Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • Jan 25 14:59
    Aaronontheweb commented #3615
  • Jan 25 14:58
    Aaronontheweb commented #3615
  • Jan 25 14:56
    Aaronontheweb commented #4126
  • Jan 25 13:16
    IgorFedchenko synchronize #4126
  • Jan 25 12:37
    IgorFedchenko commented #3615
  • Jan 25 12:37
    IgorFedchenko commented #3615
  • Jan 25 12:35
    IgorFedchenko commented #3615
  • Jan 25 12:16
    IgorFedchenko synchronize #4126
  • Jan 25 12:03
    IgorFedchenko opened #4165
  • Jan 25 11:06
    IgorFedchenko synchronize #4126
  • Jan 25 10:45
    IgorFedchenko commented #4126
  • Jan 25 10:37
    IgorFedchenko synchronize #4126
  • Jan 25 10:36
    IgorFedchenko edited #4126
  • Jan 25 10:34
    IgorFedchenko synchronize #4126
  • Jan 25 10:30
    IgorFedchenko synchronize #4126
  • Jan 24 23:00
    Danthar commented #3954
  • Jan 24 22:53
    Aaronontheweb commented #3954
  • Jan 24 21:06
    Aaronontheweb synchronize #4128
  • Jan 24 16:56
    Aaronontheweb synchronize #4128
  • Jan 24 16:46

    Aaronontheweb on explain-sharding-path

    (compare)

Onur Gumus
@OnurGumus
Use split brain then.
Vasily Kirichenko
@vasily-kirichenko
how?
resolver you mean?
Onur Gumus
@OnurGumus
It is very simple and documented
Yes
Vasily Kirichenko
@vasily-kirichenko
ah. will read about it.
Onur Gumus
@OnurGumus
to tackle with network partitioning
Vasily Kirichenko
@vasily-kirichenko
However, it's not very important in my case. The actor is just a batch worker, nobody even sends messages to it.
so a minute downtime is ok
Onur Gumus
@OnurGumus
no it is not okay
you have to down it by some means
if you don't down it then I don't know maybe then it is okay
Bartosz Sypytkowski
@Horusiath
@vasily-kirichenko kill switch invoked on PostStop seems fine.
@OnurGumus if you're able to tell cluster to leave gracefully i.e. via Cluster.LeaveAsync, it's always better - @vasily-kirichenko case regards non-graceful scenario, when the cluster must detect dead node instead of being explicitly informed about it leaving
Onur Gumus
@OnurGumus
I think it is not possible to use both loggers : ["Akka.TestKit.TestEventListener, Akka.TestKit","Akka.Logger.Serilog.SerilogLogger, Akka.Logger.Serilog"]
As first one treats curly braces as string format and the other one doesn't
Bartosz Sypytkowski
@Horusiath
I would really like to create some reference example on how to design .net apps with akka.net - the only problem is that I need some well defined domain, where such use makes sense. I don't want to invent something by myself from thin air
I can think about few domains, but they'd be using only small parts of akka. Not a comprehensive example.
Ondrej Pialek
@ondrejpialek
If you want to help me design a sharded ecommerce Akka-powered app let me know :)
Bartosz Sypytkowski
@Horusiath
@ondrejpialek the context, I was talking about need something that won't be NDA'ed ;) Something to show publicly.
but if you've stuck somewhere in your design, you can always share your doubts here
Ondrej Pialek
@ondrejpialek
I know ;) I won't have any problem with making some "framework" or "patterns / practices" to be opensourced and the domain changed slightly and released as an example too. Something along the https://github.com/pawelkaczor/akka-ddd say. We are not that possessive where I work at :)
One of the main problems I am having adopting Akka is lack of real world use cases - so definitely support your idea ;)
Most of the books I've read or videos I've seen were either too theoretical or too tactical.
Hyungho Ko
@hhko
@Aaronontheweb @vasily-kirichenko thank you for your consideration about Spark.
https://www.nuget.org/packages/Akka.Quartz.Actor/ depends on .NET Framework 4.5.2.
do you have any plans to support Akka.Quartz.Actor based on .NET Framework 4.5?
Vasily Kirichenko
@vasily-kirichenko
@hhko there were issues on production with Akka.Quartz.Actor
Hyungho Ko
@hhko
I think there's no reason to depend on .NET Framework 4.5.2.
Onur Gumus
@OnurGumus
I am seeing a weird behavior .Each time I create an actor My serialization settings are reset. Actually Serializer Object is being Set again
Does anyone know why?
Onur Gumus
@OnurGumus
I hate Akka's internal methods and interfaces
Onur Gumus
@OnurGumus
OMG! loading event journal causes all serializers to be reloaded in actor system
I mean the read journal
Onur Gumus
@OnurGumus
When ever a persistence actor is created
it forces the serialization settings to be reloaded
this doesn't sound right
Ondrej Pialek
@ondrejpialek
@OnurGumus probably best to submit an isssue to the GitHub repo page with steps to reproduce?
Bartosz Sypytkowski
@Horusiath
@OnurGumus this sounds like a serious issue
Onur Gumus
@OnurGumus
@Horusiath I think this is by design.
There is an internal interface : ISupportSerializationConfigReload
Bartosz Sypytkowski
@Horusiath
"broken by design" ;)
Onur Gumus
@OnurGumus
When ever you create a new persistence actor instance (not sure if always perhaps once per type), that ctor triggers this.
this in turn calls the private method of ActorSystemImpl
private void ConfigureSerialization()
{
    this._serialization = new Serialization(this);
}
so if somehow you customized your serializer programmatically , you lose the changes.
this is not an oversight and done deliberately but not sure if I understand the motivation
Bartosz Sypytkowski
@Horusiath
if I remember correctly, the problem was that in the past a Persistence plugin might have not been created (+ actor system not configured to use persistence) when you were creating a persistent actor for the first time
nonetheless config reload on every new actor sounds like serious issue
Onur Gumus
@OnurGumus
it has to do with persistence extensions. I have collected the full stack trace but not with me now
zbynek001
@zbynek001
@OnurGumus I've added that ISupportSerializationConfigReload interface. It's called only inside InjectTopLevelFallback, so is something maybe calling that too often?