Scaling a VNF

About this task

Scale-in a VNF
The purpose of scaling-in a VNF is to optimize resource utilization through reducing unnecessary service capacity by deleting one or more VNF instances from a cluster.
Scale-in can be requested on a VNF instantiated as a cluster and running in the ”active” state. The number of VNFs instantiated in the cluster should also exceed the minimum allowed instances.
Scale-in request can be triggered either through the IBM® Telco Network Cloud Manager - Orchestration API, or automatically by IBM Telco Network Cloud Manager - Orchestration itself based on a scaling policy set for the VNF.
Figure 1. Scaling-in a VNF
Scale-out a VNF
The purpose of scaling-out a VNF is to increase service capacity to match the offered load by creating one or more additional VNF instances to a cluster.
Scale-out can be requested on a VNF instantiated as a cluster and running in the ”active” state. The number of VNFs instantiated in the cluster should also be smaller than the maximum allowed instances.
Scale-out request can be triggered either through the IBM Telco Network Cloud Manager - Orchestration API, or automatically by IBM Telco Network Cloud Manager - Orchestration itself based on a scaling policy set for the VNF.
Figure 2. Scaling-out a VNF

Figure 3. Healing sequence

Procedure

From the IBM Telco Network Cloud Manager - Orchestration GUI, trigger scaling-in, or scaling-out.

Similar to healing, scaling of a clustered element can be triggered by navigating to the “Topology” view of a network service and selecting the target element from the service topology.

Scaling is only available for clustered service elements and thus the selected element should be defined as a cluster. Once the target cluster is selected, the scaling sequence can be triggered by clicking either the “Scale In” or “Scale Out” button shown in the header of the right-hand panel.