Akka 2.4 M2发布,Actor开发模型库

Akka 2.4 M2 发布，此版本包括一些 2.4-M1 的改进：

• incorporate Patrik’s akka-data-replication under the new module name akka-distributed-data-experimental (see akka-data-replication migration guide), the Distributed Data module is useful for sharing eventually consistent data (CRDTs) between nodes in an Akka Cluster

• improved support for Akka Persistence event migrations by Serializer with String Manifest and Event Adapters

• make Akka Persistence failure handling more robust and support rejections of events

• Persistent FSM, thanks to an awesome contribution from leonidb

• various improvements in Akka Persistence aiming to finalize its APIs, refer to the migration guide or closed persistence tickets for details

• use custom failure detection in Cluster Client to avoid quarantining in case of transient network failures

• support graceful shutdown in Cluster Sharding

• various API and configuration improvements of Cluster Sharding, Cluster Singleton, Distributed PubSub and Cluster Client

Akka 2.4 未来计划

Akka 2.4.0-RC1 将会包括的新特性：

• complete the Java API and the documentation for Akka Typed

• finalize Akka Persistence on the Write Side (i.e. PersistentActor will be non-experimental; the work on the Read Side will be deferred until Akka 2.5)

• incorporate Streams & HTTP once ready (to become an experimental module, as usual)

更多改进内容请看发行说明。

Actor模型并非什么新鲜事物，它由Carl Hewitt于上世纪70年代早期提出，目的是为了解决分布式编程中一系列的编程问题。其特点如下：

• 系统中的所有事物都可以扮演一个Actor

• Actor之间完全独立

• 在收到消息时Actor所采取的所有动作都是并行的，在一个方法中的动作没有明确的顺序

• Actor由标识和当前行为描述

• Actor可能被分成原始（primitive）和非原始（non primitive）类别

• 非原始Actor有

• 由一个邮件地址表示的标识

• 当前行为由一组知识（acquaintances）（实例变量或本地状态）和定义Actor在收到消息时将采取的动作组成

• 消息传递是非阻塞和异步的，其机制是邮件队列（mail-queue）

• 所有消息发送都是并行的