Re: [akka-user] Remote deploy of set of actors in cluster with consistent-hashing-pool

2014-05-17 Thread Eugene Dzhurinsky
Heiko, I want to clarify one point: If I have Actor A, which spawns set of actor B in a cluster, and every actor B spawns it's own set of Actor C, then: - node with Actor C to be deployed should have no routers configured - on a node with Actor B to be deployed, the router for Actor C should

Re: [akka-user] Remote deploy of set of actors in cluster with consistent-hashing-pool

2014-05-15 Thread Eugene Dzhurinsky
On Thu, May 15, 2014 at 06:11:21PM +0200, Heiko Seeberger wrote: Yep, start the router on the chunk node. Pay attention to the path in the configuration file, it must match the semi-absolute (without /user) path of your router actor. Then also remove the unnecessary and confusing

Re: [akka-user] Remote deploy of set of actors in cluster with consistent-hashing-pool

2014-05-15 Thread Heiko Seeberger
On Thu, May 15, 2014 at 6:17 PM, Eugene Dzhurinsky jdeve...@gmail.comwrote: On Thu, May 15, 2014 at 06:11:21PM +0200, Heiko Seeberger wrote: Yep, start the router on the chunk node. Pay attention to the path in the configuration file, it must match the semi-absolute (without /user) path of

[akka-user] Remote deploy of set of actors in cluster with consistent-hashing-pool

2014-05-14 Thread Eugene Dzhurinsky
Hello! I'm trying to start 2 nodes in a cluster - one with role http and another one with role chunk role. Initially there are no actors running at the http node (ones should be deployed when node chunk joins the cluster) my application.conf: akka { actor { provider =