akka-typed(7) - cluster：sharding, 集群分片

用户1150956

发布于 2020-06-16 14:50:03

1K0

发布于 2020-06-16 14:50:03

在使用akka-typed的过程中发现有很多地方都简化了不少，变得更方便了，包括：Supervision，只要用Behaviors.supervise()把Behavior包住，很容易就可以实现这个actor的SupervisorStrategy.restartWithBackoff策略了。然后集群化的group router使用起来也很方便，再就是集群分片cluster-sharding了。下面我们就通过一个例子来介绍cluster-sharding的具体使用方法。

首先，分片的意思是指在集群中多个节点上部署某种actor，即entity，的构建机制。entity的构建是动态的，ClusterSharding系统根据各节点的负载情况决定到底在哪个节点构建entity，然后返回ShardRegion：一个该类entity具体的构建工具及消息中介。也就是说我们可以把同样的一种运算通过entityId指定给任何一个entity，但具体这个entity生存在集群哪个节点上人工是无法确定的，完全靠ClusterSharding引导。先设计一个简单功能的actor，测试它作为一个entity的工作细节：

object Counter {
  sealed trait Command extends CborSerializable
  case object Increment extends Command
  final case class GetValue(replyTo: ActorRef[Response]) extends Command
  case object StopCounter extends Command
  private case object Idle extends Command

  sealed trait Response extends CborSerializable
  case class SubTtl(entityId: String, ttl: Int) extends Response


  val TypeKey = EntityTypeKey[Command]("Counter")

  def apply(nodeAddress: String, entityContext: EntityContext[Command]): Behavior[Command] = {
    Behaviors.setup { ctx =>
      def updated(value: Int): Behavior[Command] = {
        Behaviors.receiveMessage[Command] {
          case Increment =>
            ctx.log.info("******************{} counting at {},{}",ctx.self.path,nodeAddress,entityContext.entityId)
            updated(value + 1)
          case GetValue(replyTo) =>
            ctx.log.info("******************{} get value at {},{}",ctx.self.path,nodeAddress,entityContext.entityId)
            replyTo ! SubTtl(entityContext.entityId,value)
            Behaviors.same
          case Idle =>
            entityContext.shard ! ClusterSharding.Passivate(ctx.self)
            Behaviors.same
          case StopCounter =>
            Behaviors.stopped(() => ctx.log.info("************{} stopping ... passivated for idling.", entityContext.entityId))
        }
      }
      ctx.setReceiveTimeout(30.seconds, Idle)
      updated(0)
    }
  }
}

cluster-sharding的机制是这样的：在每个（或指定的）节点上构建部署一个某种EntityType的ShardRegion。这样系统可以在任何部署了ShardRegion的节点上构建这种entity。然后ClusterSharding系统会根据entityId来引导消息至正确的接收对象。我们再看看ShardRegion的部署是如何实现的吧：

object EntityManager {
  sealed trait Command
  case class AddOne(counterId: String) extends Command
  case class GetSum(counterId: String ) extends Command
  case class WrappedTotal(res: Counter.Response) extends Command


  def apply(): Behavior[Command] = Behaviors.setup { ctx =>
    val cluster = Cluster(ctx.system)
    val sharding = ClusterSharding(ctx.system)
    val entityType = Entity(Counter.TypeKey) { entityContext =>
      Counter(cluster.selfMember.address.toString,entityContext)
    }.withStopMessage(Counter.StopCounter)
    sharding.init(entityType)

    val counterRef: ActorRef[Counter.Response] = ctx.messageAdapter(ref => WrappedTotal(ref))

     Behaviors.receiveMessage[Command] {
      case AddOne(cid) =>
        val entityRef: EntityRef[Counter.Command] = sharding.entityRefFor(Counter.TypeKey, cid)
        entityRef ! Counter.Increment
        Behaviors.same
      case GetSum(cid) =>
         val entityRef: EntityRef[Counter.Command] = sharding.entityRefFor(Counter.TypeKey, cid)
         entityRef ! Counter.GetValue(counterRef)
         Behaviors.same
      case WrappedTotal(ttl) => ttl match {
        case Counter.SubTtl(eid,subttl) =>
          ctx.log.info("***********************{} total: {} ",eid,subttl)
      }
      Behaviors.same
    }
  }

}

太简单了, sharding.ini(entityType)一个函数完成了一个节点分片部署。系统通过sharding.init(entityType)来实现ShardRegion构建。这个entityType代表某种特殊actor模版，看看它的构建函数：

object Entity {

  /**
   * Defines how the entity should be created. Used in [[ClusterSharding#init]]. More optional
   * settings can be defined using the `with` methods of the returned [[Entity]].
   *
   * @param typeKey A key that uniquely identifies the type of entity in this cluster
   * @param createBehavior Create the behavior for an entity given a [[EntityContext]] (includes entityId)
   * @tparam M The type of message the entity accepts
   */
  def apply[M](typeKey: EntityTypeKey[M])(
      createBehavior: EntityContext[M] => Behavior[M]): Entity[M, ShardingEnvelope[M]] =
    new Entity(createBehavior, typeKey, None, Props.empty, None, None, None, None, None)
}

这个函数需要一个EntityTyeKey和一个构建Behavior的函数createBehavior，产生一个Entity类型。Entity类型定义如下：

final class Entity[M, E] private[akka] (
    val createBehavior: EntityContext[M] => Behavior[M],
    val typeKey: EntityTypeKey[M],
    val stopMessage: Option[M],
    val entityProps: Props,
    val settings: Option[ClusterShardingSettings],
    val messageExtractor: Option[ShardingMessageExtractor[E, M]],
    val allocationStrategy: Option[ShardAllocationStrategy],
    val role: Option[String],
    val dataCenter: Option[DataCenter]) {

  /**
   * [[akka.actor.typed.Props]] of the entity actors, such as dispatcher settings.
   */
  def withEntityProps(newEntityProps: Props): Entity[M, E] =
    copy(entityProps = newEntityProps)

  /**
   * Additional settings, typically loaded from configuration.
   */
  def withSettings(newSettings: ClusterShardingSettings): Entity[M, E] =
    copy(settings = Option(newSettings))

  /**
   * Message sent to an entity to tell it to stop, e.g. when rebalanced or passivated.
   * If this is not defined it will be stopped automatically.
   * It can be useful to define a custom stop message if the entity needs to perform
   * some asynchronous cleanup or interactions before stopping.
   */
  def withStopMessage(newStopMessage: M): Entity[M, E] =
    copy(stopMessage = Option(newStopMessage))

  /**
   *
   * If a `messageExtractor` is not specified the messages are sent to the entities by wrapping
   * them in [[ShardingEnvelope]] with the entityId of the recipient actor. That envelope
   * is used by the [[HashCodeMessageExtractor]] for extracting entityId and shardId. The number of
   * shards is then defined by `numberOfShards` in `ClusterShardingSettings`, which by default
   * is configured with `akka.cluster.sharding.number-of-shards`.
   */
  def withMessageExtractor[Envelope](newExtractor: ShardingMessageExtractor[Envelope, M]): Entity[M, Envelope] =
    new Entity(
      createBehavior,
      typeKey,
      stopMessage,
      entityProps,
      settings,
      Option(newExtractor),
      allocationStrategy,
      role,
      dataCenter)

  /**
   * Allocation strategy which decides on which nodes to allocate new shards,
   * [[ClusterSharding#defaultShardAllocationStrategy]] is used if this is not specified.
   */
  def withAllocationStrategy(newAllocationStrategy: ShardAllocationStrategy): Entity[M, E] =
    copy(allocationStrategy = Option(newAllocationStrategy))

  /**
   *  Run the Entity actors on nodes with the given role.
   */
  def withRole(newRole: String): Entity[M, E] = copy(role = Some(newRole))

  /**
   * The data center of the cluster nodes where the cluster sharding is running.
   * If the dataCenter is not specified then the same data center as current node. If the given
   * dataCenter does not match the data center of the current node the `ShardRegion` will be started
   * in proxy mode.
   */
  def withDataCenter(newDataCenter: DataCenter): Entity[M, E] = copy(dataCenter = Some(newDataCenter))

  private def copy(
      createBehavior: EntityContext[M] => Behavior[M] = createBehavior,
      typeKey: EntityTypeKey[M] = typeKey,
      stopMessage: Option[M] = stopMessage,
      entityProps: Props = entityProps,
      settings: Option[ClusterShardingSettings] = settings,
      allocationStrategy: Option[ShardAllocationStrategy] = allocationStrategy,
      role: Option[String] = role,
      dataCenter: Option[DataCenter] = dataCenter): Entity[M, E] = {
    new Entity(
      createBehavior,
      typeKey,
      stopMessage,
      entityProps,
      settings,
      messageExtractor,
      allocationStrategy,
      role,
      dataCenter)
  }
}

这里面有许多方法用来控制Entity的构建和作业。

然后我们把这个EntityManager当作RootBehavior部署到多个节点上去：

object ClusterShardingApp  {
  def main(args: Array[String]): Unit = {
    if (args.isEmpty) {
      startup("shard", 25251)
      startup("shard", 25252)
      startup("shard", 25253)
      startup("front", 25254)
    } else {
      require(args.size == 2, "Usage: role port")
      startup(args(0), args(1).toInt)
    }
  }

  def startup(role: String, port: Int): Unit = {
    // Override the configuration of the port when specified as program argument
    val config = ConfigFactory
      .parseString(s"""
      akka.remote.artery.canonical.port=$port
      akka.cluster.roles = [$role]
      """)
      .withFallback(ConfigFactory.load("cluster"))

    val entityManager = ActorSystem[EntityManager.Command](EntityManager(), "ClusterSystem", config)
...
}

一共设定了3个role=shard节点和1个front节点。

在front节点上对entityId分别为9013，9014，9015，9016几个entity发送消息：

 def startup(role: String, port: Int): Unit = {
    // Override the configuration of the port when specified as program argument
    val config = ConfigFactory
      .parseString(s"""
      akka.remote.artery.canonical.port=$port
      akka.cluster.roles = [$role]
      """)
      .withFallback(ConfigFactory.load("cluster"))

    val entityManager = ActorSystem[EntityManager.Command](EntityManager(), "ClusterSystem", config)
    if (role == "front") {
      entityManager ! EntityManager.AddOne("9013")
      entityManager ! EntityManager.AddOne("9014")
      entityManager ! EntityManager.AddOne("9013")
      entityManager ! EntityManager.AddOne("9015")
      entityManager ! EntityManager.AddOne("9013")
      entityManager ! EntityManager.AddOne("9014")
      entityManager ! EntityManager.AddOne("9014")
      entityManager ! EntityManager.AddOne("9013")
      entityManager ! EntityManager.AddOne("9015")
      entityManager ! EntityManager.AddOne("9015")
      entityManager ! EntityManager.AddOne("9016")
      entityManager ! EntityManager.GetSum("9014")
      entityManager ! EntityManager.GetSum("9015")
      entityManager ! EntityManager.GetSum("9013")
      entityManager ! EntityManager.GetSum("9016")
    }

以下是部分运算结果显示：

15:12:10.073 [ClusterSystem-akka.actor.default-dispatcher-15] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/786/9014 counting at akka://ClusterSystem@127.0.0.1:25253,9014
15:12:10.106 [ClusterSystem-akka.actor.default-dispatcher-15] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/786/9014 counting at akka://ClusterSystem@127.0.0.1:25253,9014
15:12:10.106 [ClusterSystem-akka.actor.default-dispatcher-15] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/786/9014 counting at akka://ClusterSystem@127.0.0.1:25253,9014
15:12:10.106 [ClusterSystem-akka.actor.default-dispatcher-3] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/785/9013 counting at akka://ClusterSystem@127.0.0.1:25251,9013
15:12:10.107 [ClusterSystem-akka.actor.default-dispatcher-3] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/785/9013 counting at akka://ClusterSystem@127.0.0.1:25251,9013
15:12:10.107 [ClusterSystem-akka.actor.default-dispatcher-3] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/785/9013 counting at akka://ClusterSystem@127.0.0.1:25251,9013
15:12:10.107 [ClusterSystem-akka.actor.default-dispatcher-3] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/785/9013 counting at akka://ClusterSystem@127.0.0.1:25251,9013
15:12:10.109 [ClusterSystem-akka.actor.default-dispatcher-19] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/787/9015 counting at akka://ClusterSystem@127.0.0.1:25254,9015
15:12:10.110 [ClusterSystem-akka.actor.default-dispatcher-19] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/787/9015 counting at akka://ClusterSystem@127.0.0.1:25254,9015
15:12:10.110 [ClusterSystem-akka.actor.default-dispatcher-19] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/787/9015 counting at akka://ClusterSystem@127.0.0.1:25254,9015
15:12:10.110 [ClusterSystem-akka.actor.default-dispatcher-19] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/787/9015 get value at akka://ClusterSystem@127.0.0.1:25254,9015
15:12:10.112 [ClusterSystem-akka.actor.default-dispatcher-18] INFO com.learn.akka.EntityManager$ - ***********************9015 total: 3
15:12:10.149 [ClusterSystem-akka.actor.default-dispatcher-15] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/786/9014 get value at akka://ClusterSystem@127.0.0.1:25253,9014
15:12:10.149 [ClusterSystem-akka.actor.default-dispatcher-3] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/785/9013 get value at akka://ClusterSystem@127.0.0.1:25251,9013
15:12:10.169 [ClusterSystem-akka.actor.default-dispatcher-18] INFO com.learn.akka.EntityManager$ - ***********************9014 total: 3
15:12:10.169 [ClusterSystem-akka.actor.default-dispatcher-18] INFO com.learn.akka.EntityManager$ - ***********************9013 total: 4
15:12:10.171 [ClusterSystem-akka.actor.default-dispatcher-3] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/788/9016 counting at akka://ClusterSystem@127.0.0.1:25251,9016
15:12:10.171 [ClusterSystem-akka.actor.default-dispatcher-3] INFO com.learn.akka.Counter$ - ******************akka://ClusterSystem/system/sharding/Counter/788/9016 get value at akka://ClusterSystem@127.0.0.1:25251,9016
15:12:10.172 [ClusterSystem-akka.actor.default-dispatcher-18] INFO com.learn.akka.EntityManager$ - ***********************9016 total: 1

15:19:32.176 [ClusterSystem-akka.actor.default-dispatcher-3] INFO com.learn.akka.Counter$ - ************9013 stopping ... passivated for idling.
15:19:52.529 [ClusterSystem-akka.actor.default-dispatcher-3] INFO com.learn.akka.Counter$ - ************9014 stopping ... passivated for idling.
15:19:52.658 [ClusterSystem-akka.actor.default-dispatcher-3] INFO com.learn.akka.Counter$ - ************9016 stopping ... passivated for idling.
15:19:52.662 [ClusterSystem-akka.actor.default-dispatcher-14] INFO com.learn.akka.Counter$ - ************9015 stopping ... passivated for idling.

下面是本次示范的完整源代码：

ClusterSharding.scala

package com.learn.akka
import scala.concurrent.duration._
import akka.actor.typed._
import akka.actor.typed.scaladsl._
import akka.cluster.sharding.typed.scaladsl.EntityContext
import akka.cluster.sharding.typed.scaladsl.Entity
import akka.persistence.typed.PersistenceId
//#sharding-extension
import akka.cluster.sharding.typed.ShardingEnvelope
import akka.cluster.sharding.typed.scaladsl.ClusterSharding
import akka.cluster.sharding.typed.scaladsl.EntityTypeKey
import akka.cluster.sharding.typed.scaladsl.EntityRef
import com.typesafe.config.ConfigFactory
import akka.cluster.typed.Cluster
//#counter
object Counter {
  sealed trait Command extends CborSerializable
  case object Increment extends Command
  final case class GetValue(replyTo: ActorRef[Response]) extends Command
  case object StopCounter extends Command
  private case object Idle extends Command

  sealed trait Response extends CborSerializable
  case class SubTtl(entityId: String, ttl: Int) extends Response


  val TypeKey = EntityTypeKey[Command]("Counter")

  def apply(nodeAddress: String, entityContext: EntityContext[Command]): Behavior[Command] = {
    Behaviors.setup { ctx =>
      def updated(value: Int): Behavior[Command] = {
        Behaviors.receiveMessage[Command] {
          case Increment =>
            ctx.log.info("******************{} counting at {},{}",ctx.self.path,nodeAddress,entityContext.entityId)
            updated(value + 1)
          case GetValue(replyTo) =>
            ctx.log.info("******************{} get value at {},{}",ctx.self.path,nodeAddress,entityContext.entityId)
            replyTo ! SubTtl(entityContext.entityId,value)
            Behaviors.same
          case Idle =>
            entityContext.shard ! ClusterSharding.Passivate(ctx.self)
            Behaviors.same
          case StopCounter =>
            Behaviors.stopped(() => ctx.log.info("************{} stopping ... passivated for idling.", entityContext.entityId))
        }
      }
      ctx.setReceiveTimeout(30.seconds, Idle)
      updated(0)
    }
  }
}
object EntityManager {
  sealed trait Command
  case class AddOne(counterId: String) extends Command
  case class GetSum(counterId: String ) extends Command
  case class WrappedTotal(res: Counter.Response) extends Command


  def apply(): Behavior[Command] = Behaviors.setup { ctx =>
    val cluster = Cluster(ctx.system)
    val sharding = ClusterSharding(ctx.system)
    val entityType = Entity(Counter.TypeKey) { entityContext =>
      Counter(cluster.selfMember.address.toString,entityContext)
    }.withStopMessage(Counter.StopCounter)
    sharding.init(entityType)

    val counterRef: ActorRef[Counter.Response] = ctx.messageAdapter(ref => WrappedTotal(ref))

     Behaviors.receiveMessage[Command] {
      case AddOne(cid) =>
        val entityRef: EntityRef[Counter.Command] = sharding.entityRefFor(Counter.TypeKey, cid)
        entityRef ! Counter.Increment
        Behaviors.same
      case GetSum(cid) =>
         val entityRef: EntityRef[Counter.Command] = sharding.entityRefFor(Counter.TypeKey, cid)
         entityRef ! Counter.GetValue(counterRef)
         Behaviors.same
      case WrappedTotal(ttl) => ttl match {
        case Counter.SubTtl(eid,subttl) =>
          ctx.log.info("***********************{} total: {} ",eid,subttl)
      }
      Behaviors.same
    }
  }

}

object ClusterShardingApp  {
  def main(args: Array[String]): Unit = {
    if (args.isEmpty) {
      startup("shard", 25251)
      startup("shard", 25252)
      startup("shard", 25253)
      startup("front", 25254)
    } else {
      require(args.size == 2, "Usage: role port")
      startup(args(0), args(1).toInt)
    }
  }

  def startup(role: String, port: Int): Unit = {
    // Override the configuration of the port when specified as program argument
    val config = ConfigFactory
      .parseString(s"""
      akka.remote.artery.canonical.port=$port
      akka.cluster.roles = [$role]
      """)
      .withFallback(ConfigFactory.load("cluster"))

    val entityManager = ActorSystem[EntityManager.Command](EntityManager(), "ClusterSystem", config)
    if (role == "front") {
      entityManager ! EntityManager.AddOne("9013")
      entityManager ! EntityManager.AddOne("9014")
      entityManager ! EntityManager.AddOne("9013")
      entityManager ! EntityManager.AddOne("9015")
      entityManager ! EntityManager.AddOne("9013")
      entityManager ! EntityManager.AddOne("9014")
      entityManager ! EntityManager.AddOne("9014")
      entityManager ! EntityManager.AddOne("9013")
      entityManager ! EntityManager.AddOne("9015")
      entityManager ! EntityManager.AddOne("9015")
      entityManager ! EntityManager.AddOne("9016")
      entityManager ! EntityManager.GetSum("9014")
      entityManager ! EntityManager.GetSum("9015")
      entityManager ! EntityManager.GetSum("9013")
      entityManager ! EntityManager.GetSum("9016")
    }

  }

}

cluster.conf

akka {
  actor {
    provider = cluster

    serialization-bindings {
      "com.learn.akka.CborSerializable" = jackson-cbor
    }
  }
  remote {
    artery {
      canonical.hostname = "127.0.0.1"
      canonical.port = 0
    }
  }
  cluster {
    seed-nodes = [
      "akka://ClusterSystem@127.0.0.1:25251",
      "akka://ClusterSystem@127.0.0.1:25252"]
  }
}

本文参与腾讯云自媒体同步曝光计划，分享自作者个人站点/博客。

原始发表：2020-06-12 ，如有侵权请联系 cloudcommunity@tencent.com 删除

actor