diff options
author | Patrick Wendell <pwendell@gmail.com> | 2013-12-06 20:16:15 -0800 |
---|---|---|
committer | Patrick Wendell <pwendell@gmail.com> | 2013-12-06 20:16:15 -0800 |
commit | 1b38f5f2774982d524742e987b6cef26ccaae676 (patch) | |
tree | 2269dab6d64aa2e1f2d0d88b179a76bf277cf70d | |
parent | e5d5728b72e58046cc175ab06b5f1c7be4957711 (diff) | |
parent | b9451acdf4c73ab4b0b9d9827ab1ccedbcfb144f (diff) | |
download | spark-1b38f5f2774982d524742e987b6cef26ccaae676.tar.gz spark-1b38f5f2774982d524742e987b6cef26ccaae676.tar.bz2 spark-1b38f5f2774982d524742e987b6cef26ccaae676.zip |
Merge pull request #236 from pwendell/shuffle-docs
Adding disclaimer for shuffle file consolidation
-rw-r--r-- | docs/configuration.md | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/docs/configuration.md b/docs/configuration.md index 22abe1c5a9..65b41b45cb 100644 --- a/docs/configuration.md +++ b/docs/configuration.md @@ -331,7 +331,7 @@ Apart from these, the following properties are also available, and may be useful <td>spark.shuffle.consolidateFiles</td> <td>false</td> <td> - If set to "true", consolidates intermediate files created during a shuffle. Creating fewer files can improve filesystem performance if you run shuffles with large numbers of reduce tasks. + If set to "true", consolidates intermediate files created during a shuffle. Creating fewer files can improve filesystem performance for shuffles with large numbers of reduce tasks. It is reccomended to set this to "true" when using ext4 or xfs filesystems. On ext3, this option might degrade performance on machines with many (>8) cores due to filesystem limitations. </td> </tr> <tr> |