Github user vanzin commented on a diff in the pull request:
https://github.com/apache/spark/pull/18235#discussion_r123802866
--- Diff: core/src/main/scala/org/apache/spark/deploy/SparkSubmit.scala ---
@@ -310,33 +310,28 @@ object SparkSubmit extends CommandLineUtils {
RPackageUtils.checkAndBuildRPackage(args.jars, printStream,
args.verbose)
}
- // In client mode, download remote files.
- if (deployMode == CLIENT) {
- val hadoopConf = new HadoopConfiguration()
- args.primaryResource =
Option(args.primaryResource).map(downloadFile(_, hadoopConf)).orNull
- args.jars = Option(args.jars).map(downloadFileList(_,
hadoopConf)).orNull
- args.pyFiles = Option(args.pyFiles).map(downloadFileList(_,
hadoopConf)).orNull
- args.files = Option(args.files).map(downloadFileList(_,
hadoopConf)).orNull
- }
+ val hadoopConf = new HadoopConfiguration()
+ val targetDir = Files.createTempDirectory("tmp").toFile
+ val sparkConf = new SparkConf(loadDefaults =
false).setAll(args.sparkProperties)
--- End diff --
Hmm... I think there might be a problem with logging when calling these
classes here. This code is running before the app's main class is called, and
it may initialize the logging system (e.g. `Utils.doFetchFile` later calls
logging methods in all code paths, initializing the logging system, and so does
`SecurityManager`).
This will be a problem, for example, for the spark-shell, which triggers a
slightly different code path when initializing logging. Because now the logging
system will be initialized by a class that is not the REPL's `Main` class, that
will be lost.
Long story short, with the code you're adding, the shell should still be
initialized with `WARN` as the default log level. Can you run `spark-shell`
with a remote jar in the jars list and verify it still maintains that?
---
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 have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
