doytsujin / rally-plugins

Custom rally plugins for testing platform components

Rally plugins

Rally custom plugins for testing platform components.

Installation

To add custom rally plugins to rally plugin list, just pip the repo:

pip install .

Then you can check that custom plugins included in rally plugins list:

rally plugin list | grep <plugin_name>

Uninstall

To uninstall the repo, just pip uninstall it:

pip uninstall rally-plugins

Dependencies

rally-plugins depends on rally-openstack package.

Current state

Currently there are next services:

• service for Grafana, that push metric by pushgateway and check it in specified grafana datasource
• service for Kubernetes rally plugins

Next scenarios are implemented in rally-plugins:

• Grafana:
  • check metric pushed from nova instance by pushgateway
  • check metric pushed locally by pushgateway
• Elasticsearch:
  • check data about created nova instance stored
• Kubernetes:
  • create namespace and namespaced pod and delete it then
  • create namespace and namespaced replication controller and delete it then
  • list namespaces
  • create namespace and delete it then
  • create namespace and namespaced pod with emptyDir volume and delete it then
  • create namespace and namespaced pod with emptyDir volume, check it with check command and delete it then
  • create namespace, create secret and namespaced pod with secret volume and delete it then
  • create namespace, create secret and namespaced pod with secret volume, check it with check command and delete it then
  • create namespace and namespaced pod with hostPath volume and delete it then
  • create namespace and namespaced pod with hostPath volume, check it with check command and delete it then
  • create namespace, create PV and PVC, create namespaced pod with PVC and delete it then
  • create namespace, create PV and PVC, create namespaced pod with PVC, check it with check command and delete it then
  • create namespace, create configMap and namespaced pod with configMap volume and delete it then
  • create namespace, create configMap and namespaced pod with configMap volume, check it with check command and delete it then
  • create namespace and namespaced replicaset and delete it then
  • create namespace and namespaced replicaset, scale it with number of replicas, scale revert and delete replicaset then
  • create namespace and namespaced deployment and delete it then
  • create namespace and namespaced deployment, rollout it with changes and delete it then
  • create namespace and namespaced statefulset and delete it then
  • create namespace and namespaced statefulset, scale it with number of replicas, scale revert and delete statefulset then
  • create namespace and namespaced job with "Never" restart policy, wait for successful status and delete it then.
  • create namespace and namespaced daemonset, check that each node has the daemonset pod and delete daemonset then.
  • create namespace, namespaced pod, add clusterIP service with selector on pod, get pod endpoints and run job, which curl all these endpoints, then delete them all.
  • create namespace, namespaced pod, add clusterIP service without selector and with custom endpoints on pod, run job, which curl all these endpoints, then delete them all.
  • create namespace, namespaced pod, add nodePort service with selector on pod, get node port and requests pod with node ip and port, then delete them all.

For more details how to run and analyze tests, see docs\source directory.