Hi all. I have a new set of questions this time relating to clustering.
1. How many/what percentage of the nodes in a cluster should ideally be
made seed nodes? Would there be a performance impact from making all of the
nodes in the cluster seeds?
2. Are there any implications in having too many singletons running on a
cluster? Is there a point when using cluster sharding would be better than
having a lot of singletons?
3. When using cluster-aware group routers, if my router configuration
looked something like this:
akka.actor.deployment {
/statsService/workerRouter {
router = round-robin-pool
nr-of-instances = 100
routees.paths = ["/user/worker"]
cluster {
enabled = on
allow-local-routees = on
}
}
}
then any messages that I send to said router have the potential to get sent
to any actor with that name ("worker") running in that same level ("/user")
that might be running anywhere in the cluster, right? In other words, does
this router's routees include all actors (at the user level) running in any
node of the cluster with the name "worker"?
4. Is there a point to using the "nr-of-instances" field in a configuration
for a cluster aware group router? I read that the field gets ignored if
also include "routees.paths" so I'm wondering what its purpose is in a
group router. The reason I ask is because the Typesafe activator project
for clustering used it in one of it's examples.
5. Is it possible to broadcast a message to several actors that reside in
different nodes of a cluster using ActorSelection?
6. Between using ActorSelection (assuming the answer to 3 was yes), using
cluster-aware broadcast routers, and using distributed pub-sub, which
method would be best for splitting a data stream so that a copy of it goes
to different parts of a cluster?
7. Say that I have a singleton master actor that creates a cluster-aware
pool router (which means that the router is its child). If the node
containing the master actor goes down (say, a hardware failure), will the
router (and all of it's children/routees) go down with it or can they
survive somehow until a new master actor is created by the fact that it was
declared to be singleton?
8. If I have a cluster aware pool router in a cluster with say 10 nodes and
I define:
nr-of-instances = 100
will those 100 routess be distributed evenly to each node? In this case,
will each node be running 10 routees? Also, if one of the nodes were to go
down, would the remaining 9 nodes create more routees in order to
compensate for the 10 that were lost?
--
>>>>>>>>>> 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 akka-user+unsubscr...@googlegroups.com.
To post to this group, send email to akka-user@googlegroups.com.
Visit this group at http://groups.google.com/group/akka-user.
For more options, visit https://groups.google.com/d/optout.
