Mailing-List: contact dev-help@flink.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@flink.apache.org
Date: Wed, 13 Jan 2016 16:08:39 +0000 (UTC)
From: "Ufuk Celebi (JIRA)" <jira@apache.org>
To: dev@flink.apache.org
Message-ID: <JIRA.12929861.1452701262000.99063.1452701319990@Atlassian.JIRA>
In-Reply-To: <JIRA.12929861.1452701262000@Atlassian.JIRA>
References: <JIRA.12929861.1452701262000@Atlassian.JIRA>
 <JIRA.12929861.1452701262114@arcas>
Subject: [jira] [Created] (FLINK-3232) Add option to eagerly deploy channels
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit

Ufuk Celebi created FLINK-3232:
----------------------------------

             Summary: Add option to eagerly deploy channels
                 Key: FLINK-3232
                 URL: https://issues.apache.org/jira/browse/FLINK-3232
             Project: Flink
          Issue Type: Improvement
          Components: Distributed Runtime
    Affects Versions: 0.10.1
            Reporter: Ufuk Celebi
            Assignee: Ufuk Celebi
            Priority: Minor
             Fix For: 1.0.0


Intermediate partitions are consumed via input channels. These channels are instantiated lazily when the partition producer emits the first record. This can lead to higher latencies for the first records passing the pipeline.

In order to decrease this latency, we can deploy channels eagerly.

I would like to add this as a flag to the intermediate results of the execution graph. On deployment, the task can then deploy the input channels as soon as the intermediate stream has been registered at the partition manager.


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