Mailing-List: contact issues-help@ambari.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@ambari.apache.org
Date: Mon, 11 Apr 2016 04:26:25 +0000 (UTC)
From: "Sumit Mohanty (JIRA)" <jira@apache.org>
To: issues@ambari.apache.org
Message-ID: <JIRA.12957604.1460348690000.189065.1460348785449@Atlassian.JIRA>
In-Reply-To: <JIRA.12957604.1460348690000@Atlassian.JIRA>
References: <JIRA.12957604.1460348690000@Atlassian.JIRA>
 <JIRA.12957604.1460348690748@arcas>
Subject: [jira] [Updated] (AMBARI-15795) Parallel execution should only be
 allowed on commands that have auto retry enabled
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


     [ https://issues.apache.org/jira/browse/AMBARI-15795?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Sumit Mohanty updated AMBARI-15795:
-----------------------------------
    Status: Patch Available  (was: Open)

> Parallel execution should only be allowed on commands that have auto retry enabled
> ----------------------------------------------------------------------------------
>
>                 Key: AMBARI-15795
>                 URL: https://issues.apache.org/jira/browse/AMBARI-15795
>             Project: Ambari
>          Issue Type: Bug
>          Components: ambari-agent
>    Affects Versions: 2.2.2
>            Reporter: Sumit Mohanty
>            Assignee: Sumit Mohanty
>             Fix For: 2.2.2
>
>         Attachments: AMBARI-15795.patch
>
>
> Parallel execution of commands can result in failures if the commands process resources that are shared among the commands. For example, commands may check and delete the same file resource. Deleting a file resource that is already deleted is not a functional error but its does result in command failure.
> Its not always possible to guarantee that stack scripts are written with guards that can check the failure and then determine if its harmless or not. Auto retry is a way to get around failures due to resource contention - think eventual success. Try all commands in parallel - then try only the failed ones. Each command is idempotent by definition so after a few retries all commands succeed.
> When commands without auto-retry option are executed in parallel then there is no option to retry. This is not desirable.


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