[one-users] Use cases for consideration

[Ignacio Martin Llorente]

Hi Sergio,

Both scenarios can be implemented. You may be interested in the work developed by:

- BIG Grid VM Working Group: http://lists.opennebula.org/pipermail/users-opennebula.org/2010-February/001459.html
- Batch Virtualization Project at CERN: http://indico.cern.ch/sessionDisplay.py?sessionId=15&slotId=0&confId=55893#2009-09-21

Regards
--
Ignacio M. Llorente, Full Professor (Catedratico): http://dsa-research.org/llorente
DSA Research Group:  web http://dsa-research.org and blog http://blog.dsa-research.org
OpenNebula Open Source Toolkit for Cloud Computing: http://www.OpenNebula.org
RESERVOIR European Project in Cloud Computing: http://www.reservoir-fp7.eu














On 11/03/2010, at 18:19, Sergio Maffioletti, Grid Computing Competence Center wrote:

> Dear All
> 
> I just joined the mailing list as I would like to find out whether OpenNebula could serve the purposes of our usecases.
> 
> We are going to start an explorative project trying to implement a solution for managing VMs to allow users to run complex scientific applications
> 
> Our main requirement is the possibility of integrating whatever VM management solution with the local resources (normally controlled by LRMS: Condor, SGE, PBS)
> 
> the first usecase is a sort of Local cloud:
> 
> we would like to integrate our own local resources (typically a Condor pool or an SGE cluster) and allow users to run applications in VMs (basically one VM per application).
> The overall solution should provide a user interface batch-like where users can create a job description (pretty much like in SGE,PBS or in any sort of grid).
> The solution should be able to identify which VM correspond to the user's request; deploy the VM on the local infrastructure, start it up and have it integrated as an additional execution node of the LRMS (being Condor or SGE).
> Finally the LRMS job is submitted (and hopefully executed within the started VM)
> 
> The system should also monitor the status of the LRMS job and of the VM; then, once the job is completed, stop the VM.
> The system should also be able to decide whether the VM should be stopped, paused or just left up for the next queued job.
> 
> Accounting should also be harmonized with the LRMS
> 
> 
> The second usecase is a sort of Hybrid cloud:
> 
> As an extension of the above mentioned case, it should be possible to integrate external clouds (like EC2) seamlessly.
> VM deployment and monitoring as well as accounting report should be adaptable to cope with this usecase
> Ideally the cloud should be used to host VM that, somehow, should be seen at the LRMS as just additional compute nodes
> 
> I would really appreciate some input and/or feedback on whether and how OpenNebula could provide the functionalities we are looking for,
> or knowing that I'm completely off the track :)
> 
> thanks in advance
> 
> Cheers
> Sergio :)
> 
> sergio.maffioletti at gc3.uzh.ch
> University of Zurich
> Winterthurerstrasse 190
> CH-8057 Zurich Switzerland
> Tel: +41 44 635 4222
> Fax: +41 44 635 6888
> _______________________________________________
> Users mailing list
> Users at lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org