Github user vanzin commented on a diff in the pull request:
https://github.com/apache/spark/pull/22881#discussion_r229802904
--- Diff: core/src/main/scala/org/apache/spark/deploy/SparkHadoopUtil.scala
---
@@ -471,4 +472,33 @@ object SparkHadoopUtil {
hadoopConf.set(key.substring("spark.hadoop.".length), value)
}
}
+
+ // scalastyle:off line.size.limit
+ /**
+ * Create a path that uses replication instead of erasure coding (ec),
regardless of the default
+ * configuration in hdfs for the given path. This can be helpful as
hdfs ec doesn't support
+ * hflush(), hsync(), or append()
+ *
https://hadoop.apache.org/docs/r3.0.0/hadoop-project-dist/hadoop-hdfs/HDFSErasureCoding.html#Limitations
+ */
+ // scalastyle:on line.size.limit
+ def createNonECFile(fs: FileSystem, path: Path): FSDataOutputStream = {
+ try {
+ // Use reflection as this uses apis only avialable in hadoop 3
+ val builderMethod = fs.getClass().getMethod("createFile",
classOf[Path])
+ val builder = builderMethod.invoke(fs, path)
+ val builderCls = builder.getClass()
+ // this may throw a NoSuchMethodException if the path is not on hdfs
+ val replicateMethod = builderCls.getMethod("replicate")
+ val buildMethod = builderCls.getMethod("build")
+ val b2 = replicateMethod.invoke(builder)
+ buildMethod.invoke(b2).asInstanceOf[FSDataOutputStream]
+ } catch {
+ case _: NoSuchMethodException =>
--- End diff --
If you get that exception, it's a proper error that shouldn't fall back to
this code path.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
