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
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
|
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.spark.rdd
import java.util.concurrent.atomic.AtomicLong
import scala.collection.mutable.ArrayBuffer
import scala.concurrent.{ExecutionContext, Future}
import scala.reflect.ClassTag
import org.apache.spark.{ComplexFutureAction, FutureAction, JobSubmitter, Logging}
import org.apache.spark.util.ThreadUtils
/**
* A set of asynchronous RDD actions available through an implicit conversion.
*/
class AsyncRDDActions[T: ClassTag](self: RDD[T]) extends Serializable with Logging {
/**
* Returns a future for counting the number of elements in the RDD.
*/
def countAsync(): FutureAction[Long] = self.withScope {
val totalCount = new AtomicLong
self.context.submitJob(
self,
(iter: Iterator[T]) => {
var result = 0L
while (iter.hasNext) {
result += 1L
iter.next()
}
result
},
Range(0, self.partitions.length),
(index: Int, data: Long) => totalCount.addAndGet(data),
totalCount.get())
}
/**
* Returns a future for retrieving all elements of this RDD.
*/
def collectAsync(): FutureAction[Seq[T]] = self.withScope {
val results = new Array[Array[T]](self.partitions.length)
self.context.submitJob[T, Array[T], Seq[T]](self, _.toArray, Range(0, self.partitions.length),
(index, data) => results(index) = data, results.flatten.toSeq)
}
/**
* Returns a future for retrieving the first num elements of the RDD.
*/
def takeAsync(num: Int): FutureAction[Seq[T]] = self.withScope {
val callSite = self.context.getCallSite
val localProperties = self.context.getLocalProperties
// Cached thread pool to handle aggregation of subtasks.
implicit val executionContext = AsyncRDDActions.futureExecutionContext
val results = new ArrayBuffer[T]
val totalParts = self.partitions.length
/*
Recursively triggers jobs to scan partitions until either the requested
number of elements are retrieved, or the partitions to scan are exhausted.
This implementation is non-blocking, asynchronously handling the
results of each job and triggering the next job using callbacks on futures.
*/
def continue(partsScanned: Long)(implicit jobSubmitter: JobSubmitter) : Future[Seq[T]] =
if (results.size >= num || partsScanned >= totalParts) {
Future.successful(results.toSeq)
} else {
// The number of partitions to try in this iteration. It is ok for this number to be
// greater than totalParts because we actually cap it at totalParts in runJob.
var numPartsToTry = 1L
if (partsScanned > 0) {
// If we didn't find any rows after the previous iteration, quadruple and retry.
// Otherwise, interpolate the number of partitions we need to try, but overestimate it
// by 50%. We also cap the estimation in the end.
if (results.size == 0) {
numPartsToTry = partsScanned * 4
} else {
// the left side of max is >=1 whenever partsScanned >= 2
numPartsToTry = Math.max(1,
(1.5 * num * partsScanned / results.size).toInt - partsScanned)
numPartsToTry = Math.min(numPartsToTry, partsScanned * 4)
}
}
val left = num - results.size
val p = partsScanned.toInt until math.min(partsScanned + numPartsToTry, totalParts).toInt
val buf = new Array[Array[T]](p.size)
self.context.setCallSite(callSite)
self.context.setLocalProperties(localProperties)
val job = jobSubmitter.submitJob(self,
(it: Iterator[T]) => it.take(left).toArray,
p,
(index: Int, data: Array[T]) => buf(index) = data,
Unit)
job.flatMap {_ =>
buf.foreach(results ++= _.take(num - results.size))
continue(partsScanned + p.size)
}
}
new ComplexFutureAction[Seq[T]](continue(0L)(_))
}
/**
* Applies a function f to all elements of this RDD.
*/
def foreachAsync(f: T => Unit): FutureAction[Unit] = self.withScope {
val cleanF = self.context.clean(f)
self.context.submitJob[T, Unit, Unit](self, _.foreach(cleanF), Range(0, self.partitions.length),
(index, data) => Unit, Unit)
}
/**
* Applies a function f to each partition of this RDD.
*/
def foreachPartitionAsync(f: Iterator[T] => Unit): FutureAction[Unit] = self.withScope {
self.context.submitJob[T, Unit, Unit](self, f, Range(0, self.partitions.length),
(index, data) => Unit, Unit)
}
}
private object AsyncRDDActions {
val futureExecutionContext = ExecutionContext.fromExecutorService(
ThreadUtils.newDaemonCachedThreadPool("AsyncRDDActions-future", 128))
}
|
