Mailing-List: contact dev-help@ambari.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@ambari.apache.org
Date: Mon, 14 Dec 2015 14:14:46 +0000 (UTC)
From: "Andrew Onischuk (JIRA)" <jira@apache.org>
To: dev@ambari.apache.org
Message-ID: <JIRA.12921674.1450102455000.14458.1450102486671@Atlassian.JIRA>
In-Reply-To: <JIRA.12921674.1450102455000@Atlassian.JIRA>
References: <JIRA.12921674.1450102455000@Atlassian.JIRA>
 <JIRA.12921674.1450102455293@arcas>
Subject: [jira] [Created] (AMBARI-14366) Create a script to allow config
 merge during Ambari takeover
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit

Andrew Onischuk created AMBARI-14366:
----------------------------------------

             Summary: Create a script to allow config merge during Ambari takeover
                 Key: AMBARI-14366
                 URL: https://issues.apache.org/jira/browse/AMBARI-14366
             Project: Ambari
          Issue Type: Bug
            Reporter: Andrew Onischuk
            Assignee: Andrew Onischuk
             Fix For: 2.2.1
         Attachments: AMBARI-14366.patch

_Background_:

  * Paypal 900 node cluster Ambari takeover needs to address the problem of host based configs merged to create default Ambari configs.
  * The scope of this task is to provide a script that reads a dir structure for all configs. **For now only site.xml** since env files cannot be parsed, we might need to handle yaml files in later version.
  * It produces a output of all the problems / conflicts found along with a configuration.json which can be directly incorporated in a cluster **blueprint**

_Tasks_:

  * Please find the attached patch with a partial script that you can start from (Really basic foundation patch).
  * Assume that user has copied all the configs in a directory, example:  
/tmp/my-master-configs/etc/hadoop/conf,  
/tmp/my-master-configs/etc/hbase/conf,  
/etc/my-slave-configs/hadoop/conf and so on. (There could be 5 core-site from
different master / slave / client)

  * Parse these files and merge them into a map of configs that can be exported to json based on following rules: 
    * If propertyA is found in core-site1 and core-site2 with same value, add it to the map
    * If propertyA is found in core-site1 only, add it to the map
    * If propertyA is found in core-site1 and core-site2 with different value, add it to an output file with same name as the config-type, as a conflict.  
Sample file (Choose appropriate formatting of choice) ::

        
                core-site-conflicts.txt ::
        || Conflicts ||
        | propertyA  | <path-to-config-file> | value |
        | propertyA  | <path-to-config-file> | value |
        

    * **Note**: Single run should generate all the conflicts, if no conflicts found generate the json file
    * Provide ability to force output and also a verbose mode to log every step


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