Optimize cv_facts execution.

[titom73]

Issue Type

• Performance Report reported by @noredistribution

Module Name

cv_facts

Summary

In a large environment, collecting facts from Cloudvision can take a long time increasing execution time of the overall playbook.

It has been seen in deployment with few hundreds of devices and thousands of configlets.

To reduce execution time, it might be useful to reduce number of API calls sent to Cloudvision.

2 options could be a good source of optimization:

• One option suggesting by @ankudinov is to use get /configlet/getConfigletsAndAssociatedMappers.do call to collect configlet and association in just one API call instead of current implementation described below:

• Second approach is to use get_configlets_by_device_id() in current cv_api201x wrapping provisioning/getConfigletsByNetElementId.do?netElementId API call.

def facts_configlets(module, facts):
    facts['configlets'] = []
    configlets = module.client.api.get_configlets()['data']
    # Reduce configlet data to required fields
    for configlet in configlets:
        MODULE_LOGGER.debug('  -> Working on %s', configlet['name'])
        # Get list of devices attached to configlet.
        configlet['devices'] = []
        applied_devices = module.client.api.get_devices_by_configlet(configlet['name'])
        for device in applied_devices['data']:
            configlet['devices'].append(device['hostName'])

        # Get list of containers attached to configlet.
        configlet['containers'] = []
        applied_containers = module.client.api.get_containers_by_configlet(configlet['name'])
        for container in applied_containers['data']:
            configlet['containers'].append(container['containerName'])

        # Add configlet to facts list
        facts['configlets'].append(configlet)

    return facts