SPARK-21549 has shown that downstream code is relying on the internal property
if we pulled FileOutputCommitter.getOutputPath to the PathOutputCommitter of MAPREDUCE-6956, then there'd be a public/stable way to get this. Admittedly, it does imply that the committer will always have some output path, but FileOutputFormat depends on that anyway.
- Is contained by
-
HADOOP-13786 Add S3A committers for zero-rename commits to S3 endpoints
-
- Resolved
-
- relates to
-
MAPREDUCE-6823 FileOutputFormat to support configurable PathOutputCommitter factory
-
- Resolved
-
-
SPARK-21549 Spark fails to complete job correctly in case of OutputFormat which do not write into hdfs
-
- Resolved
-
-
MAPREDUCE-6956 FileOutputCommitter to gain abstract superclass PathOutputCommitter
-
- Resolved
-