x/browbeat

akrzos 14283768b4 Telemetry Workload 2.0

+ Adjust boot and persist instances from 200 to 1000 instances
+ Boot and persist instances attached to a network (1000 instances)
+ Neutron context plugin to persist network over entire browbeat run
+ Added docs for the plugins

Change-Id: I58802218f1e2201063cf9ec3f82efa71b28ac1a0

2017-02-06 20:46:21 -05:00

1.6 KiB

Raw Blame History

Plugins

Rally

Context - browbeat_persist_network

This context creates network resources that persist upon completation of a rally run. It is used in conjunction with the nova_boot_persist_with_network plugin scenario. Beware that removal of the network resources created by this context plugin can be a lengthy process so this is best used on "throw-away-test" clouds.

Scenario - nova_boot_persist

This scenario creates instances without a network that persist upon completation of a rally run. This scenario is best used for excerising the Telemetry systems within an OpenStack Cloud. Alternatively, it can be used to put idle instances on a cloud for other workloads to compuete for resources. The scenario is referenced in the telemetry Browbeat configurations in order to build a "stepped" workload that can be used to analyze telemetry performance and scalability.

Scenario - nova_boot_persist_with_network

This scenario creates instances that are attached to a network and persist upon completation of a rally run. This scenario is best used for excerising the Telemetry systems within an OpenStack Cloud. It increases the telemetry workload by creating more resources that the telemetry services must collect and process metrics over. Alternatively, it can be used to put idle instances on a cloud for other workloads to compuete for resources. The scenario is referenced in the telemetry Browbeat configurations in order to build a "stepped" workload that can be used to analyze telemetry scalability.

1.6 KiB Raw Blame History