As long as commands lead to simple operations like adding a value to a
list, there is no different between what we do in receiveRecover
and receiveCommand so we can merge them together. Regarding DRY, you're
right but we can create a map of commands to events and use it in
receiveCommand to be
Hi Brandon,
On Tue, 2015-06-30 at 08:03 -0700, Brandon Arp wrote:
Hi Brice,
I had a similar need for operational visibility. The approach I took
was to create a ShardAllocationStrategy and configure my cluster to
use it. The strategy would save off the allocations so that they
could
I tried something like this, where I had a PushPullStage that accepted a
Source into its constructor. It seemed like the wrong approach, because it
seemed like I was just reimplementing a common merge functionality.
On Wednesday, July 1, 2015 at 3:58:49 AM UTC+3, Chad Retz wrote:
Although I
Hi,
my applicataion consists of two front end akka processes which forward
messages to 5 backend nodes.
The frontend actors use ConstantHashingRouting to devide work to the
backened nodes.
If any backend nodes become unreachable I would like the backends not
only to be auto-removed (I know
I am currently only running the actor that listens on a single node (local
to the shard coordinator), but it would be easy to run it on multiple or
all nodes. The allocation strategy had easy access to a list of all of the
cluster nodes.
On Wed, Jul 1, 2015, 07:04 Brice Figureau
Hi Patrik,
Unfortunately I cannot provide you the logs since I don't have them
anymore, however, I can provide you all my akka cluster configs (I am using
different parameters than defaults). Let me also provide more details: we
are trying to form the cluster while the nodes are running with
Hello,
I am currently in the process of evaluating Akka. I have successfully built
a proof of concept REST API using akka-http which is part of an Akka
cluster system that has a number of Actors in it.
Everything works well and I'm happy with what I am seeing until I try to
ramp up the load
also, add param -p 2551:2551 worked for me as well
On Monday, October 20, 2014 at 3:31:17 PM UTC-4, tsvika wrote:
Thanks Rafal!
--net=host worked just fine. The links you provided are great and
informative :)
On Wednesday, October 15, 2014 12:41:03 AM UTC+3, Rafał Krzewski wrote:
The
Hi all,
I just wanted to share a project I've been working on over the last 6
months. I've called it CurioDB [1], it's a clone of Redis [2] that I've
built in under 1000 lines of Scala/Akka. It makes use of Akka's persistence
and clustering to overcome some of the shortcomings of Redis [3].
Seems like the only place the term principal is located in the code is in
the akka.stream.io.SessionBytes. I am guessing this isn't possible with RC4?
https://github.com/akka/akka/blob/releasing-akka-stream-and-http-experimental-1.0-RC4/akka-stream/src/main/scala/akka/stream/io/SslTls.scala
Hi Srinivas,
You can consider using Persistent FSM
http://doc.akka.io/docs/akka/2.4-M2/scala/persistence.html?_ga=1.127649677.1441866889.1394995817#Persistent_FSM,
this way the state of the FSM can be recovered upon restart.
This will let you take down the inactive actors - they can later be
Thanks for the clarification and the tip.
On Tue, Jun 30, 2015 at 1:57 PM Richard Bradley
richard.bradley.softw...@gmail.com wrote:
how does the HTTP protocol signal it to the TCP Layer.
It doesn't; it delegates the flow control and back-pressure entirely to
the TCP layer.
When you use
12 matches
Mail list logo