[akka-user] Akka Persistent Actor stops processing messages

First a thank you to people answering questions. Against first the environment Akka 2.4.11 Scala 2.11.8 Cassandra 3.9 I've been chasing an issue for days where one of my persistent actors appears to stop processing message to the point where the JVM memory grows to 8GB before it gets an out of

Re: [akka-user] Akka persistent

Interesting (and strange), that should be fixed. Please create an issue . /Patrik On Mon, Jan 9, 2017 at 6:16 AM, Dai Yinhua wrote: > Hi Patrik, > > From Akka documentation, "Note that *Cluster Sharding* >>

Re: [akka-user] Akka persistent

Hi Patrik, >From Akka documentation, "Note that *Cluster Sharding* > > is > using snapshots, so if you use Cluster Sharding you need to define a > snapshot store plugin." > I haven't tested. Isn't it possibl

Re: [akka-user] Akka persistent

OK, it make sense to me now, thank you. On Friday, 6 January 2017 19:38:10 UTC+8, Patrik Nordwall wrote: > > > > On Fri, Jan 6, 2017 at 2:57 AM, Dai Yinhua > wrote: > >> Hi Patrik, >> >> From Akka documentation, "Note that *Cluster Sharding* >>

Re: [akka-user] Akka persistent

On Fri, Jan 6, 2017 at 2:57 AM, Dai Yinhua wrote: > Hi Patrik, > > From Akka documentation, "Note that *Cluster Sharding* > > is > using snapshots, so if you use Cluster Sharding you need to define a > snapsho

Re: [akka-user] Akka persistent

Hi Patrik, >From Akka documentation, "Note that *Cluster Sharding* is using snapshots, so if you use Cluster Sharding you need to define a snapshot store plugin." So I am wondering if akka.persistence.snapsh

Re: [akka-user] Akka persistent

For learning with Cassandra, I also strongly recommend picking up ccm -- it makes it *vastly* easier to run a local Cassandra pseudo-cluster on your machine, and works quite well for akka-persistence-cassandra development and testing. On Thu, Jan 5, 2017 at 9:09 AM

Re: [akka-user] Akka persistent

snapshots should only be treated as an optimization, so it should not be necessary to load snapshots. shared-leveldb-journal is only intended for testing, but also for that purpose I would recommend using something real such as akka-persistence-cassandra. /Patrik On Thu, Jan 5, 2017 at 9:34 AM,

[akka-user] Akka persistent

Hi Guys, I'm learning Akka cluster sharding with shared leveldb plugin. Here is my config: persistence { journal.plugin = "akka.persistence.journal.leveldb-shared" journal.leveldb-shared.store { # DO NOT USE 'native = off' IN PRODUCTION !!! native = off

[akka-user] Akka Persistent Actor Lifecycle

Hi, I'm new to AKKA world and I'm working on implementing EventSourcing using the PersistentActors, but I got confused about the life-cycle of those actors. As far as I understand, the default behavior of persistent actors is to stay in the memory until being shutdown, which may overwhelm the

Re: [akka-user] [akka-persistent]: Refine the logic regarding when channel messages should be re-delivered

Hi, I still don't think that it's the job of the Transient channel to do this throttling for you. Have you experimented with your proposed change? Does it really help? What stops you from running out of memory by just growing the number of messages that are in flight indefinitely? B/ On 15 Ma

Re: [akka-user] [akka-persistent]: Refine the logic regarding when channel messages should be re-delivered

Sorry for that. It's just redeliveries, a lot of redeliveries, we didn't really encounter redelivery failures. Thank you for the suggestions regarding throttling. We surely haven't done that yet. I just want to share with you and the akka community that when our load is high, we can achieve a

Re: [akka-user] [akka-persistent]: Refine the logic regarding when channel messages should be re-delivered

Hi Daniel, You still haven't explained if there are redelivery failures or just redeliveries, and if the destination is local or remote. If you see redeliveries or failures while the the destination is alive and healthy, then you are flooding your destination. That is a flow control issue that

Re: [akka-user] [akka-persistent]: Refine the logic regarding when channel messages should be re-delivered

On May 14, 2014, at 2:32 PM, Björn Antonsson wrote: > Hi Daniel, > > I'm not sure where your logic would be used. Are you using transient Channels > and a Processor or are you using PersistentChannels? If you are using a > PersistentChannel then you could limit the maximum number of outstand

Re: [akka-user] [akka-persistent]: Refine the logic regarding when channel messages should be re-delivered

Hi Daniel, I'm not sure where your logic would be used. Are you using transient Channels and a Processor or are you using PersistentChannels? If you are using a PersistentChannel then you could limit the maximum number of outstanding messages via the pendingConfirmationsMax setting. A redelive

Re: [akka-user] [akka-persistent]: Refine the logic regarding when channel messages should be re-delivered

I se the interval to 10 seconds. Does my suggestion make any sense to you? > 在 2014年5月13日，下午7:11，Björn Antonsson 写道： > > Hi Daniel, > > Are you talking about getting RedeliveryFailure messages when your load is > high? The Channel should handle redelivery for you. If you have a lot of > red

Re: [akka-user] [akka-persistent]: Refine the logic regarding when channel messages should be re-delivered

Hi Daniel, Are you talking about getting RedeliveryFailure messages when your load is high? The Channel should handle redelivery for you. If you  have a lot of redeliveries happening, then maybe your redeliver interval is set too tight, or you are simply overloading your system. B/ On 12 May

[akka-user] [akka-persistent]: Refine the logic regarding when channel messages should be re-delivered

I'm seeing in my application a lot of channels message redelivery when the overall load of the system is high. Is it a good idea to re-deliver channel message this way: 1. record the latest confirmed time [ of any message] 2. re-deliver the message only iff (current_timestamp - latest_

[akka-user] [akka-persistent]: Refind the way channel messages are confined

I'm seeing in my app a lot of message redelivery by channels when the overall load of the system is high. Is it a good idea to re-deliver channel message this way: 1. record the latest confirmed time 2. resend the message : iff (message_delivery_timestamp - latest_confirmed_timestamp)

Re: [akka-user] Akka persistent messages sent to two destinations thru one channel?

Does the following answer your question? If a processor emits more than one outbound message per inbound Persistent message it *must* use a separate channel for each outbound message to ensure that confirmations are uniquely identifiable, otherwise, at-least-once message delivery semantics do not

[akka-user] Akka persistent messages sent to two destinations thru one channel?

Can I use 'withPayload' to send the same persistent message to two destinations using one channel? as below: case p @ Persistent(payload, seq) => val p2 = p.withPayload(newPayload) channel forward Deliver(p2, path1) channel forward Deliver(p2, path2) If not, what