop. This causes stuff like two asm versions in the lib path.

I also verified that the dependency exclusions in the dependencyManagement section are working. (Its also in the official maven documentation .. so it should work).

@StephanEwen
Copy link
Contributor

Thanks, that looks good.

@asfgit asfgit closed this in 39fb7c9 Jan 8, 2015
mbalassi pushed a commit to mbalassi/flink that referenced this pull request Jan 8, 2015
zhijiangW pushed a commit to zhijiangW/flink that referenced this pull request Jul 7, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
3 participants