Hi,
I am using AKKA 2.3.0 with Scala 2.10.3
I am trying a basic 2 node cluster exercise in Scala using Akka. Where I
would like each node to have a cluster monitor and print console messages
on member join. So that I know, there is a cluster.
Just like the example provided in docs, reactive etc. Simple enough!
Let me first present what I have and then I will come to my
questions/confusions.
*(1) Source code of the node and monitor class *
object Platform {
final val NAME = "concenter";
}
class Platform extends Bootable {
private val system = ActorSystem( Platform.NAME );
system.actorOf(Props[ClusterMonitor], name = "clusterListener");
def startup = {
System.out.println("Greetings Earthling!");
}
def shutdown = {
system.shutdown();
System.out.println("Goodbye Blue Sky, Goodbye!");
}
}
class ClusterMonitor extends Actor with ActorLogging {
private val cluster = Cluster(context.system)
override def preStart(): Unit = {
System.out.println("Subscribing to Cluster Event");
cluster.subscribe(self, initialStateMode = InitialStateAsEvents,
classOf[MemberEvent], classOf[UnreachableMember],
classOf[ClusterDomainEvent] );
}
override def postStop(): Unit = {
cluster.unsubscribe(self);
System.out.println("Unsubscribed from Cluster Event");
}
def receive = {
case MemberUp(member) => log.info(">>>>>>>>>>> Member is Up: {}",
member.address)
case UnreachableMember(member) => log.info(">>>>>>>>>>> Member
detected as unreachable: {}", member)
case MemberRemoved(member, previousStatus) => log.info(">>>>>>>>>>>
Member is Removed: {} after {}", member.address, previousStatus)
case _: MemberEvent => log.info( "XXXXXXX" );
case cdev: ClusterDomainEvent => { log.info( "YYYYY: " + cdev.toString()
); }
}
}
(*2) Node 1 configuration. This node is always explicitly started first.*
akka {
//STDOUT logging for Debugging
loggers = ["akka.event.Logging$DefaultLogger"]
actor {
provider = "akka.cluster.ClusterActorRefProvider"
}
remote {
netty.tcp {
hostname = "127.0.0.1"
port = 9000
}
}
cluster {
log-info = on
}
}
(3*) Node 2 configuration. This node is always explicitly started after
Node 1.*
akka {
//STDOUT logging for Debugging
loggers = ["akka.event.Logging$DefaultLogger"]
actor {
provider = "akka.cluster.ClusterActorRefProvider"
}
remote {
netty.tcp {
hostname = "127.0.0.1"
port = 9090
}
}
cluster {
log-info = on
seed-nodes = [ "akka.tcp://[email protected]:9000" ]
}
}
*(4) An execution session, I have the following output from Node 1 ( Node
2 is started and taken down )*
Starting Akka...
Running Akka 2.3.0
[warning] Akka home is not defined
[DEBUG] [03/26/2014 12:22:32.446] [main] [EventStream(akka://concenter)]
logger log1-Logging$DefaultLogger started
[DEBUG] [03/26/2014 12:22:32.447] [main] [EventStream(akka://concenter)]
Default Loggers started
[INFO] [03/26/2014 12:22:32.497] [main] [Remoting] Starting remoting
[INFO] [03/26/2014 12:22:32.695] [main] [Remoting] Remoting started;
listening on addresses :[akka.tcp://[email protected]:9000]
[INFO] [03/26/2014 12:22:32.696] [main] [Remoting] Remoting now listens on
addresses: [akka.tcp://[email protected]:9000]
[INFO] [03/26/2014 12:22:32.708] [main] [Cluster(akka://concenter)] Cluster
Node [akka.tcp://[email protected]:9000] - Starting up...
[INFO] [03/26/2014 12:22:32.784] [main] [Cluster(akka://concenter)] Cluster
Node [akka.tcp://[email protected]:9000] - Registered cluster JMX MBean
[akka:type=Cluster]
[INFO] [03/26/2014 12:22:32.784] [main] [Cluster(akka://concenter)] Cluster
Node [akka.tcp://[email protected]:9000] - Started up successfully
[INFO] [03/26/2014 12:22:32.789]
[concenter-akka.actor.default-dispatcher-3] [Cluster(akka://concenter)]
Cluster Node [akka.tcp://[email protected]:9000] - Metrics will be
retreived from MBeans, and may be incorrect on some platforms. To increase
metric accuracy add the 'sigar.jar' to the classpath and the appropriate
platform-specific native libary to 'java.library.path'. Reason:
java.lang.ClassNotFoundException: org.hyperic.sigar.Sigar
[INFO] [03/26/2014 12:22:32.789]
[concenter-akka.actor.default-dispatcher-14] [Cluster(akka://concenter)]
Cluster Node [akka.tcp://[email protected]:9000] - No seed-nodes
configured, manual cluster join required
[INFO] [03/26/2014 12:22:32.790]
[concenter-akka.actor.default-dispatcher-3] [Cluster(akka://concenter)]
Cluster Node [akka.tcp://[email protected]:9000] - Metrics collection has
started successfully
Starting up concenter.Platform
Greetings Earthling!
Subscribing to Cluster Event
Successfully started Akka
[INFO] [03/26/2014 12:22:35.809]
[concenter-akka.actor.default-dispatcher-21]
[akka.tcp://[email protected]:9000/user/clusterListener] YYYYY:
ClusterMetricsChanged(Set(NodeMetrics(akka.tcp://[email protected]:9000,1395796955800,Set(Metric(heap-memory-used,24796920,Some(EWMA(2.479692E7,0.15910417746033534))),
Metric(heap-memory-max,3817865216,None), Metric(processors,8,None),
Metric(system-load-average,2.2978515625,None),
Metric(heap-memory-committed,257425408,Some(EWMA(2.57425408E8,0.15910417746033534)))))))
[INFO] [03/26/2014 12:22:59.810]
[concenter-akka.actor.default-dispatcher-21]
[akka.tcp://[email protected]:9000/user/clusterListener] YYYYY:
ClusterMetricsChanged(Set(NodeMetrics(akka.tcp://[email protected]:9000,1395796979809,Set(Metric(heap-memory-used,26584712,Some(EWMA(2.6584712E7,0.15910417746033534))),
Metric(heap-memory-max,3817865216,None), Metric(processors,8,None),
Metric(system-load-average,2.55810546875,None),
Metric(heap-memory-committed,257425408,Some(EWMA(2.57425408E8,0.15910417746033534)))))))
[INFO] [03/26/2014 12:23:08.800]
[concenter-akka.actor.default-dispatcher-14]
[akka.tcp://[email protected]:9000/user/clusterListener] YYYYY:
ClusterMetricsChanged(Set(NodeMetrics(akka.tcp://[email protected]:9000,1395796988799,Set(Metric(heap-memory-used,26810728,Some(EWMA(2.6810728E7,0.15910417746033534))),
Metric(heap-memory-max,3817865216,None), Metric(processors,8,None),
Metric(system-load-average,2.619140625,None),
Metric(heap-memory-committed,257425408,Some(EWMA(2.57425408E8,0.15910417746033534)))))))
[INFO] [03/26/2014 12:23:11.800]
[concenter-akka.actor.default-dispatcher-4]
[akka.tcp://[email protected]:9000/user/clusterListener] YYYYY:
ClusterMetricsChanged(Set(NodeMetrics(akka.tcp://[email protected]:9000,1395796991800,Set(Metric(heap-memory-used,26810728,Some(EWMA(2.6810728E7,0.15910417746033534))),
Metric(heap-memory-max,3817865216,None), Metric(processors,8,None),
Metric(system-load-average,2.619140625,None),
Metric(heap-memory-committed,257425408,Some(EWMA(2.57425408E8,0.15910417746033534)))))))
*[DEBUG] [03/26/2014 12:23:12.818]
[concenter-akka.remote.default-remote-dispatcher-22] [Remoting] Associated
[akka.tcp://[email protected]:9000] <-
[akka.tcp://[email protected]:9090]*
[DEBUG] [03/26/2014 12:23:12.873]
[concenter-akka.remote.default-remote-dispatcher-27]
[akka.serialization.Serialization(akka://concenter)] Using
serializer[akka.cluster.protobuf.ClusterMessageSerializer] for message
[akka.cluster.InternalClusterAction$InitJoinNack]
[INFO] [03/26/2014 12:23:14.800]
[concenter-akka.actor.default-dispatcher-21]
[akka.tcp://[email protected]:9000/user/clusterListener] YYYYY:
ClusterMetricsChanged(Set(NodeMetrics(akka.tcp://[email protected]:9000,1395796994799,Set(Metric(heap-memory-used,35955000,Some(EWMA(3.5955E7,0.15910417746033534))),
Metric(heap-memory-max,3817865216,None), Metric(processors,8,None),
Metric(system-load-average,2.5693359375,None),
Metric(heap-memory-committed,257425408,Some(EWMA(2.57425408E8,0.15910417746033534)))))))
[INFO] [03/26/2014 12:23:41.800]
[concenter-akka.actor.default-dispatcher-4]
[akka.tcp://[email protected]:9000/user/clusterListener] YYYYY:
ClusterMetricsChanged(Set(NodeMetrics(akka.tcp://[email protected]:9000,1395797021799,Set(Metric(heap-memory-used,36148720,Some(EWMA(3.614872E7,0.15910417746033534))),
Metric(heap-memory-max,3817865216,None), Metric(processors,8,None),
Metric(system-load-average,2.2724609375,None),
Metric(heap-memory-committed,257425408,Some(EWMA(2.57425408E8,0.15910417746033534)))))))
*Now coming to my questions:*
(1) I can see in above that I "remoting" has been established. But does
that mean a Cluster has been established? This is my primary question. As
far I understand, akka clustering is a layer on top of remoting.
(2) To determine the answer to my question, I am trying to listen to
"cluster event". In my case, as you can see above, my listener is only
getting ClusterDomainEvent. While it proves that my listener is working but
I expected a MemberUP when Node 2 is booted.
(3) Whenever I bring down any of these nodes, Ctrl+c I get console output:
^C
Shutting down Akka...
Shutting down concenter.Platform
Goodbye Blue Sky, Goodbye!
Successfully shut down Akka
Here, I get my message for kernel shutdown, where I am doing actorsystem
shutdown. To my surprise no output from postStop of my listener actor?
Shouldn't I expect that?
Thank you for your time.
Regards,
Mohammad
--
>>>>>>>>>> Read the docs: http://akka.io/docs/
>>>>>>>>>> Check the FAQ:
>>>>>>>>>> http://doc.akka.io/docs/akka/current/additional/faq.html
>>>>>>>>>> Search the archives: https://groups.google.com/group/akka-user
---
You received this message because you are subscribed to the Google Groups "Akka
User List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/akka-user.
For more options, visit https://groups.google.com/d/optout.
