RESERVOIR has defined a Reference Architecture for a next generation of Infrastructure as a Service (IaaS) Clouds capable of dealing with new requirements such as service-orientation (services managed as a whole, automating the services provisioning and scalability, and guaranteeing Service Level Agreements), separation between infrastructure and services, use of Open/Standard specifications, virtualisation technology independent, support for site federation, which allows private, public and hybrid Clouds, security and isolation reinforcement and use of utility computing business models.

The joint IBM-SAP Stand at CEBIT 2009 demonstrated the early promise of RESERVOIR. The technological value-add of RESERVOIR has been validated in numerous peer-reviewed papers.

RESERVOIR’s technology solution and spin-outs are available in open source code. Access downloads here.

RESERVOIR’s infrastructure management functions are structured in different layers, each dealing with different aspects of the Cloud.

Service Manager (SM)

Responsible for the instantiation of service applications (controlling the Service Lifecycle) and dynamically requesting virtualised resources to the underlying layer (VEEM), aiming to avoid over/under provisioning and over-costs based on SLAs and business rules protection techniques.

Virtual Execution Environment Manager (VEEM)

Responsible for the placement of VEEs (Virtual Execution Environment, a generalisation of the Virtual Machine concept) into VEE Hosts (a generalisation of the concept of hypervisor or Java Virtual Machine Container manager).

Virtual Execution Environment Host (VEEH)

Represents a virtualised resource that can host a certain type of VEEs (for example a physical machine with a hypervisor controlling it, or a Java Virtual Service Container). VEEH is also responsible for adding to the virtualisation platform the necessary functions (virtual network management, image storage, image live-migration, etc.).

Key technologies developed by RESERVOIR

  • Migration enablement of both virtual machines and Virtual Java Service Containers across network and storage boundaries.
  • Distributed management of virtual infrastructures across sites supporting private, public and hybrid cloud architectures.
  • Algorithms for the allocation of resources to conform to SLA (Service Level Agreement) requirements.
  • The creation of a formal Service Definition Language to support service deployment and life cycle management across RESERVOIR sites.
  • Security mechanisms for the safe deployment and relocation of virtual machines across physical machines, and RESERVOIR sites.
  • The development of a business information model as well as business oriented payment and billing mechanisms to charge for resources used across one or more RESERVOIR sites.
  • Testbed development to benchmark performance of actual industrial use cases in a RESERVOIR environment.
  • Automated Service Lifecycle management for service provisioning and dynamic scalability.