Spark Streaming no receivers彻底思考

数据接入Spark Streaming的二种方式：Receiver和no receivers方式

建议企业级采用no receivers方式开发Spark Streaming应用程序，好处：

1、更优秀的自由度控制

2、语义一致性

no receivers更符合数据读取和数据操作，Spark 计算框架底层有数据来源，如果只有direct直接操作数据来源则更天然。操作数据来源封装其一定是rdd级别的。

所以Spark 推出了自定义的rdd即Kafkardd，只是数据来源不同。

进入源码区：

注释基于Batch消费数据，首先确定开始和结束的offSet，特别强调语义一致性。

关键是metaData.broker.list，通过bootstrap.servers直接操作Kafka集群，操作kafka数据是一个offset范围。

/**
* A batch-oriented interface for
consuming from Kafka.
* Starting and ending offsets are
specified in advance,
* so that you can control exactly-once
semantics.
* @param kafkaParams Kafka <a href="http://kafka.apache.org/documentation.html#configuration">
* configuration parameters</a>. Requires "metadata.broker.list" or
"bootstrap.servers" to be set
* with Kafka broker(s) specified in
host1:port1,host2:port2 form.
* @param offsetRanges offset
ranges that define the Kafka data belonging to this RDD
* @param messageHandler function
for translating each message into the desired type
*/
private[kafka]
class KafkaRDD[
K:
ClassTag,
V:
ClassTag,
U <:
Decoder[_]: ClassTag,
T <:
Decoder[_]: ClassTag,
R:
ClassTag] private[spark] (
    sc: SparkContext,
    kafkaParams: Map[String, String],
    val offsetRanges: Array[OffsetRange],
  leaders: Map[TopicAndPartition, (String, Int)],
    messageHandler: MessageAndMetadata[K, V] => R
) extends
RDD[R](sc, Nil) with Logging with
HasOffsetRanges {

你要直接访问Kafka中的数据需要自定义一个KafkaRDD，如果读取hBase上的数据

也必须自定义一个hBaseRDD。有一点必须定义接口HasOffsetRange，RDD天然的是一个

A List Partitions，基于kafka直接访问RDD时必须是HasOffsetRange类型，代表了

来自kafka topicAndParttion，其实力被HasOffsetRange Create创建，从fromOffset到untilOffset ，

分布式传输Offset数据时必须序列化。

/**
* Represents any object that has a
collection of [[OffsetRange]]s. This can be used to access the
* offset ranges in RDDs generated by the
direct Kafka DStream (see
* [[KafkaUtils.createDirectStream()]]).
* {{{
*
KafkaUtils.createDirectStream(...).foreachRDD { rdd =>
*
val offsetRanges = rdd.asInstanceOf[HasOffsetRanges].offsetRanges
*
...
*
}
* }}}
*/
trait HasOffsetRanges
{
def
offsetRanges: Array[OffsetRange]
}

/**
* Represents a range of offsets from a
single Kafka TopicAndPartition. Instances of this class
* can be created with `OffsetRange.create()`.
* @param topic Kafka topic name
* @param partition Kafka
partition id
* @param fromOffset Inclusive
starting offset
* @param untilOffset Exclusive
ending offset
*/
final class OffsetRange
private(
    val
topic: String,
    val partition:
Int,
    val fromOffset:
Long,
    val untilOffset:
Long) extends Serializable {
import
OffsetRange.OffsetRangeTuple

/**
Kafka TopicAndPartition object, for convenience */
def
topicAndPartition(): TopicAndPartition = TopicAndPartition(topic, partition)

/**
Number of messages this OffsetRange refers to */
def
count(): Long = untilOffset - fromOffset

Offset是消息偏移量，假设untilOffset是10万，fromOffset是5万，第10万条消息

和5万条消息，一般处理数据规模大小是以数据条数为单位。

创建一个offSetrange实例时可以确定从kafka集群partition中读取哪些topic，从

foreachrdd中可以获得当前rdd访问的所有分区数据。Batch Duration中产生的rdd的分区数据，这个是对元数据的控制。

在看getPartitions方法，offsetRanges指定了每个offsetrange从什么位置开始到什么位置结束。

override def getPartitions: Array[Partition] = {
offsetRanges.zipWithIndex.map { case (o, i)
=>
val
(host, port) = leaders(TopicAndPartition(o.topic, o.partition))
new
KafkaRDDPartition(i, o.topic, o.partition, o.fromOffset, o.untilOffset, host, port)

}.toArray
}

看KafkaRDDPartition类，会从传入的topic和partition及offset中获取kafka数据

/** @param topic kafka topic name
* @param partition kafka
partition id
* @param fromOffset inclusive
starting offset
* @param untilOffset exclusive
ending offset
* @param host preferred kafka
host, i.e. the leader at the time the rdd was created
* @param port preferred kafka
host's port
*/
private[kafka]
class KafkaRDDPartition(
val index:
Int,
val topic: String,
val partition: Int,
val fromOffset: Long,
val untilOffset: Long,
val host: String,
val port: Int
) extends Partition {
/** Number of messages this partition refers to */
def count(): Long =
untilOffset - fromOffset
}

Host port指定读取数据来源的kfakf机器。

看kafka rdd的compute计算每个数据分片，和rdd理念是一样的，每次迭代操作获取计算的rdd一部分。

操作KafkaRDDIterator和操作rdd分片是一样的，需要迭代数据分片：

override def compute(thePart: Partition, context: TaskContext): Iterator[R] = {
val part
= thePart.asInstanceOf[KafkaRDDPartition]
assert(part.fromOffset <=
part.untilOffset, errBeginAfterEnd(part))
if (part.fromOffset
== part.untilOffset) {
    log.info(s"Beginning offset ${part.fromOffset} is the same as ending offset " +
      s"skipping ${part.topic}
${part.partition}")
    Iterator.empty
} else
{
    new KafkaRDDIterator(part, context)
}
}

private class KafkaRDDIterator(
part: KafkaRDDPartition,
context: TaskContext) extends NextIterator[R] {

context.addTaskCompletionListener{
context => closeIfNeeded() }

log.info(s"Computing topic ${part.topic}, partition ${part.partition}
" +
s"offsets ${part.fromOffset} -> ${part.untilOffset}")

val kc = new KafkaCluster(kafkaParams)
val keyDecoder
= classTag[U].runtimeClass.getConstructor(classOf[VerifiableProperties])
    .newInstance(kc.config.props)
    .asInstanceOf[Decoder[K]]
val valueDecoder
= classTag[T].runtimeClass.getConstructor(classOf[VerifiableProperties])

.newInstance(kc.config.props)
    .asInstanceOf[Decoder[V]]
val consumer
= connectLeader
var requestOffset
= part.fromOffset
var iter: Iterator[MessageAndOffset]
= null

// The idea is to use the provided preferred host, except
on task retry attempts,
// to minimize number of kafka metadata
requests
private def connectLeader: SimpleConsumer
= {
    if (context.attemptNumber
> 0) {
      kc.connectLeader(part.topic, part.partition).fold(
        errs => throw new SparkException(
          s"Couldn't connect to leader for topic ${part.topic} ${part.partition}: " +
            errs.mkString("\n")),
        consumer => consumer
      )
    } else {
      kc.connect(part.host, part.port)
    }
}

private def handleFetchErr(resp:
FetchResponse) {
    if (resp.hasError)
{
      val err
= resp.errorCode(part.topic, part.partition)
      if (err
== ErrorMapping.LeaderNotAvailableCode
||
        err == ErrorMapping.NotLeaderForPartitionCode) {
        log.error(s"Lost leader for topic ${part.topic} partition ${part.partition}, " +
          s" sleeping for ${kc.config.refreshLeaderBackoffMs}ms")
        Thread.sleep(kc.config.refreshLeaderBackoffMs)
      }
      // Let normal rdd retry sort out reconnect attempts
      throw ErrorMapping.exceptionFor(err)
    }
}

关键的地方kafkaCluster对象时在kafkaUtils中直接创建了directStream，看下之前操作kafka代码发现传入的参数是上下文、 broker.List.topic.list参数：

Spark Streaming no receivers彻底思考-LMLPHP

构建时传入topics为Set，当然可以直接指定ranges，他从kafka集群直接创建了kafkaCluster和集群进行交互，从fromOffset获取数据具体的偏移量：

/**
* Create an input stream that directly
pulls messages from Kafka Brokers
* without using any receiver. This
stream can guarantee that each message
* from Kafka is included in
transformations exactly once (see points below).
*
* Points to note:
*
- No receivers: This stream does not use any receiver. It directly
queries Kafka
*
- Offsets: This does not use Zookeeper to store offsets. The consumed
offsets are tracked
*
by the stream itself. For interoperability with Kafka monitoring tools
that depend on
*
Zookeeper, you have to update Kafka/Zookeeper yourself from the
streaming application.
*
You can access the offsets used in each batch from the generated RDDs
(see
*
[[org.apache.spark.streaming.kafka.HasOffsetRanges]]).
*
- Failure Recovery: To recover from driver failures, you have to enable
checkpointing
*
in the [[StreamingContext]]. The information on consumed offset can be
*
recovered from the checkpoint. See the programming guide for details
(constraints, etc.).
*
- End-to-end semantics: This stream ensures that every records is
effectively received and
*
transformed exactly once, but gives no guarantees on whether the
transformed data are
*
outputted exactly once. For end-to-end exactly-once semantics, you have
to either ensure
*
that the output operation is idempotent, or use transactions to output
records atomically.
*
See the programming guide for more details.
*
* @param ssc StreamingContext
object
* @param kafkaParams Kafka <a href="http://kafka.apache.org/documentation.html#configuration">
*
configuration parameters</a>. Requires "metadata.broker.list" or
"bootstrap.servers"
*
to be set with Kafka broker(s) (NOT zookeeper servers), specified in
*
host1:port1,host2:port2 form.
*
If not starting from a checkpoint, "auto.offset.reset" may be
set to "largest" or "smallest"
*
to determine where the stream starts (defaults to "largest")
* @param topics Names of the
topics to consume
* @tparam K type of Kafka message
key
* @tparam V type of Kafka message
value
* @tparam KD type of Kafka message
key decoder
* @tparam VD type of Kafka
message value decoder
* @return DStream of (Kafka
message key, Kafka message value)
*/
def createDirectStream[
K: ClassTag,
V: ClassTag,
KD <: Decoder[K]:
ClassTag,
VD <: Decoder[V]:
ClassTag] (
    ssc: StreamingContext,
    kafkaParams:
Map[String, String],
    topics:
Set[String]
): InputDStream[(K, V)] = {
val
messageHandler = (mmd: MessageAndMetadata[K, V]) => (mmd.key, mmd.message)
val
kc = new KafkaCluster(kafkaParams)
val
fromOffsets = getFromOffsets(kc, kafkaParams, topics)
new
DirectKafkaInputDStream[K, V, KD, VD, (K, V)](
    ssc, kafkaParams, fromOffsets, messageHandler)

看下getFromOffsets方法：

private[kafka] def getFromOffsets(
    kc: KafkaCluster,
    kafkaParams: Map[String, String],
    topics: Set[String]
): Map[TopicAndPartition, Long] = {
val reset
= kafkaParams.get("auto.offset.reset").map(_.toLowerCase)
val result
= for {
    topicPartitions <-
kc.getPartitions(topics).right
    leaderOffsets <- (if (reset == Some("smallest")) {

kc.getEarliestLeaderOffsets(topicPartitions)
    } else {

kc.getLatestLeaderOffsets(topicPartitions)
    }).right
} yield {
    leaderOffsets.map { case (tp, lo) =>
        (tp, lo.offset)
    }
}
KafkaCluster.checkErrors(result)
}

如果不知道fromOffsets的话直接从配置中获取fromOffsets，创建kafka DirectKafkaInputDStream的时候会从kafka集群进行交互获得partition、offset信息，通过DirectKafkaInputDStream无论什么情况最后都会创建DirectKafkaInputDStream：

/**
*
A stream of {@link org.apache.spark.streaming.kafka.KafkaRDD}
where
* each given Kafka topic/partition
corresponds to an RDD partition.
* The spark configuration
spark.streaming.kafka.maxRatePerPartition gives the maximum number
*
of messages
* per second that each '''partition''' will accept.
* Starting offsets are specified in
advance,
* and this DStream is not responsible
for committing offsets,
* so that you can control exactly-once
semantics.
* For an easy interface to Kafka-managed
offsets,
*
see {@link org.apache.spark.streaming.kafka.KafkaCluster}
* @param kafkaParams Kafka <a href="http://kafka.apache.org/documentation.html#configuration">
* configuration parameters</a>.
*
Requires "metadata.broker.list" or
"bootstrap.servers" to be set with Kafka broker(s),
*
NOT zookeeper servers, specified in host1:port1,host2:port2 form.
* @param fromOffsets per-topic/partition
Kafka offsets defining the (inclusive)
*
starting point of the stream
* @param messageHandler function
for translating each message into the desired type
*/
private[streaming]
class DirectKafkaInputDStream[
K:
ClassTag,
V:
ClassTag,
U <:
Decoder[K]: ClassTag,
T <:
Decoder[V]: ClassTag,
R:
ClassTag](
    ssc_ : StreamingContext,
    val kafkaParams: Map[String, String],
    val fromOffsets: Map[TopicAndPartition, Long],
    messageHandler: MessageAndMetadata[K, V] => R
) extends
InputDStream[R](ssc_)
with Logging {
val maxRetries
= context.sparkContext.getConf.getInt(
    "spark.streaming.kafka.maxRetries", 1)

DirectKafkaInputDStream会产生kafkaRDD，不同的topic partitions生成对应的的kafkarddpartitions，控制消费读取速度。操作数据的时候是compute直接构建出kafka rdd，读取kafka 上的数据。确定获取读取数据的期间就知道需要读取多少条数据，然后构建kafkardd实例。Kafkardd的实例和DirectKafkaInputDStream是一一对应的，每次compute会产生一个kafkardd，其会包含很多partitions，有多少partition就是对应多少kafkapartition。

看下KafkaRDDPartition就是一个简单的数据结构：

总结：

而且KafkaRDDPartition只能属于一个topic，不能让partition跨多个topic，直接消费一个kafkatopic，topic不断进来、数据不断偏移，Offset代表kafka数据偏移量指针。

数据不断流进kafka，batchDuration假如每十秒都会从配置的topic中消费数据，每次会消费一部分直到消费完，下一个batchDuration会再流进来的数据，又可以从头开始读或上一个数据的基础上读取数据。

思考直接抓取kafka数据和receiver读取数据：

好处一：

直接抓取fakfa数据的好处，没有缓存，不会出现内存溢出等之类的问题。但是如果kafka Receiver的方式读取会存在缓存的问题，需要设置读取的频率和block interval等信息。

好处二：

采用receiver方式的话receiver默认情况需要和worker的executor绑定，不方便做分布式，当然可以配置成分布式，采用direct方式默认情况下数据会存在多个worker上的executor。Kafkardd数据默认都是分布在多个executor上的，天然数据是分布式的存在多个executor，而receiver就不方便计算。

好处三：

数据消费的问题，在实际操作的时候采用receiver的方式有个弊端，消费数据来不及处理即操作数据有deLay多才时，Spark Streaming程序有可能奔溃。但如果是direct方式访问kafka数据不会存在此类情况。因为diect方式直接读取kafka数据，如果delay就不进行下一个batchDuration读取。

好处四：

完全的语义一致性，不会重复消费数据，而且保证数据一定被消费，跟kafka进行交互，只有数据真正执行成功之后才会记录下来。

生产环境下强烈建议采用direct方式读取kafka数据。

新浪微博：http://weibo.com/ilovepains

微信公众号：DT_Spark

博客：http://blog.sina.com.cn/ilovepains

手机：18610086859

QQ：1740415547

邮箱：[email protected]

Spark Streaming发行版笔记15