Mike Dusenberry created SPARK-17817:
---------------------------------------
Summary: PySpark RDD Repartitioning Results in Highly Skewed
Partition Sizes
Key: SPARK-17817
URL: https://issues.apache.org/jira/browse/SPARK-17817
Project: Spark
Issue Type: Bug
Affects Versions: 2.0.1, 2.0.0, 1.6.2, 1.6.1
Reporter: Mike Dusenberry
Calling {{repartition}} on a PySpark RDD to increase the number of partitions
results in highly skewed partition sizes, with most having 0 rows. The
{{repartition}} method should evenly spread out the rows across the partitions,
and this behavior is correctly seen on the Scala side.
Please reference the following code for a reproducible example of this issue:
{code}
# Python
num_partitions = 20000
a = sc.parallelize(range(int(1e6)), 2) # start with 2 even partitions
l = a.repartition(num_partitions).glom().map(len).collect() # get length of
each partition
min(l), max(l), sum(l)/len(l), len(l) # skewed!
# Scala
val numPartitions = 20000
val a = sc.parallelize(0 until 1e6.toInt, 2) # start with 2 even partitions
val l = a.repartition(numPartitions).glom().map(_.length).collect() # get
length of each partition
print(l.min, l.max, l.sum/l.length, l.length) # even!
{code}
The issue here is that highly skewed partitions can result in severe memory
pressure in subsequent steps of a processing pipeline, resulting in OOM errors.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
