hadoop-common-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Arun C Murthy <...@yahoo-inc.com>
Subject Re: Beginners Questions....
Date Fri, 22 Aug 2008 22:46:08 GMT

On Aug 22, 2008, at 11:15 AM, Chris Gray wrote:

> All,
> I am using Hadoop using a test case set up by Michael Noll found on  
> his web
> page (http://www.michael-noll.com).
>
> I have successfully ran a job on a single Node Cluster from his  
> examples. I
> am trying to add a additional machine to the mix and set it up the  
> same way
> as the first machine. It ran a job and it ran successfully as a  
> Single Node
> Cluster in about a minute. So I attempted to do the changes that were
> necessary for a MultiNode Cluster and everything seems to be following
> Michael's example. The only thing is when I run the same job, that  
> again
> takes about a minute on a single Node Cluster, It completes in 20  
> mins with
> occasional errors about "Too Many Fetch Failures" on the two machine
> MultiNode Cluster.
>

Check the logs of your reduce tasks - the reduces are having trouble  
fetching map-outputs via http.

Arun

> We have one more machine we would like to add to the cluster for the  
> test
> but are afraid we are missing something or have messed something  
> up.  Any
> ideas or suggestions for a person just starting out?
>
> Thanks,
> Chris


Mime
View raw message