You are viewing a plain text version of this content. The canonical link for it is here.
Posted to common-user@hadoop.apache.org by vishalsant <Vi...@gmail.com> on 2010/07/29 19:30:48 UTC

mapred.userlog.retain.hours

I have chnaged on my namenode and the datanodes mapred-site.xml , to include 

<property>
    <name>mapred.userlog.retain.hours</name>
    <value>2</value>
  </property>


And yet my job xml retains 24 .

Am I doing anything wrong
-- 
View this message in context: http://old.nabble.com/mapred.userlog.retain.hours-tp29298881p29298881.html
Sent from the Hadoop core-user mailing list archive at Nabble.com.


Re: mapred.userlog.retain.hours

Posted by Ted Yu <yu...@gmail.com>.
Have you restarted your cluster ?

You can actually specify this parameter in JobConf.
See the usage:
        TaskLog.cleanup(job.getInt("mapred.userlog.retain.hours", 24));
./src/mapred/org/apache/hadoop/mapred/Child.java

On Thu, Jul 29, 2010 at 10:30 AM, vishalsant <Vi...@gmail.com>wrote:

>
> I have chnaged on my namenode and the datanodes mapred-site.xml , to
> include
>
> <property>
>    <name>mapred.userlog.retain.hours</name>
>    <value>2</value>
>  </property>
>
>
> And yet my job xml retains 24 .
>
> Am I doing anything wrong
> --
> View this message in context:
> http://old.nabble.com/mapred.userlog.retain.hours-tp29298881p29298881.html
> Sent from the Hadoop core-user mailing list archive at Nabble.com.
>
>

Re: mapred.userlog.retain.hours

Posted by Edward Capriolo <ed...@gmail.com>.
On Thu, Jul 29, 2010 at 1:30 PM, vishalsant <Vi...@gmail.com> wrote:
>
> I have chnaged on my namenode and the datanodes mapred-site.xml , to include
>
> <property>
>    <name>mapred.userlog.retain.hours</name>
>    <value>2</value>
>  </property>
>
>
> And yet my job xml retains 24 .
>
> Am I doing anything wrong
> --
> View this message in context: http://old.nabble.com/mapred.userlog.retain.hours-tp29298881p29298881.html
> Sent from the Hadoop core-user mailing list archive at Nabble.com.
>
>

This parameter effects JobTrackers and TaskTrackers so you also
restart the wrong hadoop components.