I've thrown this guy together as a way for users to get a quick start with
Accumulo. It comes with hadoop, zookeeper, and accumulo preconfigured for a
single node. It uses BigTop for hadoop and zookeeper. It's a virtualbox
export which is set up for 4GB of memory. It does include init.d scripts
for quick start. That said, the init scripts require
Accumulo-1.4.2-SNAPSHOT, so that' why I'm only emailing dev. I will update
them when we release 1.4.2 for larger public consumption. accumulo user
password is secret, as is the root password for accumulo. Just boot it up
and wait a minute or 3 as processes start up. Firefox is configured for the
monitor page and namenode page, so you can easily pull it up and see the
state of your cluster.

If you do play with it, provide feedback. I want to make this as easily as
possible for a user who's new to Accumulo and wants to try it out without
having to configure all of the dependencies.

http://dl.dropbox.com/u/538523/accumulo.ova.gz

John

That's huge. Is there any way you can simply script what you did to
create this image, and share that, so somebody could execute it from
an internet-connected clean-install of a virtual machine? (something
like a kickstart file, or a cloudinit script?... or perhaps a single
RPM and/or DEB that grabs all the right stuff and turns a vanilla OS
install into an Accumulo system?).

On Thu, Aug 23, 2012 at 4:57 PM, John Vines <[EMAIL PROTECTED]> wrote:
> I've thrown this guy together as a way for users to get a quick start with
> Accumulo. It comes with hadoop, zookeeper, and accumulo preconfigured for a
> single node. It uses BigTop for hadoop and zookeeper. It's a virtualbox
> export which is set up for 4GB of memory. It does include init.d scripts
> for quick start. That said, the init scripts require
> Accumulo-1.4.2-SNAPSHOT, so that' why I'm only emailing dev. I will update
> them when we release 1.4.2 for larger public consumption. accumulo user
> password is secret, as is the root password for accumulo. Just boot it up
> and wait a minute or 3 as processes start up. Firefox is configured for the
> monitor page and namenode page, so you can easily pull it up and see the
> state of your cluster.
>
> If you do play with it, provide feedback. I want to make this as easily as
> possible for a user who's new to Accumulo and wants to try it out without
> having to configure all of the dependencies.
>
> http://dl.dropbox.com/u/538523/accumulo.ova.gz
>
> John

Also, it might be fun to create a Fedora "Spin"
(http://spins.fedoraproject.org/)... or the Ubuntu equivalent.

On Thu, Aug 23, 2012 at 6:18 PM, Christopher Tubbs <[EMAIL PROTECTED]> wrote:
> That's huge. Is there any way you can simply script what you did to
> create this image, and share that, so somebody could execute it from
> an internet-connected clean-install of a virtual machine? (something
> like a kickstart file, or a cloudinit script?... or perhaps a single
> RPM and/or DEB that grabs all the right stuff and turns a vanilla OS
> install into an Accumulo system?).
>
> On Thu, Aug 23, 2012 at 4:57 PM, John Vines <[EMAIL PROTECTED]> wrote:
>> I've thrown this guy together as a way for users to get a quick start with
>> Accumulo. It comes with hadoop, zookeeper, and accumulo preconfigured for a
>> single node. It uses BigTop for hadoop and zookeeper. It's a virtualbox
>> export which is set up for 4GB of memory. It does include init.d scripts
>> for quick start. That said, the init scripts require
>> Accumulo-1.4.2-SNAPSHOT, so that' why I'm only emailing dev. I will update
>> them when we release 1.4.2 for larger public consumption. accumulo user
>> password is secret, as is the root password for accumulo. Just boot it up
>> and wait a minute or 3 as processes start up. Firefox is configured for the
>> monitor page and namenode page, so you can easily pull it up and see the
>> state of your cluster.
>>
>> If you do play with it, provide feedback. I want to make this as easily as
>> possible for a user who's new to Accumulo and wants to try it out without
>> having to configure all of the dependencies.
>>
>> http://dl.dropbox.com/u/538523/accumulo.ova.gz
>>
>> John

If they have a hadoop one already, then it should be easy. If not, it's a
bit more time. I want to get the rcd setup scripts supporting
centos/redhat/fedora systems next. Elser can worry about gentoo.

Sent from my phone, so pardon the typos and brevity.
On Aug 23, 2012 6:20 PM, "Christopher Tubbs" <[EMAIL PROTECTED]> wrote:

> Also, it might be fun to create a Fedora "Spin"
> (http://spins.fedoraproject.org/)... or the Ubuntu equivalent.
>
> On Thu, Aug 23, 2012 at 6:18 PM, Christopher Tubbs <[EMAIL PROTECTED]>
> wrote:
> > That's huge. Is there any way you can simply script what you did to
> > create this image, and share that, so somebody could execute it from
> > an internet-connected clean-install of a virtual machine? (something
> > like a kickstart file, or a cloudinit script?... or perhaps a single
> > RPM and/or DEB that grabs all the right stuff and turns a vanilla OS
> > install into an Accumulo system?).
> >
> > On Thu, Aug 23, 2012 at 4:57 PM, John Vines <[EMAIL PROTECTED]> wrote:
> >> I've thrown this guy together as a way for users to get a quick start
> with
> >> Accumulo. It comes with hadoop, zookeeper, and accumulo preconfigured
> for a
> >> single node. It uses BigTop for hadoop and zookeeper. It's a virtualbox
> >> export which is set up for 4GB of memory. It does include init.d scripts
> >> for quick start. That said, the init scripts require
> >> Accumulo-1.4.2-SNAPSHOT, so that' why I'm only emailing dev. I will
> update
> >> them when we release 1.4.2 for larger public consumption. accumulo user
> >> password is secret, as is the root password for accumulo. Just boot it
> up
> >> and wait a minute or 3 as processes start up. Firefox is configured for
> the
> >> monitor page and namenode page, so you can easily pull it up and see the
> >> state of your cluster.
> >>
> >> If you do play with it, provide feedback. I want to make this as easily
> as
> >> possible for a user who's new to Accumulo and wants to try it out
> without
> >> having to configure all of the dependencies.
> >>
> >> http://dl.dropbox.com/u/538523/accumulo.ova.gz
> >>
> >> John
>

This is the first issue that I ran into. I am going to follow the
suggestions to resolve the issue.

I'm sending this email from inside the VM. The hadoop command works.
The accumulo executable does not seem to be in the path. Accumulo is
definitely running though. What accumulo userid and password should I
be using?

Sorry.. I now see you mentioned the userid and password.

-_-

A little old, but a good place to start for Gentoo:
https://github.com/joshelser/Gentoo-Ebuild-Overlay

On 08/23/2012 05:33 PM, John Vines wrote:
> If they have a hadoop one already, then it should be easy. If not, it's a
> bit more time. I want to get the rcd setup scripts supporting
> centos/redhat/fedora systems next. Elser can worry about gentoo.
>
> Sent from my phone, so pardon the typos and brevity.
> On Aug 23, 2012 6:20 PM, "Christopher Tubbs"<[EMAIL PROTECTED]>  wrote:
>
>> Also, it might be fun to create a Fedora "Spin"
>> (http://spins.fedoraproject.org/)... or the Ubuntu equivalent.
>>
>> On Thu, Aug 23, 2012 at 6:18 PM, Christopher Tubbs<[EMAIL PROTECTED]>
>> wrote:
>>> That's huge. Is there any way you can simply script what you did to
>>> create this image, and share that, so somebody could execute it from
>>> an internet-connected clean-install of a virtual machine? (something
>>> like a kickstart file, or a cloudinit script?... or perhaps a single
>>> RPM and/or DEB that grabs all the right stuff and turns a vanilla OS
>>> install into an Accumulo system?).
>>>
>>> On Thu, Aug 23, 2012 at 4:57 PM, John Vines<[EMAIL PROTECTED]>  wrote:
>>>> I've thrown this guy together as a way for users to get a quick start
>> with
>>>> Accumulo. It comes with hadoop, zookeeper, and accumulo preconfigured
>> for a
>>>> single node. It uses BigTop for hadoop and zookeeper. It's a virtualbox
>>>> export which is set up for 4GB of memory. It does include init.d scripts
>>>> for quick start. That said, the init scripts require
>>>> Accumulo-1.4.2-SNAPSHOT, so that' why I'm only emailing dev. I will
>> update
>>>> them when we release 1.4.2 for larger public consumption. accumulo user
>>>> password is secret, as is the root password for accumulo. Just boot it
>> up
>>>> and wait a minute or 3 as processes start up. Firefox is configured for
>> the
>>>> monitor page and namenode page, so you can easily pull it up and see the
>>>> state of your cluster.
>>>>
>>>> If you do play with it, provide feedback. I want to make this as easily
>> as
>>>> possible for a user who's new to Accumulo and wants to try it out
>> without
>>>> having to configure all of the dependencies.
>>>>
>>>> http://dl.dropbox.com/u/538523/accumulo.ova.gz
>>>>
>>>> John

To Christopher's first question, what about using Vagrant? You can then use
Chef or Puppet in solo/stand-alone modes to provision the VM, which also
might help some with the cross-platform issues.

Jared

On Thu, Aug 23, 2012 at 5:51 PM, David Medinets <[EMAIL PROTECTED]>wrote:

> Sorry.. I now see you mentioned the userid and password.
>