Periodic Reporting for period 2 - NECOS (Novel Enablers for Cloud Slicing)
Berichtszeitraum: 2018-11-01 bis 2019-10-31
The NECOS approach is manifested in a new developed platform whose main distinguishing features are:
1. The Slice as a Service –- a new deployment model. A slice is a grouping of resources managed as a whole, and that can accommodate service components, independent of other slices.
2. Embedded algorithms for an optimal allocation of resources to slices in the cloud and networking infrastructure, to respond to the dynamic changes of the various service demands.
3. A management and orchestration approach making use of artificial intelligence techniques in order to tackle with the complexity of large-scale virtualized infrastructure environments.
4. Making reality the lightweight principle, in terms of small footprint components deployable on large number of small network and cloud devices at the edges of the network.
The NECOS demonstrators and validation prototypes are based on state of the art open software tools, which have been carefully selected. These tools have been used to design, integrate and deploy the management and orchestration design choices, systems and the APIs that have constituted the main research activity of the project. The NECOS approach has been validated in the context of implementations, using experimental environments like FIBRE, 5TONIC and Emulab.
The System Testing Plan envisaged five different demonstrators by means of which we carried out the measurement of the different testing KPIs.
The Multi-Slice/Tenant/Service (MUSTS) demonstrator is meant to exercise the following key features of NECOS: slice creation, slice decommission, slice monitoring, service deployment, service update, VIM heterogeneity, and elasticity upgrade (both vertical and horizontal). The KPIs obtained with this demonstrator are the Average slice provisioning time, the Average service provisioning time, CPU isolation, Average elasticity response time and the Monitoring-data availability.
The marketplace (MARK) demonstrator is meant to demonstrate the marketplace concept introduced in NECOS as a dynamic resource discovery mechanism that can cope with slices of significant size and multiple geographically distributed resource providers.
The Experiments with Large-scale Lightweight Service Slices (ELSA) demonstrator is meant to show the deployment of end-to-end Slices that will be utilised by a Tenant in order to host services consisting of a very large number of lightweight elements (i.e. Virtual Network Functions (VNFs) and vLinks) deployed at the Edge of the infrastructure.
The Machine-learning based orchestration of slices (MLO) demonstrator was created to show how machine learning algorithms can add value to slice orchestration.
Finally, the Wireless Slicing Services (WISE) demonstrator shows the NECOS LSDC capabilities in expanding the cloud-network slinging concept towards wireless network domains.
The five above described demonstrators have been released by means of an open software license and constitute one of the instruments by means of which the NECOS will sustain its impact beyond the project lifetime.
The slice may be built assembling parts of resources that belong to different administrative domains, as the NECOS architecture follows a multi-domain approach. To participate in the NECOS ecosystem, a given domain must implement the appropriate APIs to offer resources through a marketplace. The NECOS LSDC contacts the marketplace to decide, based on different criteria, which resources will constitute the slice parts. Once this is done, the LSDC takes control of the resources at each local domain through on-demand created VIM/WIM or shim objects. Such a process is one of the main distinguishing features of NECOS in respect to other slicing architectures, which require major software changes of large VIM/WIM platforms and/or peer-to-peer interaction between the (slicing-ready) resource orchestrators of participating domains.
The LSDC empowers a new service model – the Slice as a Service, by dynamically mapping service components to a slice. The enhanced management capabilities of the infrastructure creates slices on demand and slice management takes over the control of all the service components, virtualized network functions, and system programmability functions assigned to the slice, and (re)-configure them as appropriate to provide the end-to-end service.
The LSDC platform offers the ability to a specific cloud provider to federate his own infrastructure with other cloud providers with different configurations in order to realize virtualized services using the Slice as a Service concept. The users of the LSDC APIs and platform will be able to create virtual services that can span the merged cloud infrastructure offered by different cloud providers. This concept is not purely technical; it can also encompass business, cultural and geographical among other domains.