hadoop-common-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Amandeep Khurana <ama...@gmail.com>
Subject Re: NN + secondary got full, even though data nodes had plenty of space
Date Wed, 12 Aug 2009 08:38:41 GMT
Is your NN doubling up as a DN? If its not, I wonder how the NN is full...

On 8/11/09, Mayuran Yogarajah <mayuran.yogarajah@casalemedia.com> wrote:
> I have a 6 node cluster running Hadoop 0.18.3.  I'm trying to figure out
> how the data was spread out like this:
>
> node001         94.15%
> node002         94.16%
> node003         48.22%
> node004         47.85%
> node005         48.12%
> node006         43.18%
>
> Node 001 (NN) and node 002( secondary NN) both got full, while the other
> data nodes had more space left.  I had assumed that Hadoop would distribute
> more blocks to nodes 3-6 since they had much more space, but it ended up
> filling up nodes1 and 2.  Is this expected?
>
> thanks,
> M
>
>


-- 


Amandeep Khurana
Computer Science Graduate Student
University of California, Santa Cruz

Mime
View raw message