Mailing-List: contact issues-help@ignite.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@ignite.apache.org
Date: Wed, 30 Nov 2016 08:18:58 +0000 (UTC)
From: "Alexey Goncharuk (JIRA)" <jira@apache.org>
To: issues@ignite.apache.org
Message-ID: <JIRA.13024351.1480493922000.398612.1480493938940@Atlassian.JIRA>
In-Reply-To: <JIRA.13024351.1480493922000@Atlassian.JIRA>
References: <JIRA.13024351.1480493922000@Atlassian.JIRA> <JIRA.13024351.1480493922242@arcas>
Subject: [jira] [Created] (IGNITE-4335) Implement cluster ACTIVE/INACTIVE
 state
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
archived-at: Wed, 30 Nov 2016 08:19:00 -0000

Alexey Goncharuk created IGNITE-4335:
----------------------------------------

             Summary: Implement cluster ACTIVE/INACTIVE state
                 Key: IGNITE-4335
                 URL: https://issues.apache.org/jira/browse/IGNITE-4335
             Project: Ignite
          Issue Type: New Feature
          Components: general
            Reporter: Alexey Goncharuk
             Fix For: 2.0


I think it might be beneficial in some cases to start cluster in some form of inactive state. In this case we can skip start of many managers and processors (basically, start only discovery + communication), which should speed up the cluster start process.

Once all nodes are started, an API call or a shell command can activate the cluster and start all managers in one pass. This should significantly reduce traffic, affinity calculations, etc.


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