1. Is there any way a node which is part of Cluster can inform other members to mark its status as Down.
Cluster().down(self) and then not terminating the actor system directly so that the information has time to reach the other nodes: Cluster cluster = Cluster.get(system); cluster.down(cluster.selfAddress()); // or more gracefully cluster.leave(cluster.selfAddress()); 2. We have multiple Cluster Singleton instances in the cluster, and when the node which is holding the instance goes unreachable, all the requests to the Singleton instance going to Deadletter box or getting added to Proxy's queue as configured That's how singletons work. Be quicker with manual downing or tweak the SBR (Split Brain Resolver http://doc.akka.io/docs/akka/akka-commercial-addons-1.0.0/java/split-brain-resolver.html) to down nodes quicker. Or avoid singletons if you want higher availability. -- Johan Akka Team Lightbend <http://www.lightbend.com/> - Reactive apps on the JVM Twitter: @akkateam On Sat, Nov 12, 2016 at 7:22 PM, Shyam kumar Akirala < [email protected]> wrote: > 1. Is there any way a node which is part of Cluster can inform other > members to mark its status as Down. We'are using Akka with Java, and we've > registered a shutdown hook to gracefully shutdown Actor System in case of > JVM going down. But seems after terminating Actor System on a node, that > particular node's status is marked as Unreachable instead of Down. > > Runtime.getRuntime().addShutdownHook( > > new Thread() { > @Override > public void run() { > if (actorSystem != null) { > Future<Terminated> terminateFut = system.terminate(); > try { > Await.result(terminateFut, Duration.Inf()); > } catch (Exception e) { > LOGGER.error("Error shutting down Actor system", e); > } > } > } > } > ); > > > 2. We have multiple Cluster Singleton instances in the cluster, and when > the node which is holding the instance goes unreachable, all the requests > to the Singleton instance going to Deadletter box or getting added to > Proxy's queue as configured. In both cases there is message loss (in case > of queue, if the limit 10000 crosses), and downtime involved. Can you share > any best practises to reduce the Cluster Singleton Instance downtime in > case of node crash or network partition. > > Thanks, > Shyam > > > > > > -- > >>>>>>>>>> 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 https://groups.google.com/group/akka-user. > For more options, visit https://groups.google.com/d/optout. > -- >>>>>>>>>> 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 https://groups.google.com/group/akka-user. For more options, visit https://groups.google.com/d/optout.
