Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • 05:22
    nagytech commented #4054
  • Nov 20 18:58
    Aaronontheweb commented #3863
  • Nov 20 18:16
    Aaronontheweb commented #4022
  • Nov 20 17:52
    MaximG1234 commented #4054
  • Nov 20 17:52
    MaximG1234 commented #4054
  • Nov 20 17:52
    MaximG1234 commented #4054
  • Nov 20 17:51
    MaximG1234 commented #4054
  • Nov 20 17:49
    Aaronontheweb commented #4054
  • Nov 20 17:49
    MaximG1234 commented #4054
  • Nov 20 17:48
    MaximG1234 commented #4054
  • Nov 20 17:47
    MaximG1234 commented #4054
  • Nov 20 17:44
    Aaronontheweb commented #4054
  • Nov 20 17:43
    MaximG1234 commented #4054
  • Nov 20 17:42
    ismaelhamed synchronize #3863
  • Nov 20 17:39
    MaximG1234 commented #4054
  • Nov 20 17:35
    MaximG1234 commented #4054
  • Nov 20 17:34
    MaximG1234 commented #4054
  • Nov 20 17:31
    MaximG1234 commented #4054
  • Nov 20 17:10
    Aaronontheweb commented #4054
  • Nov 20 17:06
    MaximG1234 commented #4054
Arsene T. Gandote
@Tochemey
Hello can someone shows me a complete sample of the circuit breaker implementation?
nathvi
@nathvi
humm
I'm confused on the example provided in the documentation. In the OnReceive method override, at the very end there is a statement "Self.Tell("continue");" I understand that this sends a message to the current actor, but I don't see any code that checks to see if the message object is "continue". Is this code handled by akka.net?
Jeff Pang
@jeff-pang
Hi everyone, may I know if any of you have any experience or resources on Akka Java on Android?
Can't seem to find much info on this topic. is Akka (Java) even compatible with Android?
AndreSteenbergen
@AndreSteenbergen
No experience for me, but if you are using remote, I would use some extra middle layer communicating with your clients, like RabbitMq
Using remote exposes your entire network.
Arjen Smits
@Danthar
@boekabart added some remarks. I suppose you need to add some more examples to illustrate the need for this change ;)
Arsene T. Gandote
@Tochemey
@jeff-pang All I can tell you is that I have used it with Xamarin and it works smoothly. Instead of Http I use akka remote to communicate between the mobile app and the backend service. So the mobile app had a mini akka system.
There is no harm in trying
Arsene T. Gandote
@Tochemey
Hello geeks how different is the Polly circuit breaker from the Akka.NEt one because I am not seeing any where how to handle the exception.
Bart de Boer
@boekabart

@Danthar

@boekabart added some remarks. I suppose you need to add some more examples to illustrate the need for this change ;)

I've elaborated a bit in the ticket, and will try to add actual code examples. But since the problem is mainly caused by the need to be able to 'compose' actors in a generic way, it's not so trivial to explain in a simple example.

Arjen Smits
@Danthar
Ok, So if i understand correctly. Sometimes you want the original sender, so you use (s) => new msg(s) as the func
and sometimes you want to use a different sender, so you use (s) => new msg(d) as the func
and the reason you want this message factory func, because it would enable you to NOT use all kinds of nasty fallback code in either the sending or receiving code
/cc @boekabart
Greatsamps
@Greatsamps
Hey Guys, i have a question about performing an orderly shutdown. I am aware that it is benefitial for an actor node to be taken offline gracefully so that the cluster is aware that it is gone for a specific reason and has not just simply failed. When running in a micro-services situation where nodes will often be running via docker, is there any way that a graceful shutdown can be performed, or do you just have to deal with it being dirty as docker stops the container?
Bart de Boer
@boekabart
@Danthar well in fact the only reason we need the Func<IActorRef,object> is the use case actor.Ask<MyResponse>( (ref) => new SendAnswerTo(ref));- your first example. This code resides either in a Unit Test (although usually we are OK with using TestActor as recipient), or in production code in an Api Controller ("HTTP to Akka converter"). The second case resides in other actors, not using ask but just .Tell( new SendAnswerTo( thisActorIVaguelyKnow )) .
Kamil Wojciechowski
@aph5
Good evening, i have one question about the actor communication in a cluster. lets say that we have 2 actors, each configured to be in a different cluster role and deployed to 2 different machines. If actor A wants to send a message to actor B, then could it be done as a normal Tell? (those actors are in the same system)
or should I use Cluster Client OR distributed pub/sub ?
Aaron Stannard
@Aaronontheweb
yep, can be done as a normal tell
ClusterClient is really for instances where something "outside" of the cluster needs to be able to connect into it
DistributedPubSub is for doing topic brokering inside the cluster
Kamil Wojciechowski
@aph5
ok, thanks !
Arsene T. Gandote
@Tochemey
Hello I would like to know whether there is a very simple tutorial on the Akka Cluster. I have gone through the WebCrawler and honestly speaking I get confused even though I have been using Akka net. If I can get a simple tutorial of three nodes on cluster I believe it will help.
Aaron Stannard
@Aaronontheweb
have you checked out docs on the website and some of our youtube videos?
Arsene T. Gandote
@Tochemey
@Aaronontheweb I have gone through the doc site and I really love it. However how to start is my issue. :worried:
I'd recommend playing around with that
Arsene T. Gandote
@Tochemey
Ok
Aaron Stannard
@Aaronontheweb
starts a basic 3 node cluster all inside a single process
Arsene T. Gandote
@Tochemey
Thanks
Aaron Stannard
@Aaronontheweb
shows you how to use the Cluster plugin too
to subscribe to events
et al
nbparrell
@nbparrell
Hey guys, I am new to Akka.net and am hoping to contribute to the project as part of a school project. I am looking forward to learning more about the project.
Bartosz Sypytkowski
@Horusiath
Hi @nbparrell . Contributions to project are most welcome! :D If you'll need some help, just call me, @Aaronontheweb , @Danthar or @marcpiechura (only note: I'm on vacations untill 8th of April, so I might be hard to catch ;) ). If you don't know where to start:
  1. There's a free bootcamp made by @Aaronontheweb, it's a by-example introduction to akka.net . There are also several different books and courses over the web, but I don't know much about their quality.
  2. There's also a Petabridge blog and youtube channel.
  3. You can also take a look at my blog with akka.net tag to find some articles.
  4. And of course there's an official documentation.
ziya
@mtmk
A quick question that’s been in my mind about cluster / remoting maintenance: how do you suggest to deploy new releases without down time? Also about Hyperion serialisation is there backward / forward compatibility concepts? I might’ve missed a good article or documentation might be covering that. Feel free to rtfm me for this :fire:
Lutando Ngqakaza
@Lutando

I have a small question. How can I best test my persistent actors recovery from persistence.

my first test case is to do it "black box style" i send x commands resulting in some good internal state changes. I kill the actor and then revive the actor and send a y command and observe the event coming from that actor using a probe actor who subscribes to that event being published

problem is when i kill the actor i just end up getting dead letters
Lutando Ngqakaza
@Lutando
another question wrt my assumptions, if an actor is recovering does that mean it wont process any other non-recovery messages?
Lutando Ngqakaza
@Lutando

docs read: By default, a persistent actor is automatically recovered on start and on restart by replaying journaled messages.

i interpret this as my assumptions as being right?

nbparrell
@nbparrell
Thank you @Horusiath for the warm welcome and the information in regards with where to start. I have started will reading through the documentation but I am glad to see that there is a youtube channel out there. That will definitely make it easier as I sink my teeth in to akka.net. :)
Bartosz Sypytkowski
@Horusiath
@mtmk my idea was to add version tag as one of the roles for a deployed node. Regarding Hyperion, it's using index-based fields, however unlike protobuf, the index is assigned implicitly, not explicitly. Therefore, you can add new fields, change their names, sometimes change their type, but you cannot reorder them or remove them.
@Lutando if actor is killed, and then resurrected, it will first try to replay state from snapshots and events - if any other message came in the meantime, it will be stashed aside, until recovery completes.
Lutando Ngqakaza
@Lutando
Thanks @Horusiath
Ismael Hamed
@ismaelhamed

if any other message came in the meantime, it will be stashed aside, until recovery completes.

@Horusiath current implementation is not complete akkadotnet/akka.net#3366

Bartosz Sypytkowski
@Horusiath
@ismaelhamed I wouldn't call it incomplete, it's more like it doesn't snap you in the face when you try to do something you shouldn't at the moment.
Lutando Ngqakaza
@Lutando
@ismaelhamed I dont understand that PR, you want the ability to Persist right after Recovery is complete? but not as a result of any incoming commands?
Ismael Hamed
@ismaelhamed
@Lutando see akka/akka#21736 and akka/akka#22609
Lutando Ngqakaza
@Lutando
I see