[Fiware-cloud] FI-Ware WP4 (Cloud Hosting) weekly conf call -- 17-May-2011 minutes

[Alex Glikson]

All,

Please, see bellow minutes from today's conference call.
Please, let me know if you have any comments (or if you can't see the
formatting well).
Also, please, let me know if the person that will be responsible for action
items and deliverables mentioned below on behalf of your company is
different than the one I've indicated.

Thanks!
Alex

Attendees:
   IBM, TID, TRDF, UPM (missing: Intel, INRIA, FT)

Agenda:
   update on finalizing the role of each partner
   further meetings to work on the content of HLDesc(M2) deliverable
   review & update the list of functions and capabilities
   action items until next meeting, owners, schedule

Minutes:
   FT still has not representative in WP4 (was escalated to Juanjo)
   The exact role of INRIA is still not clear, due to inconsistency in the
   DoW (was escalated to Juanjo)
   We will have weekly 1-hour conference calls to work on the HLDesc
   towards the M2 milestone. it was agreed to use the same time slot as
   today (every Tuesday)
   It should be mentioned in the HLDesc that per Agile process, some of the
   capabilities described in the M2 deliverable may be identified as low
   priority during the backlog prioritization phases, and hence could be
   deferred or dropped
   The list of functions & capabilities was reviewed and updated. Also,
   owners were identified for each function/capability (see below).
   It was agreed that the goal for the upcoming week is to produce the
   first version of high-level description of each function/capability --
   few paragraphs per function/capability, plus the initial set of
   important terms which should be defined in the HLDesc (derived from the
   description of the function/capability).
   In order to let all the partners to review the outcome of what others
   have done *before* the next conf call, it was agreed that the deadline
   for providing the above content is Monday, 23/5, noon (CET).

Action Items:
   Schedule next meetings [Alex, done. All: please, confirm that you've
   received the invitation and can see the details properly]
   Follow-up with Juanjo regarding the role and representation of FT, as
   well as the role of INRIA [Alex, sent a note]
   Provide initial content and list of keywords/terms for each of the
   functions/capabilities [Andy, Marina, Alex, David, Fernando, Serge,
   Juaquin -- send to the mailing list by Monday, 23/5, noon (CET)]

List of functions/capabilities and owners (see more details below):
   Physical & virtual resource infrastructure -- Intel/Andy
   Resource management
      Note-level -- Intel/Andy + IBM/Marina
      Cluster-level -- IBM/Alex
      Cloud-level -- IBM/David
   Management fabric -- IBM/Alex
   Service management - TID/Fernando + IBM/David + FT/?? (??)
   PaaS -- TID/Fernando + INRIA/Denis?
   Cloud Proxy -- TRDF/Serge
   Interfaces -- UPM/Joaquin (portal), TID/Fernando (APIs)


Detailed list of functions and capabilities (working draft):
   Physical & virtual resource infrastructure
      Object store
      Block storage infrastructure
      Network infrastructure
      Image repository
   Resource management: Note-level, Cluster-level, Cloud-level
      Request orchestration and dispatching
      Capacity management & admission control (note: related to
      elasticity/SLOs aspects of service management)
      Placement optimization (for multiple targets)
      QoS management & resource allocation guarantees
      Resource reservation (e.g., for certain time period)
      Monitoring
   Management fabric
      communication, synchronization, health monitoring, etc
   Service management
      Request orchestration and dispatching
      Catalog
      Service definition & life cycle management
      Elasticity governance
      SLOs (management, enforcement)
      Monitoring
      Metering
   PaaS
      Application/service definition (extensions to OVF to support platform
      concepts).
      Deployment design (establish the concrete structure of the
      application deployment: number of VMs, software in each VM, final
      network structure, etc.
      Provisioning and deployment: Configuring and passing of information
      to the different elements in the PaaS to prepare and later deploy the
      application/service into the IaaS. During the deployment phase
      different components need to be orchestrated: the configuration
      interfaces of the services offered integrated into the Platform (e.g.
      database services), configuration of the monitoring and accounting
      system, configuration of the VM (drive the creation of the VM with
      the application components inside), etc.
      IaaS/NaaS control layer, which interacts with the IaaS provider to
      deploy the VM, install and configure the software into the different
      VMs.
      Adaptation and scalability of the application/services deployed.
      Based on rules defined by the service provider or to enforce the SLAs
      in place.
   Cloud Proxy
   Interfaces
      Admin vs User
      API vs Portal vs Toolkit