Sorry, something went wrong with my computer and it posted the query half
way.
Here is my question again:-)
I have another question regarding Agent. Based on my scenario explained in
my post - there is a situation where i need to share some mutable state
between DIFFERENT actor.
For example:
The DatabaseWorker is happily processing data in a while loop for Request 1.
Assume it has split the query into 4 times slices ….
while (moreTimeSlicesToProcess)
{
results = sendQueryToDataBase()
resultProcessor.tell(results)
}
In the mean time a CANCEL message comes in for request 1. I want the
DatabaseWorker to stop processing any more time slices.
SO
(1) When RequestProcessor gets a CANCEL message, it sets an AGENT<Boolean>
field to true/false
(2) The DatabaseWorker should look for this Agent<Boolean> value in the
while loop and stop future processing.
(3) Something like:
while (agent<boolean>.get() && moreTimeSlicesToProcess)
{
result = sendQueryToDatabase()
resultProcessor.tell(result)
}
Is this even feasible. I looked at Ref which worked more nicely but it has
been DEPRECATED in 2.3.1
So is Agent the best way for me to achieve this? Is this pseudocode correct?
Thanks in advance,
Meena
On Tuesday, April 8, 2014 9:44:16 PM UTC-4, MV wrote:
>
> Hello Akka experts,
>
> I have another question regarding Agent. Based on my scenario explained in
> my post - there is a situation where i need to share some mutable state
> between DIFFERENT actor.
>
> For example:
> The DatabaseWorker is happily processing data in a while loop for Request
> 1.
> Assume it has split the query into 4 times slices ….
>
> while (moreTimeSlicesToProcess)
> {
>
>
> }
>
> On Monday, April 7, 2014 1:05:06 PM UTC-4, MV wrote:
>>
>> Hi Bjorn,
>>
>> Thanks a lot for getting back to me. I was able to try out the
>> ConsistentHashoungRouter as per your suggestion and worked really nice. We
>> did a little stress test over last week and worked as expected.
>>
>> Thanks again for all the help.
>>
>> Meena
>>
>> On Tuesday, April 1, 2014 4:20:59 AM UTC-4, Björn Antonsson wrote:
>>>
>>> Hi Meena,
>>>
>>> On 27 March 2014 at 17:22:21, MV ([email protected]) wrote:
>>>
>>> Hi Endre,
>>>
>>> Thank you again and I am trying the code out. I had another quick
>>> question though - how do you release and actor reference if the specific
>>> request is ALL done?
>>> For example let's say that the Result Processor has finished all the
>>> work for the 3 data points (R1, R2, R3) and we do not have to hold this
>>> hash mapping anymore - how do I release that hash mapping?
>>> Does AKKA automatically release resources and assign that actor to
>>> another request? How is that taken care of?
>>>
>>> I apologize in advance if my wording of the question is off.
>>>
>>>
>>> There is not really a reference from a hash value to an actorRef. The
>>> actors that build up the pool of the router live on a logical ring, and the
>>> messages you send to the pool gets mapped (via hash values) onto that ring,
>>> ending up at some actor. If you send in messages that have the same hash
>>> value, they will go to the same actor. Actors don't get assigned to
>>> requests, but the requests get sent to the actors and ends up in the actors
>>> mailbox. Once the actor is finished with processing a message (request)
>>> then it takes the next message from its mailbox.
>>>
>>> Hope that helps.
>>>
>>> B/
>>>
>>> Thanks
>>> Meena
>>>
>>> On Wednesday, March 26, 2014 10:19:53 AM UTC-4, drewhk wrote:
>>>>
>>>> Hi,
>>>>
>>>>
>>>> On Wed, Mar 26, 2014 at 3:15 PM, MV <[email protected]> wrote:
>>>>
>>>>> Hi Andre,
>>>>>
>>>>> Thanks a lot for getting back to me. Yes your understanding of my
>>>>> problem is correct. Can I google for more information on
>>>>> ConsistentHashingPool and how to use it?
>>>>> I will also look into CustomRouter and let you know how that works out.
>>>>>
>>>>
>>>> It is on the same documentation page I linked:
>>>> http://doc.akka.io/docs/akka/2.3.1/scala/routing.html#ConsistentHashingPool_and_ConsistentHashingGroup
>>>>
>>>> -Endre
>>>>
>>>>
>>>>>
>>>>> Thanks
>>>>> MV
>>>>>
>>>>>
>>>>> On Wednesday, March 26, 2014 8:39:57 AM UTC-4, Akka Team wrote:
>>>>>
>>>>>> Hi MV,
>>>>>>
>>>>>> I don't see how Agents would help here. Your problem is basically
>>>>>> that the three messages for the same ID should be processed by the same
>>>>>> Result Processor. You can use ConsistentHashingPool or Group for that
>>>>>> (assuming Akka 2.3.x). If you need more specific domain dependent
>>>>>> routing
>>>>>> that is not expressible by consistent hashing you can implement your own
>>>>>> routing logic:
>>>>>> http://doc.akka.io/docs/akka/2.3.1/scala/routing.html#Custom_Router
>>>>>>
>>>>>> -Endre
>>>>>>
>>>>>>
>>>>>> On Wed, Mar 26, 2014 at 12:40 AM, MV <[email protected]> wrote:
>>>>>>
>>>>>>> Hi,
>>>>>>>
>>>>>>> I have a specific question regarding Akka Agent.
>>>>>>>
>>>>>>> My scenario is as follows:
>>>>>>>
>>>>>>> 1. I have 3 different Akka RoundRobin Workers. For example 3
>>>>>>> Request Processors, 3 Workers(Database Worker) that connect to a
>>>>>>> database
>>>>>>> and 3 workers (Result Processor) which received the database results
>>>>>>> and
>>>>>>> process the results into some kind of count table for analysis.
>>>>>>> 2. The reason I have 3 of each is that, for simulation purposes,
>>>>>>> assume 3 requests come in from the outside world into this system
>>>>>>> and each
>>>>>>> worker handles a request.
>>>>>>>
>>>>>>> My Issue:
>>>>>>>
>>>>>>> Assume that 1 (ONE) Database worker can send upto 3 or 4 data tables
>>>>>>> for a specified time slice to the Result Processor(s).
>>>>>>> The result processors have to:
>>>>>>>
>>>>>>> - From the first data ID result, create a table with some count
>>>>>>> in it.
>>>>>>> - When a second data result comes FOR THE SAME data ID update
>>>>>>> the SHARED count table.
>>>>>>> - When a third data result comes FOR THE SAME data
>>>>>>> ID, update the SHARED count table again
>>>>>>>
>>>>>>> The problem is that since DatabaseWorker is a worker of 3 instance,
>>>>>>> they can simultaneously process 3 Data Points (ID =1, ID = 2, ID =3)
>>>>>>> Since each Data Pont (ID =1 ) can produce more than 1 Result (R=1,
>>>>>>> R=2, R=3…), The Result Processor actor not only need to share the count
>>>>>>> table for each Data ID but also be able to handle all the ID's
>>>>>>> correctly
>>>>>>> and update the shared location for that SPECIFIC ID.
>>>>>>>
>>>>>>> How can I achieve this with AKKA Agent? Or is there another
>>>>>>> technique to handle this in Akka?
>>>>>>>
>>>>>>> Thanks in advance,
>>>>>>> MV
>>>>>>> --
>>>>>>> >>>>>>>>>> Read the docs: http://akka.io/docs/
>>>>>>> >>>>>>>>>> Check the FAQ:
>>>>>>> http://doc.akka.io/docs/akka/current/additional/faq.html
>>>>>>> >>>>>>>>>> Search the archives:
>>>>>>> https://groups.google.com/group/akka-user
>>>>>>> ---
>>>>>>> You received this message because you are subscribed to the Google
>>>>>>> Groups "Akka User List" group.
>>>>>>> To unsubscribe from this group and stop receiving emails from it,
>>>>>>> send an email to [email protected].
>>>>>>> To post to this group, send email to [email protected].
>>>>>>>
>>>>>>> Visit this group at http://groups.google.com/group/akka-user.
>>>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> Akka Team
>>>>>> Typesafe - The software stack for applications that scale
>>>>>> Blog: letitcrash.com
>>>>>> Twitter: @akkateam
>>>>>>
>>>>> --
>>>>> >>>>>>>>>> Read the docs: http://akka.io/docs/
>>>>> >>>>>>>>>> Check the FAQ:
>>>>> http://doc.akka.io/docs/akka/current/additional/faq.html
>>>>> >>>>>>>>>> Search the archives:
>>>>> https://groups.google.com/group/akka-user
>>>>> ---
>>>>> You received this message because you are subscribed to the Google
>>>>> Groups "Akka User List" group.
>>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>>> an email to [email protected].
>>>>> To post to this group, send email to [email protected].
>>>>> Visit this group at http://groups.google.com/group/akka-user.
>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>
>>>>
>>>> --
>>> >>>>>>>>>> Read the docs: http://akka.io/docs/
>>> >>>>>>>>>> Check the FAQ:
>>> http://doc.akka.io/docs/akka/current/additional/faq.html
>>> >>>>>>>>>> Search the archives:
>>> https://groups.google.com/group/akka-user
>>> ---
>>> You received this message because you are subscribed to the Google
>>> Groups "Akka User List" group.
>>> To unsubscribe from this group and stop receiving emails from it, send
>>> an email to [email protected].
>>> To post to this group, send email to [email protected].
>>> Visit this group at http://groups.google.com/group/akka-user.
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>>> --
>>> *Björn Antonsson*
>>> Typesafe <http://typesafe.com/> – Reactive<http://reactivemanifesto.org/>
>>> Apps
>>> on the JVM
>>> twitter: @bantonsson <http://twitter.com/#!/bantonsson>
>>>
>>> JOIN US. REGISTER TODAY! <http://www.scaladays.org/>
>>> Scala <http://www.scaladays.org/>
>>> Days <http://www.scaladays.org/>
>>> June 16th-18th, <http://www.scaladays.org/>
>>> Berlin <http://www.scaladays.org/>
>>>
>>>
--
>>>>>>>>>> Read the docs: http://akka.io/docs/
>>>>>>>>>> Check the FAQ:
>>>>>>>>>> http://doc.akka.io/docs/akka/current/additional/faq.html
>>>>>>>>>> Search the archives: https://groups.google.com/group/akka-user
---
You received this message because you are subscribed to the Google Groups "Akka
User List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/akka-user.
For more options, visit https://groups.google.com/d/optout.
