Return-Path: <dev-return-13004-archive-asf-public=cust-asf.ponee.io@reef.apache.org>
X-Original-To: archive-asf-public-internal@cust-asf2.ponee.io
Delivered-To: archive-asf-public-internal@cust-asf2.ponee.io
Received: from cust-asf.ponee.io (cust-asf.ponee.io [163.172.22.183])
	by cust-asf2.ponee.io (Postfix) with ESMTP id 2EE04200B89
	for <archive-asf-public-internal@cust-asf2.ponee.io>; Wed, 21 Sep 2016 22:12:22 +0200 (CEST)
Received: by cust-asf.ponee.io (Postfix)
	id 2D818160ADB; Wed, 21 Sep 2016 20:12:22 +0000 (UTC)
Delivered-To: archive-asf-public@cust-asf.ponee.io
Received: from mail.apache.org (hermes.apache.org [140.211.11.3])
	by cust-asf.ponee.io (Postfix) with SMTP id 69FE7160ABC
	for <archive-asf-public@cust-asf.ponee.io>; Wed, 21 Sep 2016 22:12:21 +0200 (CEST)
Received: (qmail 31113 invoked by uid 500); 21 Sep 2016 20:12:20 -0000
Mailing-List: contact dev-help@reef.apache.org; run by ezmlm
Precedence: bulk
List-Help: <mailto:dev-help@reef.apache.org>
List-Unsubscribe: <mailto:dev-unsubscribe@reef.apache.org>
List-Post: <mailto:dev@reef.apache.org>
List-Id: <dev.reef.apache.org>
Reply-To: dev@reef.apache.org
Delivered-To: mailing list dev@reef.apache.org
Received: (qmail 31096 invoked by uid 99); 21 Sep 2016 20:12:20 -0000
Received: from arcas.apache.org (HELO arcas) (140.211.11.28)
    by apache.org (qpsmtpd/0.29) with ESMTP; Wed, 21 Sep 2016 20:12:20 +0000
Received: from arcas.apache.org (localhost [127.0.0.1])
	by arcas (Postfix) with ESMTP id 6EDE02C0050
	for <dev@reef.apache.org>; Wed, 21 Sep 2016 20:12:20 +0000 (UTC)
Date: Wed, 21 Sep 2016 20:12:20 +0000 (UTC)
From: "Julia (JIRA)" <jira@apache.org>
To: dev@reef.apache.org
Message-ID: <JIRA.12972262.1464056015000.630521.1474488740449@Atlassian.JIRA>
In-Reply-To: <JIRA.12972262.1464056015000@Atlassian.JIRA>
References: <JIRA.12972262.1464056015000@Atlassian.JIRA> <JIRA.12972262.1464056015047@arcas>
Subject: [jira] [Commented] (REEF-1404) IMRU task state Maintenance and
 Preservation in Evaluator for fault tolerant
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
X-JIRA-FingerPrint: 30527f35849b9dde25b450d4833f0394
archived-at: Wed, 21 Sep 2016 20:12:22 -0000


    [ https://issues.apache.org/jira/browse/REEF-1404?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15511023#comment-15511023 ] 

Julia commented on REEF-1404:
-----------------------------

To summaries, there are two ways to do the State management. 

Use empty interface
* Define an empty interface IStateManager at Evaluator.Context so that both context and task can inject from it and share the same singleton
* Client like IMRU can define UpdateTaskState and MapTaskState which implement IStateManager and wraps the data.e.g.
{noformat}
       internal class UpdateTaskState<TMapInput, TResult> : IStateManager
    {
        private static readonly Logger Logger = Logger.GetLogger(typeof(UpdateTaskState<TMapInput, TResult>));

        /// <summary>
        /// The state that needs to be reserved
        /// </summary>
        internal UpdateResult<TMapInput, TResult> State { get; set; }

        [Inject]
        private UpdateTaskState()
        {
            Logger.Log(Level.Info, "UpdateTaskState constructor");
        }
    }
{noformat}
* pros - The injection at context is easier as there is no T involved. 
* cons - There is no API defined. It is up to the user how to defined the State method. Gives user flexibility but also no constrains. 

Define APIs in the interface
* IStatemanager<T> has well defined interface 
{noformat}
        void AddState(T stateObject);
        T GetLatestState();
{noformat}
* Client like IMRU can have its own implementation for the interface
* Client can replace T with any class which contains the data they want to reserve.
* pros - As the API mostly will end up as get and set data, fixed API signature is clear.
* corns - The client needs to specify the type T for the injection in context. 

ICodec<IStateManager> can be implemented/injected if serialization is needed. 

> IMRU task state Maintenance and Preservation in Evaluator for fault tolerant
> ----------------------------------------------------------------------------
>
>                 Key: REEF-1404
>                 URL: https://issues.apache.org/jira/browse/REEF-1404
>             Project: REEF
>          Issue Type: Task
>            Reporter: Julia
>              Labels: FT
>
> IMRU task should be able to 
> * Maintenance and preservation the state
> * When restart, able to recover from the previous sate



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)