hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Vinod K V (JIRA)" <j...@apache.org>
Subject [jira] Resolved: (HADOOP-4375) Accesses to CompletedJobStore should not lock the JobTracker.
Date Mon, 23 Feb 2009 07:57:01 GMT

     [ https://issues.apache.org/jira/browse/HADOOP-4375?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Vinod K V resolved HADOOP-4375.
-------------------------------

    Resolution: Duplicate

HADOOP-5285 fixed this.

> Accesses to CompletedJobStore should not lock the JobTracker.
> -------------------------------------------------------------
>
>                 Key: HADOOP-4375
>                 URL: https://issues.apache.org/jira/browse/HADOOP-4375
>             Project: Hadoop Core
>          Issue Type: Bug
>            Reporter: Vinod K V
>
> When CompletedJobStatusStore is configured to use dfs, it can cause lockup of JT due
to inline dfs access.
> For e.g.
> {code}
>   public synchronized JobStatus getJobStatus(JobID jobid) {
>     JobInProgress job = jobs.get(jobid);
>     if (job != null) {
>       return job.getStatus();
>     } else {
>       return completedJobStatusStore.readJobStatus(jobid);
>     }
>   }
> {code}
> Instead, JT should be locked only when accessing its internal memory structures. The
other such methods are JobTracker.getJobCounters, JobTracker.getJobProfile and JobTracker.finalizeJob.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message