Skip to content

Conversation

rahulsinghaliitd
Copy link

Generate an unique app name which is used to create events and metric folders.

@AmplabJenkins
Copy link

Can one of the admins verify this patch?

Copy link
Contributor

Choose a reason for hiding this comment

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

Hi @rahulsinghaliitd , do we really need to use SparkConf to get this unique app name? I think metrics system is driven by its own configuration system, here you involved SparkConf as another configuration system just only for CsvSink, this will give people the chance to bypass the original way and lose its controllability. I think it would be nice to modify the way to get appUniqueName (eg. get it from SparkEnv).

Copy link
Author

Choose a reason for hiding this comment

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

@jerryshao Thanks for the feedback! I was not aware of the fact the SparkEnv provides access to SparkConf. Will follow up with the suggested modification.

Copy link
Contributor

Choose a reason for hiding this comment

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

Forgot to mention that several sinks like CsvSink will be started in master and worker for standalone mode if configured, in this situation you may not get unique app name or SparkEnv (something created only when app started). So maybe you should consider this situation when there's no app name available.

Copy link
Author

Choose a reason for hiding this comment

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

Hi @jerryshao ,

  1. At the moment the Sinks are being created, SparkEnv has not been created. I may be able to modify the Properties being passed to this Sink or even get the SparkConf from SecurityManager. But neither of those approaches seems generic to me. For e.g. we will need hadoopConf if we wanted the csv directory to be on HDFS.
  2. Thanks for pointing out the problem with Master and Worker. I have for now added app names to these classes. Please let me know if you think adding null checks in CsvSink would also be useful.

1. Generate an unique app name which is used to create events and metric folders.
2. Add app name & unique app name to Standalone mode's Master and Worker.
3. Log the unique app directory that will be used by CsvSink.
@SparkQA
Copy link

SparkQA commented Sep 5, 2014

Can one of the admins verify this patch?

@asfgit asfgit closed this in 79e45c9 Oct 3, 2014
@nitindexter nitindexter deleted the SPARK-2127 branch January 16, 2015 09:10
wangyum pushed a commit that referenced this pull request May 26, 2023
…n for Hadoop platform (#1067)

* [CARMEL-6200] Add unified command to extract table lineage information for Hadoop platform

* Add tests
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants