Mailing-List: contact yarn-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Date: Sat, 21 May 2016 18:03:12 +0000 (UTC)
From: "Lavkesh Lahngir (JIRA)" <jira@apache.org>
To: yarn-issues@hadoop.apache.org
Message-ID: <JIRA.12971813.1463852619000.259443.1463853792880@Atlassian.JIRA>
In-Reply-To: <JIRA.12971813.1463852619000@Atlassian.JIRA>
References: <JIRA.12971813.1463852619000@Atlassian.JIRA> <JIRA.12971813.1463852619723@arcas>
Subject: [jira] [Comment Edited] (YARN-5123) SQL based RM state store
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
archived-at: Sat, 21 May 2016 18:03:14 -0000


    [ https://issues.apache.org/jira/browse/YARN-5123?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15295166#comment-15295166 ] 

Lavkesh Lahngir edited comment on YARN-5123 at 5/21/16 6:02 PM:
----------------------------------------------------------------

Config definitions.

RM_DB_TYPE -> it's the name of jdbc driver, in our case it would be "postgresql".
RM_DB_HOST -> DB host name.
RM_DB_DBNAME -> Database name.
RM_DB_USERNAME -> Database user.
RM_DB_PASSWORD -> Database password.
RM_DB_PASSWORD_FILE -> Password file from which the state store will read the password. In our setup we have password file with special permission. If the password file is not provided, the value of RM_DB_PASSWORD will be used.
RM_DB_RETRIES -> number of retries before any DB operation is given up. If these retries are exhausted RM will go on stand by.
DEFAULT_RM_DB_RETRIES -> Default retries 
RM_DB_RETRIES_INTERVAL -> Fixed interval between retries 
DEFAULT_RM_DB_RETRIES_INTERVAL -> Default interval for retries in seconds.
RM_DB_VERIFICATION_TIMEOUT -> A timeout interval in milliseconds for the verification thread which keeps checking the status of active RM.
DEFAULT_RM_DB_VERIFICATION_TIMEOUT -> Default verification timeout


was (Author: lavkesh):
RM_DB_TYPE -> it's the name of jdbc driver, in our case it would be "postgresql".
RM_DB_HOST -> DB host name.
RM_DB_DBNAME -> Database name.
RM_DB_USERNAME -> Database user.
RM_DB_PASSWORD -> Database password.
RM_DB_PASSWORD_FILE -> Password file from which the state store will read the password. In our setup we have password file with special permission. If the password file is not provided, the value of RM_DB_PASSWORD will be used.
RM_DB_RETRIES -> number of retries before any DB operation is given up. If these retries are exhausted RM will go on stand by.
DEFAULT_RM_DB_RETRIES -> Default retries 
RM_DB_RETRIES_INTERVAL -> Fixed interval between retries 
DEFAULT_RM_DB_RETRIES_INTERVAL -> Default interval for retries in seconds.
RM_DB_VERIFICATION_TIMEOUT -> A timeout interval in milliseconds for the verification thread which keeps checking the status of active RM.
DEFAULT_RM_DB_VERIFICATION_TIMEOUT -> Default verification timeout

> SQL based RM state store
> ------------------------
>
>                 Key: YARN-5123
>                 URL: https://issues.apache.org/jira/browse/YARN-5123
>             Project: Hadoop YARN
>          Issue Type: New Feature
>    Affects Versions: 2.6.0
>            Reporter: Lavkesh Lahngir
>         Attachments: sqlstatestore.patch
>
>
> In our setup,  zookeeper based RM state store didn't work. We ended up implementing our own SQL based state store. Here is a patch, if anybody else wants to use it. 


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

---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: yarn-issues-help@hadoop.apache.org