hadoop-common-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Eli Collins (Commented) (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HADOOP-7991) HA: the FailoverController should check the standby is ready before failing over
Date Thu, 02 Feb 2012 19:10:54 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-7991?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13199108#comment-13199108
] 

Eli Collins commented on HADOOP-7991:
-------------------------------------

Yea, I tested putting the SBN in SM, failover, which fails, then failover with forceactive
which succeeds even though the SBN is still in SM.

{noformat}
hadoop-0.24.0-SNAPSHOT $ ./bin/hdfs dfsadmin -fs hdfs://localhost:8022 -safemode enter
Safe mode is ON
hadoop-0.24.0-SNAPSHOT $ ./bin/hdfs haadmin -failover localhost:8021 localhost:8022
Failover failed: localhost/127.0.0.1:8022 is not ready to become active
hadoop-0.24.0-SNAPSHOT $ ./bin/hdfs haadmin -failover localhost:8021 localhost:8022 --forceactive
Failover from localhost:8021 to localhost:8022 successful
hadoop-0.24.0-SNAPSHOT $ /bin/hdfs dfsadmin -fs hdfs://localhost:8022 -safemode get
Safe mode is ON
{noformat}

                
> HA: the FailoverController should check the standby is ready before failing over
> --------------------------------------------------------------------------------
>
>                 Key: HADOOP-7991
>                 URL: https://issues.apache.org/jira/browse/HADOOP-7991
>             Project: Hadoop Common
>          Issue Type: Sub-task
>          Components: ha
>    Affects Versions: HA Branch (HDFS-1623)
>            Reporter: Eli Collins
>            Assignee: Eli Collins
>         Attachments: hadoop-7991.txt
>
>
> The FC in preFailoverChecks should check that the target service is ready for failover
(eg in the case of HDFS, that it's not in SM) before failing over to it. We'll need to provide
a force option as well since the standby may not actually be ready until the failover has
been initiated (eg because becoming active kicks it to check the log and see newly allocated
blocks).

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message