Github user vanzin commented on a diff in the pull request:
https://github.com/apache/spark/pull/22504#discussion_r226456519
--- Diff:
core/src/main/scala/org/apache/spark/deploy/history/FsHistoryProvider.scala ---
@@ -117,6 +118,13 @@ private[history] class FsHistoryProvider(conf:
SparkConf, clock: Clock)
// Visible for testing
private[history] val fs: FileSystem = new
Path(logDir).getFileSystem(hadoopConf)
+ private val driverLogFs: Option[FileSystem] =
+ if (conf.get(DRIVER_LOG_DFS_DIR).isDefined) {
+ Some(FileSystem.get(hadoopConf))
--- End diff --
This is not right. It assumes the directory defined is in the defaultFS.
See call right above this one for what you should do here.
I also don't see a good reason to keep this in a field. Just get the FS
reference when the log cleaner runs.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
