Reynold Xin created SPARK-8888:
----------------------------------
Summary: DynamicPartitionWriterContainer.outputWriterForRow is
very inefficient
Key: SPARK-8888
URL: https://issues.apache.org/jira/browse/SPARK-8888
Project: Spark
Issue Type: Sub-task
Reporter: Reynold Xin
A few things:
1. We should use a java.util.HashMap, which is faster than Scala's map,
although this one barely matters.
2. Don't use zip and scala collection methods to avoid garbage collection
{code}
val partitionPath = partitionColumns.zip(row.toSeq).map { case (col,
rawValue) =>
val string = if (rawValue == null) null else String.valueOf(rawValue)
val valueString = if (string == null || string.isEmpty) {
defaultPartitionName
} else {
PartitioningUtils.escapePathName(string)
}
s"/$col=$valueString"
}.mkString.stripPrefix(Path.SEPARATOR)
{code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
