[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Possible GSoC idea: Cloud images for bioinformatics.

Hi all,

Seems like a good idea to me; happy to have Debian-med cloud images in the Debian AWS account, and to put them into the AWS Marketplace (@$0, naturally).

The scripts we have used to generate Squeeze (and now Wheezy) images have been totally Ander's Ingemann's ec2debian-build-ami; the biggest missing component right now is the cloud-init and cfn-init package for AWS; a Wheezy backport is underway AFAIK (its too late to include into Wheezy, sadly). Other than that, if Debian-mid is just an additional set of packages, then we should be able to easily generate a suitable image.

Looking at the Task list http://debian-med.alioth.debian.org/tasks/; are you suggesting a Cloud image for each task, or some subset?

I'm conscious of the point release 6.0.7 due real soon now - I've spoken with Anders to tag a new point release of the ec2debian-build-ami script which fixes a number of issues since 6.0.6 (such as deleting the root EBS volume on Terminate in AWS). Beyond that I think we should now be looking to target Wheezy - is that a fair target, or would people like a Debian-med Squeeze AMI available as well?

(AWS AMI Maintainer)
PS: I am not on the debian-med list or soc-coordination; please CC me for my attention.
PPS: I am travelling to Seattle tonight for the next week, but will still be on email to some fashion, and will be around to generate 6.0.7 images as the point release rolls out. Contact details in db.debian.org.

On 22/02/2013 8:04 AM, Charles Plessy wrote:
Dear all,

2012 has been a year of progresses in Debian's support for clouds,
and Debian is planning to apply again to the Google Summer of Code.


I would like to propose a bottom-up project with the goal of setting an
infrastructure or workflow allowing one to build and use easily a Debian (Med)
image where given bioinformatics tools are available at a version selected by
the user (tapping in snapshots.debian.org or perhaps other repository within
our web of trust if necessary).

The project would be bottom-up in the sense that it would be up to the
prospective students to propose a sensible design, and that in the absence of
something convincing the project will not start (that is: I do not intend to
propose a top-down implementation design).

What do you think ?

Mobile: +61 422 166 708, Email: james_AT_rcpt.to

Reply to: