Mailing-List: contact hadoop-dev-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: hadoop-dev@lucene.apache.org
Received-SPF: neutral (asf.osuosl.org: local policy)
DomainKey-Signature: a=rsa-sha1; s=serpent; d=yahoo-inc.com; c=nofws; q=dns;
	h=mime-version:in-reply-to:references:content-type:message-id:
	content-transfer-encoding:from:subject:date:to:x-mailer;
	b=L099uGs9MeMQVzJOAnOtW+1GoCCHQvnjCGiFqXVf/wl6c5HMNeJdMnvTCVtVgJrY
Mime-Version: 1.0 (Apple Message framework v752.2)
In-Reply-To: <cac652840607060103o5cfb63a9nef9a5d40e40436e8@mail.gmail.com>
References: <cac652840607051909j60480860h4b5601a4688c45db@mail.gmail.com>
 <B736A161-1DCE-4905-B7B0-BFE552C730AD@yahoo-inc.com>
 <cac652840607060103o5cfb63a9nef9a5d40e40436e8@mail.gmail.com>
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <DE4A9625-1F5D-4D6C-B232-901AF64E30F9@yahoo-inc.com>
Content-Transfer-Encoding: 7bit
From: Eric Baldeschwieler <eric14@yahoo-inc.com>
Subject: Re: Stopping datanode and namenode
Date: Thu, 6 Jul 2006 10:42:50 -0700
To: hadoop-dev@lucene.apache.org

Thats sounds cool.  Sounds like something that would be great in the  
examples directory.  Then we could put some regressions in place to  
make sure we don't break it as we uplevel things too.

Might it make more sense to run the datanode as a distinct process  
and use the client in your servelet?  I'm no expert on web farms,  
just curious.

On Jul 6, 2006, at 1:03 AM, barry.kaplan@gmail.com wrote:

> I am not sure of what you are asking, but we are using DFS inside our
> servlet, which runs in tomcat, to allow us to distribute files in a
> cleaner, easier, method then an NFS share.
>
> To make things easier for us (especially operations) we have each of
> our servlets act as a datanode. This way each of our servers (web
> server), which only uses around 10GB of disk space, can take advantage
> of the rather cheap 40-100GB drives that are in them for the purpose
> of DFS.