Github user andrewor14 commented on a diff in the pull request:
https://github.com/apache/spark/pull/450#discussion_r11880274
--- Diff: core/src/main/scala/org/apache/spark/util/FileLogger.scala ---
@@ -85,11 +87,13 @@ private[spark] class FileLogger(
private def createWriter(fileName: String): PrintWriter = {
val logPath = logDir + "/" + fileName
val uri = new URI(logPath)
+ val defaultFs = FileSystem.getDefaultUri(hadoopConfiguration).getScheme
+ val isDefaultLocal = (defaultFs == null || "file".equals(defaultFs))
/* The Hadoop LocalFileSystem (r1.0.4) has known issues with syncing
(HADOOP-7844).
* Therefore, for local files, use FileOutputStream instead. */
val dstream = uri.getScheme match {
- case "file" | null =>
+ case null | "file" if isDefaultLocal =>
--- End diff --
Actually, this is incorrect I think. If the default is not local, but the
user specifies `file://` then this will use the next case and fail silently
because of sync issues.
By the way the logic is a little obfuscated to me right now. Since we only
match on 2 cases here, I think it's more straightforward to write it out with
`if-else`.
---
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 [email protected] or file a JIRA ticket
with INFRA.
---
