hadoop-common-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From li ping <li.j...@gmail.com>
Subject Re: HDFS disk consumption.
Date Thu, 30 Dec 2010 01:09:45 GMT
If you remove the file from the command line. The file will be moved to
trash folder(if you enable the feature).otherwise the file will be moved
immediately.


On Thu, Dec 30, 2010 at 1:09 AM, Jane Chen <jxchen_us_1999@yahoo.com> wrote:

> You are right.  There's only one replica.  When does the deleted file space
> get reclaimed?
>
> --- On Tue, 12/28/10, Hemanth Yamijala <yhemanth@gmail.com> wrote:
>
> > From: Hemanth Yamijala <yhemanth@gmail.com>
> > Subject: Re: HDFS disk consumption.
> > To: common-user@hadoop.apache.org
> > Date: Tuesday, December 28, 2010, 8:43 PM
> > Hi,
> >
> > On Wed, Dec 29, 2010 at 5:51 AM, Jane Chen <jxchen_us_1999@yahoo.com>
> > wrote:
> > > Is setting dfs.replication to 1 sufficient to stop
> > replication?  How do I verify that?  I have a pseudo
> > cluster running 0.21.0.  It seems that the hdfs disk
> > consumption triples the amount of data stored.
> >
> > Setting to 1 is sufficient to stop replication. Can you
> > check if the
> > web UI for NameNode has a way to show the replicas of
> > blocks for a
> > file ?
> >
> > >
> > > Thanks,
> > > Jane
> > >
> > >
> > >
> > >
> >
>
>
>
>


-- 
-----李平

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message