You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Dexter Hu (Jira)" <ji...@apache.org> on 2020/10/07 07:53:00 UTC
[jira] [Commented] (SPARK-33083) Optionally skip remote/local
dependency resolution at submission client in k8s cluster mode
[ https://issues.apache.org/jira/browse/SPARK-33083?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17209360#comment-17209360 ]
Dexter Hu commented on SPARK-33083:
-----------------------------------
The main improvement request is that the submission client doesn't have to validate spark properties (jars, files, py-files), I just want it to pass to the driver and executor in *cluster* mode, as I will not use *client* mode.
> Optionally skip remote/local dependency resolution at submission client in k8s cluster mode
> -------------------------------------------------------------------------------------------
>
> Key: SPARK-33083
> URL: https://issues.apache.org/jira/browse/SPARK-33083
> Project: Spark
> Issue Type: Improvement
> Components: Kubernetes, Spark Submit
> Affects Versions: 2.4.7
> Reporter: Dexter Hu
> Priority: Minor
> Fix For: 2.4.7
>
>
> Usecase:
> # Users use Apache Livy to submit spark jobs. Livy pod's spark-submit command will be invoked to submit jobs to k8e cluster with *--deploy-mode cluster*
> # The Livy pod (submission client) does *NOT* have permission to access remote dependencies specified by *--files, --jars, --py-files* from S3, gcs, hdfs:// . Only future driver pod and executor pod will have permissions with dynamic secure tokens to download those remote files.
> Is it possible to support options to *disable* remote resource resolution/download at Livy pod or submission client host?
> Of course, users/developers will make sure that
> # Remote files are accessible by driver and executor pods with secure tokens.
> # files with local paths are placed into driver and executor in advance.
>
>
>
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org