http://www.fit-pc.com/web/fit-pc2-models/ - I wonder if there is any
reason why Hadoop & Accumulo would not work on this kind of computer.
Any opinions? Buying a few desktops for my home use will make my
office too hot to use.

Doesn't seem like enough RAM or disk to me. I guess it depends on what
you were wanting to process though. Ive had a localhost setup of
Hadoop/ Zookeeper/ Accumulo on my macbook with 4GB of RAM for testing
but its brutal. Is there a reason to not just go with an EC2 setup on
amazon?

Sent from my iPhone

On Jul 7, 2012, at 11:14 AM, David Medinets <[EMAIL PROTECTED]> wrote:

> http://www.fit-pc.com/web/fit-pc2-models/ - I wonder if there is any
> reason why Hadoop & Accumulo would not work on this kind of computer.
> Any opinions? Buying a few desktops for my home use will make my
> office too hot to use.

Part of the impulse is the novelty. But also, I am concerned about the
price of EC2; how pausing the instances would affect accumulo; or with
continuously loading data.

With something like the Fit PC I can leave them running all the time
in my office. Of course, simply getting a Linode 512 from
http://www.linode.com/ might be better.

On Sat, Jul 7, 2012 at 11:37 AM, Jim Klucar <[EMAIL PROTECTED]> wrote:
> Doesn't seem like enough RAM or disk to me. I guess it depends on what
> you were wanting to process though. Ive had a localhost setup of
> Hadoop/ Zookeeper/ Accumulo on my macbook with 4GB of RAM for testing
> but its brutal. Is there a reason to not just go with an EC2 setup on
> amazon?
>
> Sent from my iPhone
>
> On Jul 7, 2012, at 11:14 AM, David Medinets <[EMAIL PROTECTED]> wrote:
>
>> http://www.fit-pc.com/web/fit-pc2-models/ - I wonder if there is any
>> reason why Hadoop & Accumulo would not work on this kind of computer.
>> Any opinions? Buying a few desktops for my home use will make my
>> office too hot to use.

I took my own advice to buy a Linode 512 system. It took less than 40
minutes to have accumulo v1.5 working on it - start to finish. I
should have done this a long time ago.

Since 1.4 we've provided configurations that fit intoas small as 512mb of
ram. While the user cases are limited, we provide these for users to get a
taste of running Accumulo without servers class systems.

Sent from my phone, so pardon the typos and brevity.
On Jul 7, 2012 11:14 AM, "David Medinets" <[EMAIL PROTECTED]> wrote:

> http://www.fit-pc.com/web/fit-pc2-models/ - I wonder if there is any
> reason why Hadoop & Accumulo would not work on this kind of computer.
> Any opinions? Buying a few desktops for my home use will make my
> office too hot to use.
>

Reminded me of Raspberry Pi: http://downloads.element14.com/raspberryPi2.html?SKU=83T1943

Accumulo + Hadoop + Raspberry Pi! :)

Cheers,
Chris

On Jul 7, 2012, at 8:13 AM, David Medinets wrote:

> http://www.fit-pc.com/web/fit-pc2-models/ - I wonder if there is any
> reason why Hadoop & Accumulo would not work on this kind of computer.
> Any opinions? Buying a few desktops for my home use will make my
> office too hot to use.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Chris Mattmann, Ph.D.
Senior Computer Scientist
NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA
Office: 171-266B, Mailstop: 171-246
Email: [EMAIL PROTECTED]
WWW:   http://sunset.usc.edu/~mattmann/
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Adjunct Assistant Professor, Computer Science Department
University of Southern California, Los Angeles, CA 90089 USA
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Do it!

On Sat, Jul 7, 2012 at 3:47 PM, Mattmann, Chris A (388J) <
[EMAIL PROTECTED]> wrote:

> Reminded me of Raspberry Pi:
> http://downloads.element14.com/raspberryPi2.html?SKU=83T1943
>
> Accumulo + Hadoop + Raspberry Pi! :)
>
> Cheers,
> Chris
>
> On Jul 7, 2012, at 8:13 AM, David Medinets wrote:
>
> > http://www.fit-pc.com/web/fit-pc2-models/ - I wonder if there is any
> > reason why Hadoop & Accumulo would not work on this kind of computer.
> > Any opinions? Buying a few desktops for my home use will make my
> > office too hot to use.
>
>
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> Chris Mattmann, Ph.D.
> Senior Computer Scientist
> NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA
> Office: 171-266B, Mailstop: 171-246
> Email: [EMAIL PROTECTED]
> WWW:   http://sunset.usc.edu/~mattmann/
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> Adjunct Assistant Professor, Computer Science Department
> University of Southern California, Los Angeles, CA 90089 USA
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>
>

You could probably do it, but only small amounts of data. You can cut down
on running so many processes by not running HDFS.
On Jul 7, 2012 11:38 AM, "Jim Klucar" <[EMAIL PROTECTED]> wrote:

> Doesn't seem like enough RAM or disk to me. I guess it depends on what
> you were wanting to process though. Ive had a localhost setup of
> Hadoop/ Zookeeper/ Accumulo on my macbook with 4GB of RAM for testing
> but its brutal. Is there a reason to not just go with an EC2 setup on
> amazon?
>
> Sent from my iPhone
>
> On Jul 7, 2012, at 11:14 AM, David Medinets <[EMAIL PROTECTED]>
> wrote:
>
> > http://www.fit-pc.com/web/fit-pc2-models/ - I wonder if there is any
> > reason why Hadoop & Accumulo would not work on this kind of computer.
> > Any opinions? Buying a few desktops for my home use will make my
> > office too hot to use.
>

Hdfs is necessary for Accumulo, as we currently don't support running of
the local filesystem. You can cut out MapReduce though.

Sent from my phone, so pardon the typos and brevity.
On Jul 8, 2012 8:48 AM, "Ed Kohlwey" <[EMAIL PROTECTED]> wrote:

> You could probably do it, but only small amounts of data. You can cut down
> on running so many processes by not running HDFS.
> On Jul 7, 2012 11:38 AM, "Jim Klucar" <[EMAIL PROTECTED]> wrote:
>
> > Doesn't seem like enough RAM or disk to me. I guess it depends on what
> > you were wanting to process though. Ive had a localhost setup of
> > Hadoop/ Zookeeper/ Accumulo on my macbook with 4GB of RAM for testing
> > but its brutal. Is there a reason to not just go with an EC2 setup on
> > amazon?
> >
> > Sent from my iPhone
> >
> > On Jul 7, 2012, at 11:14 AM, David Medinets <[EMAIL PROTECTED]>
> > wrote:
> >
> > > http://www.fit-pc.com/web/fit-pc2-models/ - I wonder if there is any
> > > reason why Hadoop & Accumulo would not work on this kind of computer.
> > > Any opinions? Buying a few desktops for my home use will make my
> > > office too hot to use.
> >
>

Hdfs is required as we don't support running over the local filesystem.
However, you can cut out MapReduce related tasks.

Sent from my phone, so pardon the typos and brevity.
On Jul 8, 2012 8:48 AM, "Ed Kohlwey" <[EMAIL PROTECTED]> wrote:

> You could probably do it, but only small amounts of data. You can cut down
> on running so many processes by not running HDFS.
> On Jul 7, 2012 11:38 AM, "Jim Klucar" <[EMAIL PROTECTED]> wrote:
>
> > Doesn't seem like enough RAM or disk to me. I guess it depends on what
> > you were wanting to process though. Ive had a localhost setup of
> > Hadoop/ Zookeeper/ Accumulo on my macbook with 4GB of RAM for testing
> > but its brutal. Is there a reason to not just go with an EC2 setup on
> > amazon?
> >
> > Sent from my iPhone
> >
> > On Jul 7, 2012, at 11:14 AM, David Medinets <[EMAIL PROTECTED]>
> > wrote:
> >
> > > http://www.fit-pc.com/web/fit-pc2-models/ - I wonder if there is any
> > > reason why Hadoop & Accumulo would not work on this kind of computer.
> > > Any opinions? Buying a few desktops for my home use will make my
> > > office too hot to use.
> >
>

Chris,
Have you actually tried a Raspberry Pi Hadoop cluster?  I've been considering it for a while, but I'd love to know the results if anyone has actually done it.

Tejay

-----Original Message-----
From: Mattmann, Chris A (388J) [mailto:[EMAIL PROTECTED]]
Sent: Saturday, July 07, 2012 1:48 PM
To: <[EMAIL PROTECTED]>
Subject: EXTERNAL: Re: Using a really small computer for accumulo experimentation.

Reminded me of Raspberry Pi: http://downloads.element14.com/raspberryPi2.html?SKU=83T1943

Accumulo + Hadoop + Raspberry Pi! :)

Cheers,
Chris

On Jul 7, 2012, at 8:13 AM, David Medinets wrote:

> http://www.fit-pc.com/web/fit-pc2-models/ - I wonder if there is any
> reason why Hadoop & Accumulo would not work on this kind of computer.
> Any opinions? Buying a few desktops for my home use will make my
> office too hot to use.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Chris Mattmann, Ph.D.
Senior Computer Scientist
NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA
Office: 171-266B, Mailstop: 171-246
Email: [EMAIL PROTECTED]
WWW:   http://sunset.usc.edu/~mattmann/
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

A raspberry pi cluster is an interesting idea. Other people have toyed with
the idea of lots of small, low wattage hosts attached to relatively slow,
cheap storage and making up the difference in parallelism.
http://www.cs.cmu.edu/~fawnproj/

On Mon, Jul 9, 2012 at 9:53 AM, Cardon, Tejay E <[EMAIL PROTECTED]>wrote:

> Chris,
> Have you actually tried a Raspberry Pi Hadoop cluster?  I've been
> considering it for a while, but I'd love to know the results if anyone has
> actually done it.
>
> Tejay
>
> -----Original Message-----
> From: Mattmann, Chris A (388J) [mailto:[EMAIL PROTECTED]]
> Sent: Saturday, July 07, 2012 1:48 PM
> To: <[EMAIL PROTECTED]>
> Subject: EXTERNAL: Re: Using a really small computer for accumulo
> experimentation.
>
> Reminded me of Raspberry Pi:
> http://downloads.element14.com/raspberryPi2.html?SKU=83T1943
>
> Accumulo + Hadoop + Raspberry Pi! :)
>
> Cheers,
> Chris
>
> On Jul 7, 2012, at 8:13 AM, David Medinets wrote:
>
> > http://www.fit-pc.com/web/fit-pc2-models/ - I wonder if there is any
> > reason why Hadoop & Accumulo would not work on this kind of computer.
> > Any opinions? Buying a few desktops for my home use will make my
> > office too hot to use.
>
>
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> Chris Mattmann, Ph.D.
> Senior Computer Scientist
> NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA
> Office: 171-266B, Mailstop: 171-246
> Email: [EMAIL PROTECTED]
> WWW:   http://sunset.usc.edu/~mattmann/
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> Adjunct Assistant Professor, Computer Science Department University of
> Southern California, Los Angeles, CA 90089 USA
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>
>

You can Accumulo on the local filesystem by setting instance.dfs.uri,
but its not well tested.  I have tried it once.

On Sun, Jul 8, 2012 at 10:11 AM, John Vines <[EMAIL PROTECTED]> wrote:
> Hdfs is necessary for Accumulo, as we currently don't support running of
> the local filesystem. You can cut out MapReduce though.
>
> Sent from my phone, so pardon the typos and brevity.
> On Jul 8, 2012 8:48 AM, "Ed Kohlwey" <[EMAIL PROTECTED]> wrote:
>
>> You could probably do it, but only small amounts of data. You can cut down
>> on running so many processes by not running HDFS.
>> On Jul 7, 2012 11:38 AM, "Jim Klucar" <[EMAIL PROTECTED]> wrote:
>>
>> > Doesn't seem like enough RAM or disk to me. I guess it depends on what
>> > you were wanting to process though. Ive had a localhost setup of
>> > Hadoop/ Zookeeper/ Accumulo on my macbook with 4GB of RAM for testing
>> > but its brutal. Is there a reason to not just go with an EC2 setup on
>> > amazon?
>> >
>> > Sent from my iPhone
>> >
>> > On Jul 7, 2012, at 11:14 AM, David Medinets <[EMAIL PROTECTED]>
>> > wrote:
>> >
>> > > http://www.fit-pc.com/web/fit-pc2-models/ - I wonder if there is any
>> > > reason why Hadoop & Accumulo would not work on this kind of computer.
>> > > Any opinions? Buying a few desktops for my home use will make my
>> > > office too hot to use.
>> >
>>

Hi Tejay,

Thanks, no I haven't tried, but would love to. I missed the first Raspberry Pi window
but will be checking each day when they open sales back up in July.

Thanks!

Cheers,
Chris

On Jul 9, 2012, at 6:53 AM, Cardon, Tejay E wrote:

> Chris,
> Have you actually tried a Raspberry Pi Hadoop cluster?  I've been considering it for a while, but I'd love to know the results if anyone has actually done it.
>
> Tejay
>
> -----Original Message-----
> From: Mattmann, Chris A (388J) [mailto:[EMAIL PROTECTED]]
> Sent: Saturday, July 07, 2012 1:48 PM
> To: <[EMAIL PROTECTED]>
> Subject: EXTERNAL: Re: Using a really small computer for accumulo experimentation.
>
> Reminded me of Raspberry Pi: http://downloads.element14.com/raspberryPi2.html?SKU=83T1943
>
> Accumulo + Hadoop + Raspberry Pi! :)
>
> Cheers,
> Chris
>
> On Jul 7, 2012, at 8:13 AM, David Medinets wrote:
>
>> http://www.fit-pc.com/web/fit-pc2-models/ - I wonder if there is any
>> reason why Hadoop & Accumulo would not work on this kind of computer.
>> Any opinions? Buying a few desktops for my home use will make my
>> office too hot to use.
>
>
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> Chris Mattmann, Ph.D.
> Senior Computer Scientist
> NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA
> Office: 171-266B, Mailstop: 171-246
> Email: [EMAIL PROTECTED]
> WWW:   http://sunset.usc.edu/~mattmann/
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> Adjunct Assistant Professor, Computer Science Department University of Southern California, Los Angeles, CA 90089 USA
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Chris Mattmann, Ph.D.
Senior Computer Scientist
NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA
Office: 171-266B, Mailstop: 171-246
Email: [EMAIL PROTECTED]
WWW:   http://sunset.usc.edu/~mattmann/
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Adjunct Assistant Professor, Computer Science Department
University of Southern California, Los Angeles, CA 90089 USA
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++