[转载] How Many Partitions Does An RDD Have

Posted

tags:

篇首语:本文由小常识网(cha138.com)小编为大家整理,主要介绍了[转载] How Many Partitions Does An RDD Have相关的知识,希望对你有一定的参考价值。

From https://databricks.gitbooks.io/databricks-spark-knowledge-base/content/performance_optimization/how_many_partitions_does_an_rdd_have.html

For tuning and troubleshooting, it‘s often necessary to know how many paritions an RDD represents. There are a few ways to find this information:

View Task Execution Against Partitions Using the UI

When a stage executes, you can see the number of partitions for a given stage in the Spark UI. For example, the following simple job creates an RDD of 100 elements across 4 partitions, then distributes a dummy map task before collecting the elements back to the driver program:

scala> val someRDD = sc.parallelize(1 to 100, 4)
someRDD: org.apache.spark.rdd.RDD[Int] = ParallelCollectionRDD[0] at parallelize at <console>:12

scala> someRDD.map(x => x).collect
res1: Array[Int] = Array(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100)

In Spark‘s application UI, you can see from the following screenshot that the "Total Tasks" represents the number of partitions:

技术分享

View Partition Caching Using the UI

When persisting (a.k.a. caching) RDDs, it‘s useful to understand how many partitions have been stored. The example below is identical to the one prior, except that we‘ll now cache the RDD prior to processing it. After this completes, we can use the UI to understand what has been stored from this operation.

scala> someRDD.setName("toy").cache
res2: someRDD.type = toy ParallelCollectionRDD[0] at parallelize at <console>:12

scala> someRDD.map(x => x).collect
res3: Array[Int] = Array(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100)

Note from the screenshot that there are four partitions cached.

技术分享

Inspect RDD Partitions Programatically

In the Scala API, an RDD holds a reference to it‘s Array of partitions, which you can use to find out how many partitions there are:

scala> val someRDD = sc.parallelize(1 to 100, 30)
someRDD: org.apache.spark.rdd.RDD[Int] = ParallelCollectionRDD[0] at parallelize at <console>:12

scala> someRDD.partitions.size
res0: Int = 30

In the python API, there is a method for explicitly listing the number of partitions:

In [1]: someRDD = sc.parallelize(range(101),30)

In [2]: someRDD.getNumPartitions()
Out[2]: 30

Note in the examples above, the number of partitions was intentionally set to 30 upon initialization.

以上是关于[转载] How Many Partitions Does An RDD Have的主要内容,如果未能解决你的问题,请参考以下文章

How to remove the "Active" mark on partitions

How to choose the number of topics/partitions in a Kafka cluster?

UVA - 10061 How many zero&#39;s and how many digits ?

hdu 1316 How Many Fibs?

P - How many

HDOJ 1316 How Many Fibs?