RDD Lineage — Logical Execution Plan-spark技术分享

RDD Lineage — Logical Execution Plan

RDD Lineage (aka RDD operator graph or RDD dependency graph) is a graph of all the parent RDDs of a RDD. It is built as a result of applying transformations to the RDD and creates a logical execution plan.

Note	The execution DAG or physical execution plan is the DAG of stages.

Note	The following diagram uses `cartesian` or `zip` for learning purposes only. You may use other operators to build a RDD graph.

Figure 1. RDD lineage

The above RDD graph could be the result of the following series of transformations:


val r00 = sc.parallelize(0 to 9)
val r01 = sc.parallelize(0 to 90 by 10)
val r10 = r00 cartesian r01
val r11 = r00.map(n => (n, n))
val r12 = r00 zip r01
val r13 = r01.keyBy(_ / 20)
val r20 = Seq(r11, r12, r13).foldLeft(r10)(_ union _)

val r00 = sc.parallelize(0 to 9)

val r01 = sc.parallelize(0 to 90 by 10)

val r10 = r00 cartesian r01

val r11 = r00.map(n => (n, n))

val r12 = r00 zip r01

val r13 = r01.keyBy(_ / 20)

val r20 = Seq(r11, r12, r13).foldLeft(r10)(_ union _)

A RDD lineage graph is hence a graph of what transformations need to be executed after an action has been called.

You can learn about a RDD lineage graph using RDD.toDebugString method.

Logical Execution Plan

Logical Execution Plan starts with the earliest RDDs (those with no dependencies on other RDDs or reference cached data) and ends with the RDD that produces the result of the action that has been called to execute.

Note	A logical plan, i.e. a DAG, is materialized and executed when `SparkContext` is requested to run a Spark job.

Getting RDD Lineage Graph — `toDebugString` Method


toDebugString: String

toDebugString: String

You can learn about a RDD lineage graph using toDebugString method.


scala> val wordCount = sc.textFile("README.md").flatMap(_.split("\\s+")).map((_, 1)).reduceByKey(_ + _)
wordCount: org.apache.spark.rdd.RDD[(String, Int)] = ShuffledRDD[21] at reduceByKey at <console>:24

scala> wordCount.toDebugString
res13: String =
(2) ShuffledRDD[21] at reduceByKey at <console>:24 []
 +-(2) MapPartitionsRDD[20] at map at <console>:24 []
    |  MapPartitionsRDD[19] at flatMap at <console>:24 []
    |  README.md MapPartitionsRDD[18] at textFile at <console>:24 []
    |  README.md HadoopRDD[17] at textFile at <console>:24 []

scala> val wordCount = sc.textFile("README.md").flatMap(_.split("\\s+")).map((_, 1)).reduceByKey(_ + _)

wordCount: org.apache.spark.rdd.RDD[(String, Int)] = ShuffledRDD[21] at reduceByKey at <console>:24

scala> wordCount.toDebugString

res13: String =

(2) ShuffledRDD[21] at reduceByKey at <console>:24 []

+-(2) MapPartitionsRDD[20] at map at <console>:24 []

| MapPartitionsRDD[19] at flatMap at <console>:24 []

| README.md MapPartitionsRDD[18] at textFile at <console>:24 []

| README.md HadoopRDD[17] at textFile at <console>:24 []

toDebugString uses indentations to indicate a shuffle boundary.

The numbers in round brackets show the level of parallelism at each stage, e.g. (2) in the above output.


scala> wordCount.getNumPartitions
res14: Int = 2

scala> wordCount.getNumPartitions

res14: Int = 2

With spark.logLineage property enabled, toDebugString is included when executing an action.


$ ./bin/spark-shell --conf spark.logLineage=true

scala> sc.textFile("README.md", 4).count
...
15/10/17 14:46:42 INFO SparkContext: Starting job: count at <console>:25
15/10/17 14:46:42 INFO SparkContext: RDD's recursive dependencies:
(4) MapPartitionsRDD[1] at textFile at <console>:25 []
 |  README.md HadoopRDD[0] at textFile at <console>:25 []

$ ./bin/spark-shell --conf spark.logLineage=true

scala> sc.textFile("README.md", 4).count

...

15/10/17 14:46:42 INFO SparkContext: Starting job: count at <console>:25

15/10/17 14:46:42 INFO SparkContext: RDD's recursive dependencies:

(4) MapPartitionsRDD[1] at textFile at <console>:25 []

| README.md HadoopRDD[0] at textFile at <console>:25 []

Settings

Table 1. Spark Properties
Spark Property	Default Value	Description
`spark.logLineage`	`false`	When enabled (i.e. `true`), executing an action (and hence running a job) will also print out the RDD lineage graph using RDD.toDebugString.

RDD Lineage — Logical Execution Plan