spark LocalCheckpointRDD 源码
spark LocalCheckpointRDD 代码
文件路径：/core/src/main/scala/org/apache/spark/rdd/LocalCheckpointRDD.scala
/*
* 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 scala.reflect.ClassTag
import org.apache.spark.{Partition, SparkContext, TaskContext}
import org.apache.spark.errors.SparkCoreErrors
import org.apache.spark.storage.RDDBlockId
/**
* A dummy CheckpointRDD that exists to provide informative error messages during failures.
*
* This is simply a placeholder because the original checkpointed RDD is expected to be
* fully cached. Only if an executor fails or if the user explicitly unpersists the original
* RDD will Spark ever attempt to compute this CheckpointRDD. When this happens, however,
* we must provide an informative error message.
*
* @param sc the active SparkContext
* @param rddId the ID of the checkpointed RDD
* @param numPartitions the number of partitions in the checkpointed RDD
*/
private[spark] class LocalCheckpointRDD[T: ClassTag](
sc: SparkContext,
rddId: Int,
numPartitions: Int)
extends CheckpointRDD[T](sc) {
def this(rdd: RDD[T]) = {
this(rdd.context, rdd.id, rdd.partitions.length)
}
protected override def getPartitions: Array[Partition] = {
(0 until numPartitions).toArray.map { i => new CheckpointRDDPartition(i) }
}
/**
* Throw an exception indicating that the relevant block is not found.
*
* This should only be called if the original RDD is explicitly unpersisted or if an
* executor is lost. Under normal circumstances, however, the original RDD (our child)
* is expected to be fully cached and so all partitions should already be computed and
* available in the block storage.
*/
override def compute(partition: Partition, context: TaskContext): Iterator[T] = {
throw SparkCoreErrors.checkpointRDDBlockIdNotFoundError(RDDBlockId(rddId, partition.index))
}
}
相关信息
相关文章
0
赞
- 所属分类： 前端技术
- 本文标签：
热门推荐
-
2、 - 优质文章
-
3、 gate.io
-
8、 golang
-
9、 openharmony
-
10、 Vue中input框自动聚焦