[ 
https://issues.apache.org/jira/browse/TRAFODION-2650?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16354291#comment-16354291
 ] 

ASF GitHub Bot commented on TRAFODION-2650:
-------------------------------------------

Github user zcorrea commented on a diff in the pull request:

    https://github.com/apache/trafodion/pull/1437#discussion_r166396386
  
    --- Diff: core/sqf/sqenvcom.sh ---
    @@ -47,6 +47,15 @@ export PRODUCT_COPYRIGHT_HEADER="2015-2017 Apache 
Software Foundation"
     ##############################################################
     export 
TRAFODION_ENABLE_AUTHENTICATION=${TRAFODION_ENABLE_AUTHENTICATION:-NO}
     
    +# Set the Trafodion Configuration store type
    +if [[ -n "$CLUSTERNAME" ]]; then
    +  # This is a cluster environment, not a workstation
    +  #export TRAF_CONFIG_DBSTORE=MySQL
    +  export TRAF_CONFIG_DBSTORE=Sqlite
    +else
    +  export TRAF_CONFIG_DBSTORE=Sqlite
    --- End diff --
    
    This is an initial step to prepare the code base before adding the MySQL 
storage method.


> The SQLite Trafodion Configuration database storage method can become stale.
> ----------------------------------------------------------------------------
>
>                 Key: TRAFODION-2650
>                 URL: https://issues.apache.org/jira/browse/TRAFODION-2650
>             Project: Apache Trafodion
>          Issue Type: Bug
>          Components: foundation
>    Affects Versions: 2.2-incubating
>            Reporter: Gonzalo E Correa
>            Assignee: Gonzalo E Correa
>            Priority: Major
>             Fix For: 2.3
>
>
> The SQLite storage method can prevent a node from initializing properly. This 
> is due to the nature of SQLIte, a database in a file. The file is updated 
> locally on each node. In certain failure scenarios, such as Cloudera Manager 
> and potentially Ambari installations, the file will not be updated locally. 
> This happens when the monitor process is not running and changes are made to 
> the configuration, the file in the node where the monitor process is not 
> executing will become stale and on a node up event, may contain old 
> configuration information.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Reply via email to