Mailing-List: contact dev-help@hama.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@hama.apache.org
Received-SPF: pass (athena.apache.org: domain of thomas.jungblut@gmail.com
 designates 209.85.220.175 as permitted sender)
MIME-Version: 1.0
In-Reply-To: 
 <CAGQgZQRcYkNhKbmX+R2iVD4j1LLAXuc8dUHyDh9k80TW2bRi1g@mail.gmail.com>
References: 
 <CAGQgZQRcYkNhKbmX+R2iVD4j1LLAXuc8dUHyDh9k80TW2bRi1g@mail.gmail.com>
Date: Thu, 23 Aug 2012 09:27:14 +0200
Message-ID: 
 <CAJ-=ysmhX9sG3zgM9OjGA19qQCh+-cc9N7zGkmUenjWwKnBXuQ@mail.gmail.com>
Subject: Re: K-means test scenario
From: Thomas Jungblut <thomas.jungblut@gmail.com>
To: "Edward J. Yoon" <edwardyoon@apache.org>
Cc: dev@hama.apache.org
Content-Type: multipart/alternative; boundary=20cf307f380edea45404c7e9c9f7

--20cf307f380edea45404c7e9c9f7
Content-Type: text/plain; charset=ISO-8859-1

Hi,

yet another benchmark?^^ Just do the same stuff like you did last time.
What I would find very interesting, is how much the disk caching comes into
play in terms of performance. You can configure whether you want to cache
the input in the heap or rely on OS disk caching.

But don't hesitate, I don't think this is very urgent.

2012/8/23 Edward J. Yoon <edwardyoon@apache.org>

> Hi thomas,
>
> I wanna test your k-means and evaluate its performance tomorrow.
> What's the best scenario to test k-means?
>
> --
> Best Regards, Edward J. Yoon
> @eddieyoon
>

--20cf307f380edea45404c7e9c9f7--