ajantha-bhat commented on a change in pull request #3621: [HOTFIX] Support both
listfile() and listfile(maxCount) in InsertStag…
URL: https://github.com/apache/carbondata/pull/3621#discussion_r379985373
##########
File path:
integration/spark2/src/main/scala/org/apache/spark/sql/execution/command/management/CarbonInsertFromStageCommand.scala
##########
@@ -502,13 +503,23 @@ case class CarbonInsertFromStageCommand(
): Array[(CarbonFile, CarbonFile)] = {
val dir = FileFactory.getCarbonFile(loadDetailsDir, hadoopConf)
if (dir.exists()) {
- // Only HDFS/OBS/S3 server side can guarantee the files got from
iterator are sorted
- // based on file name so that we can use iterator to get the A and
A.success together
- // without loop all files which can improve performance compared with
list all files.
- // One file and another with '.success', so we need *2 as total and this
value is just
- // an approximate value. For local files, as can it can we not guarantee
the order, we
- // just list all.
- val allFiles = dir.listFiles(false, batchSize * 2)
+ // It is possible that the filename of stage files is not in order of
time,
Review comment:
Can you explain in which scenario it won't be ordered of time ?
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
