RFC : Create separate section for neutron trunk workloads

[Sanjay Chari]

We have workloads related to trunks[1] under the "neutron" section in browbeat-config.yaml.
These workloads require a different value for times when compared to the other workloads in the same block.

For example, for create-list-network, we set times as 1000 and concurrency as 16 as a kind of Neutron API test.

On the other hand, we cannot have times that high for the trunk workloads. For example, in boot-server-and-batch-add-subports, with subports_per_batch=10 and batches=5, that would mean close to 50000 ports on the overcloud with times=1000. There is no major need to test ports at that scale usually, and the environment is likely to have issues at that scale.

Changing the number of subports per VM is possible, but since the subports are meant to mimic containers on Openshift nodes, having 10 subports per VM seems good.

I propose that we have a separate section in browbeat-config.yaml specifically for the workloads mentioned in [1], so that we can set times and concurrency separately for those workloads.

[1] https://github.com/cloud-bulldozer/browbeat/blob/master/browbeat-config.yaml#L250-L272

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.opendev.org/c/x/browbeat/+/865000