Resource consumption metrics from GCP

Proposers

• David Aguilar (Indra)
• Gerardo García (Telefónica)
• Francisco Javier Ramón (Telefónica)

Description

So far OSM is only able to collect resource consumption metrics from VMs deployed on Openstack-based VIMs and Azure. Since Release FOURTEEN the default monitoring architecture has been changed to one based on Apache Airflow and Prometheus, but Airflow's DAGs are only set up to collect resource consumption metrics from Openstack and Azure VMs. With this feature, the functionality related to resource consumption metrics will be extended to support VMs deployed on Google Cloud Platform. In addition, this functionality will allow metric-based closed-loop operations, such as auto-scaling, on VNFs deployed on GCP.

Demo or definition of done

Robot tests based on VNF metrics, such as SA01 and SA02, should pass using GCP as VIM once the feature is implemented.