Thanks Konrad.
~
Karthik
On Wed, Dec 2, 2015 at 1:56 AM, Konrad Malawski
wrote:
> No, the sharded actors do not need to be persistent.
> Akka Persistence *is* however needed internally by Cluster Sharding in order
> to be able to cope with recovery / dying of nodes
No, the sharded actors do not need to be persistent.
Akka Persistence *is* however needed internally by Cluster Sharding in order to
be able to cope with recovery / dying of nodes etc.
So:
- you need to configure persistence with a distributed journal, yes
- no, your sharded actors do not have
Konrad,
Does cluster sharding feature require the actors to be persistent
actors? I looked at the typesafe examples for cluster sharding -
http://www.typesafe.com/activator/template/akka-cluster-sharding-scala and
they are all based on Persistent actors. I'm trying to use cluster sharding
We have a system where we require all requests pertaining to a given
user-id be processed in order and by the same actor. For example : if we
get requests for userid - u1,u2,u3,u1,u4,u5,u5 in that order then the
requests for user-id u1 must always go to the same actor so that they can
You should read up on cluster sharding:
http://doc.akka.io/docs/akka/current/scala/cluster-sharding.html
Such "to the same Actor, given an identifier" is exactly what it is designed
for.
--
Cheers,
Konrad 'ktoso’ Malawski
Akka @ Typesafe
On 28 November 2015 at 11:59:52, Karthik Deivasigamani
Thanks Konard
> On Nov 28, 2015, at 3:47 AM, Konrad Malawski
> wrote:
>
> You should read up on cluster sharding:
> http://doc.akka.io/docs/akka/current/scala/cluster-sharding.html
>
> Such "to the same Actor, given an identifier" is exactly what it is designed