spark RateLimitedOutputStream 源码

spark RateLimitedOutputStream 代码

文件路径：/streaming/src/main/scala/org/apache/spark/streaming/util/RateLimitedOutputStream.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.streaming.util

import java.io.OutputStream
import java.util.concurrent.TimeUnit._

import scala.annotation.tailrec

import org.apache.spark.internal.Logging

private[streaming]
class RateLimitedOutputStream(out: OutputStream, desiredBytesPerSec: Int)
  extends OutputStream
  with Logging {

  require(desiredBytesPerSec > 0)

  private val SYNC_INTERVAL = NANOSECONDS.convert(10, SECONDS)
  private val CHUNK_SIZE = 8192
  private var lastSyncTime = System.nanoTime
  private var bytesWrittenSinceSync = 0L

  override def write(b: Int): Unit = {
    waitToWrite(1)
    out.write(b)
  }

  override def write(bytes: Array[Byte]): Unit = {
    write(bytes, 0, bytes.length)
  }

  @tailrec
  override final def write(bytes: Array[Byte], offset: Int, length: Int): Unit = {
    val writeSize = math.min(length - offset, CHUNK_SIZE)
    if (writeSize > 0) {
      waitToWrite(writeSize)
      out.write(bytes, offset, writeSize)
      write(bytes, offset + writeSize, length)
    }
  }

  override def flush(): Unit = {
    out.flush()
  }

  override def close(): Unit = {
    out.close()
  }

  @tailrec
  private def waitToWrite(numBytes: Int): Unit = {
    val now = System.nanoTime
    val elapsedNanosecs = math.max(now - lastSyncTime, 1)
    val rate = bytesWrittenSinceSync.toDouble * 1000000000 / elapsedNanosecs
    if (rate < desiredBytesPerSec) {
      // It's okay to write; just update some variables and return
      bytesWrittenSinceSync += numBytes
      if (now > lastSyncTime + SYNC_INTERVAL) {
        // Sync interval has passed; let's resync
        lastSyncTime = now
        bytesWrittenSinceSync = numBytes
      }
    } else {
      // Calculate how much time we should sleep to bring ourselves to the desired rate.
      val targetTimeInMillis = bytesWrittenSinceSync * 1000 / desiredBytesPerSec
      val elapsedTimeInMillis = NANOSECONDS.toMillis(elapsedNanosecs)
      val sleepTimeInMillis = targetTimeInMillis - elapsedTimeInMillis
      if (sleepTimeInMillis > 0) {
        logTrace("Natural rate is " + rate + " per second but desired rate is " +
          desiredBytesPerSec + ", sleeping for " + sleepTimeInMillis + " ms to compensate.")
        Thread.sleep(sleepTimeInMillis)
      }
      waitToWrite(numBytes)
    }
  }
}

相关信息

spark 源码目录

相关文章

spark BatchedWriteAheadLog 源码

spark FileBasedWriteAheadLog 源码

spark FileBasedWriteAheadLogRandomReader 源码

spark FileBasedWriteAheadLogReader 源码

spark FileBasedWriteAheadLogSegment 源码

spark FileBasedWriteAheadLogWriter 源码

spark HdfsUtils 源码

spark RawTextHelper 源码

spark RawTextSender 源码

spark RecurringTimer 源码