Re: Debian Med server available for Cloud Computing?
Stephen,
Amazon Web Services provides small grants (as compute credit)
through their education and research programs:
http://aws.amazon.com/education/
>From experience I have with our own work with Cloud BioLinux
(http://tinyurl.com/ycxuau2, http://tinyurl.com/kmdpuh) projects that
might not fit the funded categories exactly - like a linux distro -
still can be funded.
You basically need a very small amount (100$ or less) to create a VM
on amazon and maintain it for 3 or more years (their prices per GB per
month are very low).
Of course people that might want to run Debian Med apps on Amazon will
have to pay and rent compute time. But scientists for example who
have research funds might not have a problem using Amazon, and wanting
to pay for having stable and scalable compute available.
Plus Amazon is widely known, so I think you get much more exposure of
the debian-med VM you's put up there. Of course you can still have a
server with Eucalyptus for people who don't have the funds to rent
compute from Amazon, and since the two cloud platforms have compatible
APIs, there's not a problem to port applications between the two.
If anyone is in Boston, MA area in July 7-8, we have a codefest around
those topics, and would love to meet fellow hackers:
http://open-bio.org/wiki/Codefest_2010
best,
Ntino
On Fri, Feb 19, 2010 at 9:07 AM, Steffen Möller <steffen_moeller@gmx.de> wrote:
> Dear all,
>
> I am aware of at least two efforts to prepare images for the Amazon
> services to help the bioinformatics community. And I wish us to be at
> the front here. From the packaging aspect we are, but we have no
> (project-)funds to pay for the storage of the Image at Amazon for
> instance. In a way this may be fortunate, since this opens our eyes for
> alternatives.
>
> The alternative I see is to prepare an Eucalyptus server (see
> http://www.eucalyptus.org) and run it ourselves. We would then invite
> the various efforts to use that rather than Amazon's services to develop
> the reference bioinformatics image - well, we all know that it won't be
> a single big image but rather a collection of images. That site may then
> become a major source for the practical exchange of concepts for dealing
> with the packages we prepared.
>
> There was a notion to have a server for Debian Med when we did not have
> a use for it. Does that offer still hold? Would the machine be
> sufficiently big to accommodate for the demands that several parallel
> run instances would have (say 8 GB main memory, 200GB disk (we also want
> some data, right))? Or are here some philanthropic readers on this list
> to help out?
>
> I cannot tell exactly where we will end up with this effort. I could for
> instance well imagine that we will acquire more such servers over time
> at different institutes to develop some mixture between a clouds and a
> grid. We'll see.
>
> Many thanks and regards,
>
> Steffen
>
>
>
> --
> To UNSUBSCRIBE, email to debian-med-REQUEST@lists.debian.org
> with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
> Archive: 4B7E9B0D.1010109@gmx.de">http://lists.debian.org/4B7E9B0D.1010109@gmx.de
>
>
--
Konstantinos (Ntino) Krampis, Ph.D.
Bioinformatics Engineer
J.Craig Venter Institute
Cell. 540-209-1029
http://semanticlifescience.wordpress.com/
http://twitter.com/agbiotec
Reply to: