dongjoon-hyun commented on code in PR #54280:
URL: https://github.com/apache/spark/pull/54280#discussion_r2797272677
##########
core/src/main/scala/org/apache/spark/deploy/history/EventLogFileWriters.scala:
##########
@@ -131,7 +131,20 @@ abstract class EventLogFileWriter(
}
protected def closeWriter(): Unit = {
+ // 1. Flush first to check the errors
+ writer.foreach(_.flush())
+ if (writer.exists(_.checkError())) {
+ logWarning("Spark detects errors while flushing event logs.")
+ }
+ hadoopDataStream.foreach(_.hflush())
Review Comment:
This PR aims to avoid a silent failure inside `checkError`. For other
propagatable unhandled IOException, `SparkContext.stop` already logs like the
following. So, I didn't use `try ... catch ...` or `Utils.tryLog...`
intentionally.
https://github.com/apache/spark/blob/59f3a16590d8a8e1c942be8d129343a378e674d1/core/src/main/scala/org/apache/spark/SparkContext.scala#L2380-L2382
https://github.com/apache/spark/blob/59f3a16590d8a8e1c942be8d129343a378e674d1/core/src/main/scala/org/apache/spark/deploy/history/EventLogFileWriters.scala#L356-L363
In my case, the silent failure happens before `renameFile`. So, the last log
file is not uploaded correctly and `inprogress` file remains. As a result, SHS
shows running stages always because it's not finished.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
