[akka-user][deprecated] Akka remoting via kafka

Hello, I wonder if this question has already been asked (I couldn't find a direct answer) - but would it be possible to use Kafka as a transport for akka remoting? I mean is akka open for implementing alternative transports for remoting and whether it makes sense in general thanks Vadim -- *

Re: [akka-user] Backpressure and bounded mailbox

t; > I don't fully understand what you mean by timeouts and blocking. > > Regards, > Patrik > > On Thu, May 21, 2015 at 8:32 PM, Vadim Bobrov > wrote: > >> Endre, >> >> I did. However this is a multiple-thousand lines app and a complete >> r

Re: [akka-user] Backpressure and bounded mailbox

pattern On Thursday, May 21, 2015 at 2:25:50 PM UTC-4, drewhk wrote: > > Hi Vadim, > > Why don't you look at Akka Streams first? It is all about handling > backpressure in a simple way. > > -Endre > > On Thu, May 21, 2015 at 7:39 PM, Vadim Bobrov > wrote: > >

[akka-user] Backpressure and bounded mailbox

Hello all! I am thinking of using bounded mailbox as a simple solution to backpressure implementation. My app basically consumes messages from a queue (kafka), unpacks and processes them and sends to HBase storage. The birds-eye overview of the principal actors is: W - writes to Hbase C - cons

Re: [akka-user] No CurrentState when preStart is overriden

On Friday, January 31, 2014 2:02:34 AM UTC-5, Martin Krasser wrote: > > > On 30.01.14 23:22, Vadim Bobrov wrote: > > That is what I meant, yes. Leaving messages on the JMS queue wouldn't work > for me because I have other actors (like retrieving info over REST) whose

Re: [akka-user] No CurrentState when preStart is overriden

; JmsManager. While you still end up with two actors you also create a > parent/child relationship which would allow you to still supervise only one > actor - JmsManager since its lifecycle would affect all its children. > > Oleg > > > On Thu, Jan 30, 2014 at 1:34

Re: [akka-user] No CurrentState when preStart is overriden

> Are you coupling too much responsibility into a single component? here is a birds-eye overview of what my system is. I am picking up messages from ActiveMQ, do some processing and save to HBase. All principal actors are FSM, I found it convenient to orchestrate their cooperative behaviour by

Re: [akka-user] No CurrentState when preStart is overriden

> It seems to me that the Processor is based on the assumption that it must preserve the sequence of Messages it received... Oleg, thanks. I'll give it a thought. However: if the Processor is based on the assumption that it basically cannot do any processing until after the recovery is done, th

Re: [akka-user] No CurrentState when preStart is overriden

Oleg, Martin, thanks I did not realize everything is stashed by Processor until after the recovery ends, regardless customized or not. My idea was to run a controlled recovery after my FSM actors are in a certain "ready" state (which involves coordinating several actors). But since a Processor

Re: [akka-user] No CurrentState when preStart is overriden

Thanks, Oleg, If you make FsmActor in your example inherit from Processor (from akka-persistence) you'll see the effect (class FsmActor extends Processor with FSM[State, Data] {) Vadim -- >> Read the docs: http://akka.io/docs/ >> Check the FAQ: http://akka.io/faq/ >>

[akka-user] No CurrentState when preStart is overriden

Hi, I ran into this problem: when preStart is overriden (empty) for an FSM actor it stops sending CurrentState in response to SubscribeTransitionCallBack. Tried putting startWith initialize into preStart, still not getting CurrentState. What is the correct solution in this case? I am disabli

[akka-user] Re: leveldb journal growth

et to figure out what version it is On Wednesday, January 8, 2014 6:53:03 PM UTC-5, Andy R wrote: > > On Wednesday, January 8, 2014 1:50:30 PM UTC-5, Vadim Bobrov wrote: >> >> Also - is there any way to look into what's in the journal? >> > There is leveldbutil: >

Re: [akka-user] Persistent messages out-of-order when sent via channel

> There are numerous ways (impl options) for resending messages. Using a > channel is only one. > Other than persistent channel what are the options? I actually believe it is important to note in the documentation that messages in mailbox are not preserved by a processor, it was not obviou

Re: [akka-user] leveldb journal growth

how do you run the LevelDB journal, with native = on or off? Please note > that native = off (LevelDB Java port) is only for testing purposes and > doesn't do data compression. With the native LevelDB journal I end up > with 40M disk space after 1M messages and 200M and 5M messages (whereas >

Re: [akka-user] Persistent messages out-of-order when sent via channel

good catch, that's a bug in the actor ref resolution in channels. Ticket created . > > Thanks for reporting. > > Anytime :) Here is one more fact that I learned from this example - message

Re: [akka-user] leveldb journal growth

Martin, I understand. But here is what I am trying to achieve - I have an app that processes potentially large amounts of incoming data (hundreds of thousands of messages per sec). I want to allow the app to crash anytime and not take any data with it (write-ahead log). Given the amount of data

[akka-user] Persistent messages out-of-order when sent via channel

class A extends Processor { val b = context.actorFor("/user/top/b") val channel = context.actorOf(Channel.props(ChannelSettings(redeliverInterval = 180 seconds, redeliverMax = 10)), name = "bc") override def receive = { case Start => for(i <- 1 to 10) { channel ! Deliver(Persistent(i), b, Reso

[akka-user] Persistent messages out-of-order when sent via channel

Hi, Here are 2 processors communicating via a channel: class A extends Processor { val b = context.actorFor("/user/top/b") val channel = context.actorOf(Channel.props(ChannelSettings(redeliverInterval = 180 seconds, redeliverMax = 10)), name = "bc") override def receive = { case Start => for

Re: [akka-user] leveldb journal growth

age in any combination - database stays at a constant size On Thursday, January 9, 2014 9:40:02 AM UTC-5, Vadim Bobrov wrote: > > UPDATE: actually running deleteMessage (single) instead of deleteMessages > did fix it. Could be a bug in deleteMessages? I am using 1.3-M2 > > On Thursday

Re: [akka-user] leveldb journal growth

UPDATE: actually running deleteMessage (single) instead of deleteMessages did fix it. Could be a bug in deleteMessages? I am using 1.3-M2 On Thursday, January 9, 2014 9:23:31 AM UTC-5, Vadim Bobrov wrote: > > Oops, I mean to add permanent = true but forgot, apologies. It doesn't

Re: [akka-user] leveldb journal growth

January 9, 2014 1:02:27 AM UTC-5, Martin Krasser wrote: > > Hi Vadim, > > On 08.01.14 19:50, Vadim Bobrov wrote: > > Hi, > > I have this leveldb journal problem when playing with the new > persistence module. With this simple setup: > > class ProcAct

[akka-user] leveldb journal growth

Hi, I have this leveldb journal problem when playing with the new persistence module. With this simple setup: class ProcActor extends Processor { override def receive = { case Persistent(_, sequenceNr) => deleteMessages(sequenceNr) } } object Main extends App { implicit lazy val system =

[akka-user] Re: resolve of path sequence failed

Hi Martin, I am not using channels, just processors and persistent messages. Here is the code object Main extends App { val config = ConfigFactory.load().getConfig("evsourd") implicit lazy val system = ActorSystem("evsourd", config) val top = system.actorOf(Props[TopActor], name = "top") for

[akka-user] resolve of path sequence failed

Hello, I am trying out the new persistence module in 2.3-M2 and it works like a charm except for about every recovered message I am getting this in the log: LocalActorRefProvider(akka://evsourd) - resolve of path sequence [/user/top#-228886491] failed My understanding is that the recovery star